`
`AAC
`Jo-.J'nals F\'.:iil.I'l.L1lg
`
`MINIHEUIEW
`
`Pharmacokinetic and Pharmacodynamic Considerations in
`Antimalarial Dose Optimization
`
`Nicholas J. White
`
`I"-.-"l 3|'..-Ll»: I '.-.~_--_! u;:..-'_ :7-.—:-:.—_-an.-;|'. -_i';-n_.F,.;u.
`
`".-.-.l-:_:r-:i'_=r:
`-_
`.:
`_
`..
`'.ru.Ic-'1 l-..r|-._.-._....'
`
`|_‘-.-' -:|
`
`'|-,'-I.- -lsl .‘-.-1:-lg.-_'ru~_.
`
`."-.]_1.'.:-_'I._u|
`
`,|'
`
`:-..r—_-:
`
`-'__1I'._.I
`
`-'.-I .TI"-5-Iar:-:"'Z-%:'{I% "cu Ti-I't_':I1.':l J-.-'l-:r;n;::'~+,CIu...-_'nIi I ;;u:::.I_.
`
`Antimalarial drugs have usually been first deployed in areas ofmalaria endemicity at doses which were too low, particularly for
`high—risk groups such as young children and pregnant women. This may accelerate the emergence and spread of resistance,
`thereby shortening the usefi.I.l life of the drug, but it is an inevitable consequence o fthe current imprecise method ofdose ii ud-
`ing. An alternative approach to dose finding is suggested in which phase 2 studies concentrate initially on phat-macokinetic—
`pharmacodynamic (PK~PD) characterization and in viva calibration of in vitra susceptibility information. PD assessment. is
`facilitated in malaria because serial parasite densities are readily assessed by microscopy, and at low densities by quantitative
`PCR, so that initial therapeutic responses can be quantitated accurately. If the in viva MIC could be characterized early in phase
`2 studies, it would provide a sound basis for the choice of dose in all target populations in subsequent combination treatments.
`Population PK assessments in phase 2b and phase 3 studies which characterize PK differences between different age groups, clin-
`ical disease states, and human populations can then be combined with the PK-PD observations to provide a sound evidence base
`for dose recommendations in different target groups.
`
`he primary objective of treating severe malaria is to save life.
`Other considerations such as preventing recrudescence or mi-
`nor toxicity are secondary. In uncomplicated malaria. the main
`objective of antimalarial drug treatment is cure of the infection.
`Speed of response is also important. as this reflects the rate at
`which the disease is controlled and the corresponding reduction
`in the risk of progression to severe malaria. Less—serious adverse
`effects therefore become a more important factor in determining
`dose. The therapeutic response in malaria is determined by the
`concentration profile (pharmacokinetics lPi=il_} of active antima-
`larial drug or drugs in the blood (as the asexual parasites which
`cause malaria pathology are confined to the blood). then-intrinsic
`pharmacodynamic {PD} properties, the susceptibility of the in-
`fecting parasites to the drugfsl, the number of asexual malaria
`parasites in the blood, and the activity of host-defense mecha-
`nisms. ldeally. antimalarial treatment should be 100% effective in
`everyone, but this may not be possible without producing toxicity
`or recommending a long course of treatment with consequent
`poor adlierence.
`It is now recommended that all antirnalarial
`treatments for uncomplicated malaria should aim at 21 .“>95% cure
`rate for the blood-stage infection (1). In recent years. a general
`agreement has been reached on methods ofclinical and parasito—
`logical assessment to measure the cure rates in cases of uncompli-
`cated falciparurn m:-ilaria (L-3). ln Plnsmodinm 1u'vu.\: and P. [I1-‘tilt?
`infections, persistent
`liver—stage parasites fhypnozoitesj cause
`later relapses, despite cure of the blood~stage infection, which
`complicates therapeutic assessment. These infections require ad-
`ditional treatment with 8—aminoquinoline5 [radical cure). Re-
`lapses are often genetically heterologous and cannot be distin-
`guished reliably from recrudescen ces or new infections. This
`necessitates a different approach for assessment of treatment eth-
`cacy in the relapsing rnalarias-wln'ch is yet to be agreed upon.
`Many ofthe antimalarial drugs in current use were introduced
`at suboptimal doses. For various reasons, quinine, sulfadoxine
`pyrimethamine, primaquine {for radical cure oftropical frequent
`relapsing P. v1'vnxinfections). mefloquine. halofantrine. artetnis-
`
`inin derivatives, artemether—lumefautrine, and dihydroartemis—
`inin—piperaquine (i.e., 7 of the 12 current antimalarials) were all
`deployed initially at doses which were too low in some or all age
`groups. Pyrimethamine and sulfadoitine doses for children were
`extrapolated from experience in Caucasian and Asian adults.
`Their pharrnacoltineticproperties were not studied in younger age
`groups before widespread deployment in Africa, where children
`are the main target group (4). The dose was too low in young
`children. The primaquine dose regimen [15 mg basefday adult
`dose} was developed largely on the basis of studies of the long-
`latency Korean vivax rnalaria. but this dose was then recom-
`mended widely in areas with the more resistant tropical relapse P.
`vfvnx phenotypes (5). in Southeast Asia and Oceania, this dose is
`too low. Five—day primaquine regimens were deployed very widely
`for radical cure ofvivax malaria for over 30 ye-ars—yet these reg-
`imens were largely ineffective. Fourteetrday courses are now rec-
`ommended. Mefloquine was first introduced at a single dose of] 5
`mg baseikg of body weight (-6, 7}. which may have hastened the
`emergence ofresistance (-8). The total dose now recommended is
`25 mgfkg divided over 2 or 3 days. The doses ofartemisinin deriv-
`atives used iuitially as monotherapy. and then subsequently in
`combination treatments (artemether at 1.6 mgfkgfdose in arte-
`rr1ether—lu1nefantrine and dihydroartemisinin at 2.5 rngfkg/‘dose
`togetherwith piperaquine}, may not provide rnaximal effectsirt all
`patients. The initial treatment regimen ofartemether—lun1efan—
`trine deployed was a four—dose regimen which provided insuffi-
`cient lL1mefant1'ine and gave high failure rates (six doses are now
`recommended) ('9). The dose of dihydroartemisinin in the first
`
`'_::n'-
`1._-[.I,;-r:._',.,'_|
`Published ahead of pn'nr E-
`»\.d _1Ir_-.
`-.
`-.-,:.-_.- I_:j-|junu:-._1I:_-'_...;._
`-,.'y'|.I1.'i_-_ "l'_ .-
`'.-:._ u._.|_3_ _
`'_ |'_--_-r_1'-_-n:_-:
`"_-_-
`
`|-_':_-.1.-;.tI|r_.:1:.j I][ul:‘[ 1|.-
`._-‘-.i|u.'v_e
`ii.._
`_-_o||ru'1.fr|,'«.:r_:__.l._
`
`
`
`
`
`.~.-.-1.- ii _'~:!A#.r' NI_-."
`
`
`
`
`
`isanfiliq,i_|,oz‘gtManneruo,rfSio'uJse'oee;;:d1J,uwon|‘.'J9pE0|UMOG
`
`5?92 aacssn1.org
`
`Antimicrobial hgents and Chemotherapy
`
`p. S?92—58O.7
`
`December 2013 Volume 5? Number 12
`
`InnoPharma Exhibit ‘lO30.000‘l
`
`
`
`Minireview
`
`formulations of the dihydroartemisinin-piperaquine combina-
`tion was <2 rnglkg (it is now 2.5 mg/kg, which may still be too
`low) [10]. The pharmacokinetic properties ofpiperaquine are dif-
`ferent in children from in adults, and there is evidence that current
`
`dosing schedules in children may be suboptimal (l 1}. After 3 cen-
`turies of reasonable dosing based originally upon the Schedula
`Roniana. treatment recommendations for quinine in severe ma—
`laria were suddenly reduced in the 19705 to a dose as low as 5
`niglkg,f.‘»'_4 h. which is eight times lower than that now recom—
`mended. In contrast. the quinine loading dose in severe malaria
`was not introduced until the early 1980s. and it is still not recom—
`mended universally (.12). The initial recommendation for artesu—
`nate treatment in severe cases ofmalaria was a daily maintenance
`dose ofltalfrhe initial dose (1.2 mgfkg}. As oral bioavailability is
`approximately 60%. this corresponds to an oral dose of 2 mglkg
`(1-3. 14}. The currently recommended parenteral close is twice this
`and is the same as the recommended initial dose, 2.4 mglkglday
`(I. 15.
`lei}. Recent evidence suggests that this dose should be in-
`creased in young children {l'7').
`Optimizing drug dosing requires characterization ofthe phar-
`niacokinetic and phartnacodyn-antic properties of the drug in the
`target populations. There are four main determinants of the ther-
`apeutic response: antimalarial pharniacokinetics {affected byvari-
`ables such as coadministration with food. age, pregnancy, disease
`severity, vital organ dysfunction. partner drug, and coinfections!
`other drugs), parasite susceptibility {incorporating effects on dif-
`ferent stages of asexual parasite development. dormancy, propen-
`sity for resistance to develop, and level ofresistan ce first selected],
`host defense (ilifluenced by age, pregnancy, and transmission in-
`tensitylexpostire history}. and parasite burden. In addition. mixed
`infections can be a Factor. For antimalarials. ex viva systems are
`useful for predicting resistance (18) and they provide valuable
`pharmacodynamic information (19), but they are simply not
`good enough yet to replace in viva evaluations for dose finding. In
`uncomplicated falciparum malaria,
`it
`is generally agreed that
`combinations, preferably,
`fixed—dose
`combinations
`[FDCL
`should be used. The same should apply to vivax malaria. although
`chloroquine and prirnaquine can be considered a combination.
`When drugs are first developed. there is -a limited window of op-
`portunity to define the dosevresponse (or concentration—effect__l
`relationship for the single new compound. but this opportunity
`must be taken (20). On ce the drug is available only as an FDC. the
`dose ratio is. by definition. fixed and it is too late for optimization
`ofthe individual component doses. Characterizing the individual
`drug dose— response relationships is essential for rational dose op-
`timization, and so a good drug development approach involves
`documenting the blood concentrations that are associated with
`5I.tl3l'l"lI-J..V{‘lt'I‘tal antimalarial effects. Studies in animal models. partic-
`ularly with P. jiilcipamrm, may be inforrnative. but studies in hu-
`mans will also be needed. it is important to accept that this may
`result in temporary therapeutic failures in some vol unteers. There
`is a natural reluctance to accept this. but sensitive detection meth-
`ods to measure low parasite densities now provide us with safe
`methods that should avoid any risk or discomfort to the patient
`(21). Suggestions are provided here for an alternative PK—PD ap-
`proach for dose finding which, ifvalidated, may improve and ac-
`celerate dose finding and so avoid systematic underprescribing
`and thus underdosing. It might also prove more rapid and less
`expensive. The primary objective is determination of the in viva
`MIC as the basis for rational dosing (the MIC is the concentration
`
`Total
`
`Detection limit
`
`parasites
`
`
`
`
`
`isanfiliq,i_|,oz‘gtManneruo,rf3.Io'uJse'oea;,r:diiumoi;pepeolumog
`
`Weeks
`
`FIG I Poplllation PK -PD responses lollowinga 3 —day Ircannenl with a l|ypoll1el—
`l( :1] slowly eliniinatvd a.ril'i1'n alarial drug. The total numbers olinalaria ptuasitcs in
`the body over time are depicted in blue in it range of patients presenting with
`parasite densities between apprrtcitruately 50 and 20U,000.",ul. The ranges nltlrug
`concentration profiles are sliowll in red. with the corresponding ranges {ill p-.1ra.si-
`tological rcsportscs in blue. Parasitelnia levels cannot be counted reliably by Ini
`croscopy below 50;’p.l [corresponding to -
`l[lll,00l).l)00 parasites in the body olan
`adult}. The M'PC is the lowest blood, plasma, or free plasma cm Icerltrnlion which
`produces Ille l11:L\‘ll11tIl1‘| pautslllcitlal eilecl {i.e.. Ill-Ell‘laJ{llI11ll11 parasite reduction
`n1n'ol.‘lbis corresponds to the conccntratlort associated willi firs] slowing ollhe
`lirst order (log linear) decline in pamsiteinia.
`
`at which the parasite multiplication factor per asexual cycle is 1). It
`is necessary first to consider the factors which affect the pharma-
`cokinetic properties of antim alarials in malaria and then to con-
`sider antimalarial pharmacodynarriics and how Pl'C—Pl) relation-
`ships should be assessed.
`
`PHARMACOKINETICS
`
`The pharrnacoltinetic (PK) properties of antimalarial drugs are
`often altered in patients with malaria compared with healthy sub-
`jects. The PK properties therefore change as the patient recovers.
`PK properties are also often significantly different in important
`patient subgroups such as young children and pregnant women
`(22). Several of the antimalarial drugs, notably those which are
`hydrophobic and lipopltilic. are poorly absorbed after oral or in—
`tram uscular administration and show wide interindividnal differ-
`
`ences in coiicentration profiles. In general. this variation in blood
`concentrations is inversely proportional to bioavailability, which
`emphasizes the importance of improving bioavail-ability in drug
`development. Increasing bioavailability provides the twin benefits
`of reducing the required dose and thus the cost ofthe drugs and
`reducing the individual probabilities of underdosing or overdos-
`ing. In considering antimalarial closing in the past, we tended to
`concentrate on mean or median values ofPK variables, but itis the
`
`patients with the lowest blood concentrations who are most likely
`to fail treatment and facilitate the emergence of resistance and
`those with the highest concentrations who are most likely to ex-
`perience drug toxicity (23). These extremes need to be defined,
`which means that characterizing the distributions of PK variables
`in important target groups is as important as assessing their mea-
`sures of central tendency (Fig. 1). Characterizing these distribu-
`tions well eventually requires sampling of relatively large numbers
`ofpatients, which in turn usually necessitates sparse sampling and
`population PK modeling. Optimal design approaches can be used
`to ensure that the information is gathered most efficiently (24). It
`
`December 2013 Volume 5}‘ Number ‘I2
`
`aac.asn'-torg 5?‘93
`
`lnnoPharma Exhibit 10300002
`
`
`
`Mirlirevlew
`
`is essential that key patient groups such as young children and
`pregnant women are studied specifically, and there should be a
`postregistration commitment to this if such investigations have
`not been conducted during preregistration studies. There may
`also be clinically relevant pharmacogenetic differences in drug
`metabolism between different ethnicgroups. Thus. characterizing
`the distributions ofpharrnacokinetic variables is a gradual process
`accrued during phase 2 and phase 3 of drug development, but it
`must continue into phase 4 to cover all relevant populations.
`Malaria is often worst in remote rural areas. The recent devel-
`
`opment ofsimple methodologies such as drug measurement from
`capillary blood filter paper samples (25. 26) will facilitate commu-
`nity—b:-ised assessments in remote settings and make sampling of
`infants and children feasible. Thus. population PK information
`will eventually be needed in all important target groups (i.e.. in-
`fants, cliildren, pregnant women. lactating women, malnourished
`patients, patients receiving antituberculosis lanti—TB] and antiret—
`roviral drugs. etc.) (22) to provide optimal dose recommenda—
`tions. There is currently limited bioanalytical capacity to support
`such studies. but there are international schemes to assist antima-
`
`larial drug measurement and ensure the accuracy of the results,
`which should facilitate future laboratory bioanalytical capacity
`development in tropical countries (27. 28}.
`in drug developcnent. where a new compound has not been
`used previously, there is little information on distributions of PK
`variables and so the important but difficult issue is to determine
`how much Pl{—PlJ information is enough to decide upon a dosage
`recommendation. For safety reasons, the PK information is usu-
`ally gathered in the following standard sequence: experimental
`animals. healthy normal volunteers. adult patients with uncom-
`plicated malaria. children, and. much later, infants and pregnant
`women.
`
`PHARMACODYNAMICS
`
`(i) Action oflhedrugs. The antimalarial drugs differ in their stage
`specificities of action against malaria parasites. The 8—an1ino—
`quinolines are unusual in killing pre—ei'ythrocy'tic—stage parasites,
`liypnozoites. and mature ganietocytes ofP.falciparrmr but having
`weak activity against its asexual stages ('29). They are more active
`against asexual stages of P. vivnx and P. lcmiwlesi. All other anti-
`malarial drugs in current use kill the asexual and sexual stages of
`sensitive P. vii-tract, P. mrtlnrirtc, P. ovals, and P. lcriowlcsi and the
`asexual stages and early gatnetocytes (stages 1 to lll_'J ofsensitive P.
`}'ulci'pttr‘tt.-n. but they do not kill the mature P. jitir.‘iptti'rrrn gameto-
`cytes (stage V). The artemisinins have a broader range ofeffect on
`developing P. ,-‘Erlcfpnrmrr sexual stages, as they also kill stage IV
`and younger stage Vgametocytes. Atovaquone and the antifols kill
`preerythrocytic stages and have spororitocidal activity in the mos-
`quito {interfering with oocyst formation and therefore blocking
`transmission). Apart from the 8—aminoquinolines, none of these
`drugs have significant effects on P. vivax or P. (mile hypnozoites.
`Even within the asexual cycle there are differences in antimalarial
`activity in relation to parasite developltierit. None ofthe currently
`used drugs have significant effects on very young ring stages or
`mature schizonts. and all have their greatest effects on mature
`trophozoites in the middle of the asexual Cycle (30). In addition,
`the artemisinins (and other antimalarial peroxides) have substan-
`tial ring—stage activity which underlies their life—saving benefit in
`treatment ofsevere falciparum malaria (15, 16. 31}. Several anti-
`rnalarials. notably. some antibiotics with antirnalarial activity.
`
`have greater effects in the second than in the first drugexposed
`asexual cycle (23). The pharmacokinetiopharmacodynamic rela-
`tionships {PI<—PD) have not been very well characterized for any
`ofthese activities.
`
`(ii) In vivo pharmacodynamic measures. in severe malaria.
`the primary therapeutic concern is the speed of parasite killing
`and, in particular, the killing of circulating ring—stage parasites
`before they mature and sequester (39, 31). Rapid killing ofyoung
`P. jnlcipumm parasites by artemisinin and its derivatives explains
`much of the superiority of artesunate over quinine in the treat~
`ment ofsevere falciparum malaria [15, 16). in uncomplicated ma-
`laria, rapid ring—forrn killing is also important, as it contributes to
`the speed ofpatient recovery, but the main tlierapeutic objective is
`to reduce para site multiplication. Once zuitimala rial treatment is
`started. then. after a va_1'i:-:ble lag phase, parasite killing in vivo
`approximates to a first-order process (32-34) as represented by
`the following equation:
`
`P, — n,.~.‘ tr
`
`(1 i
`
`where P, is the parasitemia level at any time I after starting treat-
`ment, PL, is the parasitemia level immediately before starting treat-
`ment. and kg, is the first—order parasite elimination rate constant.
`The parasite clearance halt? life is therefore 0.693119. In equation I.
`parasite killing equates with parasite removalfrom the circulation.
`but in falciparum malaria (but not the other malariasl there is an
`additional major factor removing parasites from the circulation.
`and that is cytoadherence. Only parasites in the first third of the
`asexual cycle circulate, and the more mature parasites are seques-
`tered. This complicates interpretation of the parasite clearance
`curve following treatment with drugs which do not kill ring-form
`parasites. as initial declines in parasitemia result mainly from se-
`questration ancl not drug effects (33). Parasite killing can be ex-
`pressed as the parasite reduction ratio {P1111}. which is the frac-
`tional reduction in parasite numbers per asexual cycle. or the
`reciprocal of ring-form kg. per cycle (32). This cancels out the
`effects ofcytoadherence, as the parasite populations are assessed at
`the same stages ofdevelopment separated by one cycle. The shape
`ofthe concentratiomeffect relationship in viva is assumed always
`to be sigmoid. as it is in vitro (Fig. 2), per the following equation:
`
`ll: _ kin-1.x' lcn/ECSHH l
`
`(-ml
`
`where it is the parasite killing rate and low is the maximum par-
`asite killing rate [i.e., the maximum effect, orfinml for that drug in
`that infection. C is the concentration of drug in blood or plasma.
`ECEU is the blood or plasma C0]1C€11lI_'l'ElT_'lDl1 resulting in 50% of the
`maximum effect, and r1 is a parameter defining the steepness ofthe
`dose—response relationship. For most drugs, maximum effects are
`probably achieved initially. The evidence for this is the lack of a
`relationship between peak concentrations and parasite clearance
`(the exception is quinine treatment of severe malaria without a
`loading dose, which provides subinaxinial effects in some pa~
`tients} (12). So while concentrations exceed the minimum para-
`siticidal concentration l_MPC).lc in equation I is equal to knm. It
`should be noted that each end ofthe sigmoid curve approaches 0%
`and 100% effects asymptotically—-—so the MPC is an approxima-
`tion, whereas the ECSU is a more robust and precise estimate. Once
`antimalarial concentrations in blood decline to a level below the
`
`MPC. the parasite killing rate declines (see the “Antimalarial pl1ar—
`macokinetic—pharrnacodynamic relationships" section below).
`For drugs in current use, maximum PRRs range from approxi-
`
`5?94 at-1Last1'I.arg
`
`Antlmicrobial Agents and Chemotherapy
`
`
`
`isanfisq,r_|,oz‘gtMenueruo,lfiJO'LUSE‘OEE;‘_.tId1J,l.|won|‘.'J9pE0|UMOG
`
`lnnoPharma Exhibit 10300003
`
`
`
`
`
`Leg concentration
`
`FIG 2 Tilt’ co|1cent'rat'ion ellect relationship; for antintztlarial drugs, the effect
`is parasite killing, which can be Ineastlred in tlifferettt ways. The F.I1Ifl.X is the
`maxirnttni parasite killing lliat :1 drug c:1n produce. wliiclt tmrislutes in vivo into
`the inaxiintmt parasite redui.'tinn ratio. The EC5", is the blood or plztsnta con
`cent ration providing 50% ot'n12n:i1n1nn killing. The median and range values
`for a ltyptithelical ptipltlat ion nfnialaria parasites are sltmvn in blue. and Ill:
`clislribtlliolt ufztverztge drug levels in patients is shown as a rcxl be|l—sl|apt-cl
`curve l i.e., cnnct-nt'raI:ions are log normally distributed}. Clearly, sonic of lhe
`patients ltave average drug levels below the MPG and would not have maxi
`nitltti responses with this close regimen.
`
`rnately 10-fold to approximately l0,0t.l0—fold reductions in para-
`sitemia per asexual cycle. The mean values and their variance in
`vivo have not been established for several important antimalarial
`drugs in current use (notably lumefantrine and piperaquine}. and
`for others. where rnonotherapies have been evaluated, the esti—
`mates are often imprecise. There is no evidence for saturation of
`parasite clearance, but, obviously, the higher the initial bio-
`mass, the longer it takes to eliminate all the parasites from the
`body (3-3). Consequently, patients with high—biomass infec-
`tions need more antimalarial drug exposure than those with
`low—bion1ass infections.
`
`(ii) In vitro susceptibility. For antimalarial effects, the shape
`and position ofthe concentration-effect curve studied ex viva de-
`pends on the susceptibility of the infecting parasites and the PD
`readout (typically, for blood stages. inhibition ofgrowth or mat-
`uration. inhibition of hypoxanthine uptake. inhibition ofprotein
`or nucleic acid synthesis, etc.). Furthermore, each in virro method
`assesses a slightly different section ofthe asexual life cycle, which
`may result in important differences between methods in the re-
`sults for drugs with ring—stage activity. It is not clear exactly how
`the effects ofthese static drug concentrations in a small volume of
`dilute blood in the laboratory correspond with in vivo effects I 1-8,
`19, 35). Neither is the relationship between inhibition of parasite
`growth and subsequent inhibition of multiplication well estab
`lished. inhibition of growth is measured in most in vitro tests,
`whereas in in viva patient studies, inhibition of multiplication
`(parasite clearance} is recorded. In the absence of in viva i nforma—
`tion on the concentratiomeffect relationship. for predictive mod-
`eling purposes the slopes ofthe linear segments of the in vitro and
`in viva sigmoid concentration-effect relationships have been as-
`sumed to be similar (8, 35), but whether or not such an assump-
`tion is justified remains to be established. Most agree that the
`antimalarial drug concentration that is biologically relevant in as-
`sessing blood~stage effects is the lun|.iound,l fraction in plasma.
`Total red cell concentrations are less informative as the parasitized
`
`Minireview
`
`red cells behave very differently from their unparasitized counter-
`parts. ln the patient, the blood concentrations of the antimalarial
`drug are changing constantly, and the parasite age distributions
`may differ considerably between patients. Ex vivo systems with
`changing antimalarial concentrations that are more biologically
`relevant than the simple static drug susceptibility assays have
`therefore been developed, and measurement ofmultiplication in-
`hibition can yield valuable information (19). Rodent models
`capable of sustaining human malaria infections have also been
`developed recently {S6}. Human malaria infections in irnmuno—
`deficient mice allow PK—PD characterization and thus provide
`useful information in predicting therape utic responses in patients.
`These laboratory studies have the great advantage that parasites
`from many different locations or with known resistance profiles
`can be studied and compared. It is argued below that if the rela-
`tionship between the standard in vin-in susceptibility measures
`l_'S0°/ii inhibitory concentrations [[C50I,
`lC._,.,_,, etc.) and in viva
`PK—PD responses in patients with malaria could be characterized.
`then this would facilitate dose finding.
`
`ANTIMALARIAL PHARMACOKINETIC-PHARMACODYNAMIC
`RELATIONSHIPS
`
`Some of the best research on antimalarial PK-PD relationships
`came from the period ofintense antimalarial drug investigation hi
`the United States during and shortly after the Second World War
`(Fig. 3]. Studies were conducted to determine tl1e optimum dos-
`ing strategies for mepacrine (atebrine, quinacrine), the Cinchona
`alkaloids, and both the 4- and S—arni:noquinolines (3-7-39). Phar-
`macokinetic analysis had yet to he invented, and methods for
`quantitation ofdrugs in serum or plasma were in their infancy, but
`the spectrophotometric assays that were conducted still provided
`valuable information. Relatively large numbers of nonimmune
`adult male volunteers artificially infected with single “strains" of
`P. frtlrriprtritrrt or P. vivax received dilterent dose regimens. serum
`levels were measured, and therapeutic responses were assessed.
`This research provided dose~response or concentration-effect re-
`lationships and led to the mepacrine loading—dose regimen. char-
`acterization of the comparative antinialarial effects of the four
`main Cinchona alkaloids (quinine, quinidine, cinclionine, and.
`cinchonidine), and development of the standard dosing regimen
`for chloroquine [one ofthe few antimalarial dose recommenda-
`tions which has stood the test of time). This was still the era of
`malaria therapy, and the war had focused military attention on
`malaria. Such volunteer studies are no longer possible today. Since
`that time, PK— PD relationships have been inferred mainly from
`clinical studies ofatltimalarial t1‘eatment [8, 9, 4043).
`(i) PK-PD correlates. Studies ofPK—PD relationships for anti-
`bacterial effects have shown that for some antibiotics (those with
`steep concentration-et'fect relationships and without postantibi~
`otic effects), bacterial killing is dependent on the duration for
`which the antibiotic exceeds the MIC for the bacterial population
`(“time above MIC"). For other antibiotics (where concentrations
`achieved with current regimens remain on the steep part ofthe
`concentration—effect relationship), it is the maximum concentra-
`tion achieved (Cum) or the related area under the plasma concen-
`tration—’cime curve (AUC) that is the best correlate of bacterial
`killing (Fig. 4}. These PK variables are all interrelated li.e., the
`higher the Cm“, the larger the All C and the longer the time above
`the MIC]. With some adjustments. these PK measures can be ap-
`plied to antimalarial effects {S2}, although correlates with parasite
`
`
`
`isanfiitq,r_|,oz‘gtiuenueruo,lfiJO'tUSE'OEEflZd1J,l.|moi;pepeoiumoq
`
`
`
`
`
`December 2013 Volume 5}‘ Number ‘I2
`
`aac.asn'-morg 5795
`
`InnoPharma Exhibit 10300004
`
`
`
`Minireview
`
`Plasma
`
`quininemg/L
`
`Effect
`
`0.1
`
`0.2
`
`0.3
`
`0.4
`
`0.5
`
`Daily dose {Grams}
`
`Plasma concentration {r'rIg!l.}
`
`‘l0
`
`WC 3 Dose—responsc relationsliips oblziinecl between lhe years [9115 and 19-16 for quinine in l3lDU(l-llIICll1L'|?(l vivrtx malaria il\‘icCoy strain} in volunteers (.53)-
`Plasrna cmlcenlrulinlis after protein precipiialirnl were menmrcd spectropliotonielricnlly, which overestirmiles parenl compouliti concenlralions. The left but
`shows the variable rt*la1'ionsl1ips he-tween dose and nicali plasma t'oI1cer1tral'ions. and the right graph shows the coliccritraljtiri effect‘ relal'iunsl1ip divided into
`three eff;-cl Iiieasurcs: class 1. no certain ellcct; class 1], lcinporary siipprcssioli of parasitt-min anclior fever: class III, “perInanem" effect.
`i.-2.. absence ol
`parasilenlia for M days.
`
`killing have not been established for most antirnalarial drugs.
`Whereas most bacteria replicate every 20 to 40 min, asexual rna~
`laria parasites infecting humans replicate every I to 3 days. Syrup»
`tornatic infections usually comprise one predominant brood of
`malaria parasites, but multiple genotypes are often present—par—
`ticularly in higher—transmi5sion settiI1gs—and so within one host
`there may be subpopulations with different drug susceptibilities
`(and alsfl diflerent stages of asexual development). The lowest
`blood. plasma, or free plasma concentration which produces the
`maxim Lll'I'l PRRis the MPC (Fig. 5}. These PK—PD variables reflect
`the antiparasitic effects of the antimalarial drug and host immu-
`nity and so are specific for an individual and that indi\«"idu.=.tl’s
`
`infection. Innate host—det'ense mechanisms and acquired irnni une
`responses contribute significantly to therapeutic responses—ef—
`fectively shifting dose—response curves to the left. The contribu-
`tion of the host immune response, which may be significant even
`in previously 110l1l.IIlIIllJl}€ patients 44). has not been well char-
`acterized.
`
`with current dosing for all antimalarial drugs except the arte~
`misinin derivatives. drug elimination is sufficiently slow that the
`antirnalarial efiects ofa treatment persist for longer than one asex—
`
`Concentration
`
`Weeks
`
`FIG 4 Plasma or blood concuntra Lion profile of a slowly eliniinalud a.nti.ma
`larial drug showingnn arbitrary MIC. The AUC is the area under the c11rve,ancl
`Cnlax is Ihe lI'lfl..\'ilIllIll'l colicenlration in blood or plasma. AUC from 7 days to
`infinity is shown in darker pink. Blond conccnI1"at_ions are increasingly mea-
`sured on day Sr’ in thcrapcutiu.‘ assessments of slowly eliminated antirnalarials
`H9).
`
`Total
`
`parasites
`
`Weeks
`
`FIG 5 l'Jif'i'ert-I11 tlieraipeulin: responses to .1 slowly eliminated anliilialarial
`drug in at rIla.l:iri:J infeclirnl of I01" p:1rasiles {parasite density,
`-2,(l(l(l,i|.Ll). The
`blood concciitration profile in gray is shown in the backgrouml. Parasitologi
`cal responses range from fully sensiriw {green} to highly re.sisia1il.(blw.‘}. Each
`respunse is associated wil h :1 dillerelil level of snsce}: Iibili ty and lhus a diilerenl
`MIC and MPC [arrows polrnlillg, to collceniraliorn profile}. The inset repre-
`sents the coriceiitratitsii ell}-cl. relatioiishjp tor lhc lowesl. level of resistance
`(resulting in it late failure). showing -:oi'respni'ICling points for the 1\‘flC and
`h-'iPC(or;1nge curve}.
`
`
`
`
`
`isanfiliq,i_|,oz‘gtMenueruojl3JO'LUSE'OEE[fZd11l.|moi;papeoiumog
`
`5?96 aacasrruorg
`
`Antimicrobial Agents and Chemotherapy
`
`lnnoPharma Exhibit ‘l030.0005
`
`
`
`ual cycle (3, 22, 32. 45). Indeed, many antimalarials have terminal
`elimination half—lives (r, mid) of several days or weeks. in order to
`cure the blood-stage infection in a nonimmune patient, antima-
`larial concentrations in blood {free plasma concentrations) must
`exceed the MIC for the infecting parasites until the last parasite is
`killed (8, 22, 23. .32. 45). The higher the initial parasite burden. the
`longer this takes. With host immunity. cure of malaria may be
`achieved even ifdrug levels fall below the MIC before complete
`elimination ofall parasites (44). Thus, the time above the MIC is
`an important PK determinant oftherapeutic outcome, although
`the AUC above the MIC is also relevant. as the rate of parasite
`killing is determined by the conceiitratiomeffect relationship
`above the MIC for the infecting parasites and by the antimalarial
`concentration profile in the treated patient.
`Assuming that the parasites are exposed to the antimalarial
`drug at a sensitive stage, what duration o