Journal of Curz1iu\'ux('ulur P/mrn1u('ulz);;_\'
`7532-537 (1;
`I985 Raven Press. New York
`
`Application of Pharmacokinetic-Pharmacodynamic
`Modelling for the Comparison of Quinazoline
`a—Adrenoceptor Agonists in
`Normotensive Volunteers
`
`Peter A. Meredith, Henry L. Elliott, Andrew W. Kelman, and John L. Reid
`
`University of Glasgow, Department of Materia Medica, Stobhill General Hospital. Glasgow, Smtlzind
`
`Summary: Prazosin. doxazosin, and trimazosin are
`structural analogues with relatively selective peripheral
`oi,-adrenoceptor antagonist properties. The relationships
`between the pharmacokinetics and pharmacodynamics
`for these three drugs have been studied following acute
`intravenous administration in normotensive subjects. The
`mean terminal elimination half-life (:SD) for prazosin
`was 2.0 1- 0.4 h. and that for trimazosin was comparable
`at 3.1 : 0.3 h, whilst the mean terminal elimination half-
`life for doxazosin was significantly longer at 9.4 : 1.5 h.
`The clearance of prazosin (mean, 327 1- 78 ml/min) was
`greater than that of both doxazosin (139 : 30 ml/min)
`
`and trimazosin (67 : 29 ml/min). The hypotensive effects
`of prazosin and doxazosin were fitted to an integrated
`pharmacokinetic—pharmacodynamic model with similar
`resulting values for the parameter describing responsive-
`ness. The pharmacodynamic profile of trimazosin was
`subjected to similar analysis and was most appropriately
`fitted to a model incorporating an effect compartment as-
`sociated with both parent drug and its major metabolite
`1-hydroxy trimazosin. Key Words: cx—Blockers—Quina-
`zoline derivatives—Pharmacokinetics—Effect model-
`ling.
`
`In recent years considerable effort has been de-
`voted to the use of pharmacokinetics to model drug
`disposition in the body and thereby to optimise
`dosage regimens. However, comparatively little at-
`tention has been focused on pharmacodynamic
`modelling and the interrelationship of the effect of
`a drug and its concentration in blood. Recently, a
`unified modelling approach has been developed to
`integrate kinetic and dynamic data to characterise
`the drug concentration—effect relationship in indi-
`vidual subjects. Using this technique, the pharma-
`cological effects of a number of drugs have been
`correlated with their pharmacokinetic properties—
`for example. the prolongation of the QT interval on
`the electrocardiogram in response to disopyramide
`(1) or quinidine (2), the change in the force of
`muscle concentration following administration of D-
`tubocurarine (3), and the improvement in respira-
`tory function in response to theophylline (4).
`In the management of hypertension the choice of
`
`Received September 10. 1984: revision accepted January 17,
`1985.
`Address correspondence and reprint requests to Dr. P. A.
`
`532
`
`drug is usually empirical and dose titration is by
`trial and error. The most commonly used antihy-
`pertensive drugs (B-adrenoceptor antagonists and
`thiazide diuretics) have flat dose—response curves,
`and for most other antihypertensives there is no di-
`rect simple relationship between plasma drug level
`and fall in blood pressure. However. for the a-ad-
`renoceptor antagonist prazosin such a relationship
`has been described following acute dosing in groups
`of normotensive and hypertensive subjects (5-7).
`More recent studies have demonstrated similar re-
`
`lationships in normotensive subjects for the other
`quinazoline oi—blockers, trimazosin (8) and dox-
`azosin (9).
`Although prazosin. trimazosin, and doxazosin
`are related quinazoline derivatives with closely sim-
`ilar chemical structures (Fig. 1) and peripheral (11-
`adrenoceptor antagonist properties, they have sig-
`nificant pharmacokinetic and metabolic differences
`and different profiles of hypotensive activity.
`
`Meredith at Department of Materia Medica. Stobhill General
`Hospital, Glasgow G2! 3UW. Scotland.
`
`Par Pharm., Inc.
`Exhibit 1 053
`
`
`
`Par Pharm., Inc. v. Novartis AG
`Case lPR201 6-00084
`
`EX. 1053-0001
`
`Ex. 1053-0001
`
`

`
`on-BLOCKERS AND EFFECT MODELLING
`
`533
`
`case
`
`CH?
`
`W30
`
`Cl-I30
`
`also
`
`C330
`
`W30
`
`3
`CH 0
`
`N
`
`N/—\N— A
`T \—/
`ii
`°
`
`/ N
`
`NH2
`
`0
`
`0
`
`_
`
`n
`
`N
`
`\—’
`
`c-—-[
`
`0
`
`ll
`0
`
`N\\'_
`
`~
`
`/
`NH?
`
`0130
`
`Ct-130
`
`N — c -——oCH
`N
`\—/
`H
`0
`
`on
`
`CH3
`(I:
`2: _.
`I
`CH3
`
`N\
`
`N
`
`/
`NH2
`
`CH3
`N—- c -—ocH—— tl: —oH
`2
`
`0
`
`cnlon
`
`N\
`
`N
`
`/ ”
`NH?
`
`This study compared the relationships between
`concentration and effect for these three drugs, fol-
`lowing acute dosing in normotensive subjects, using
`the integrated pharmacokinetic and pharmacody-
`namic modelling approach.
`
`METHODS
`
`Six healthy normotensive male subjects, aged 24-39
`years, gave informed written consent to participate in the
`study which was reviewed and approved by the Hospital
`Research and Ethical Committee. The study was con-
`ducted in the Clinical Pharmacology Research Unit, to
`which subjects reported at 9 a.m. on each study day
`having avoided alcohol, nicotine, and caffeine from 10
`p.m. the previous evening.
`In random order, on 4 separate study days at least 1
`week apart, subjects received an intravenous bolus injec-
`tion of prazosin (1 mg), doxazosin (1 mg), or trimazosin
`(100 mg), or an equivalent volume of 0.9% saline vehicle.
`Blood pressure was measured with a Roche Arteriosonde
`semiautomatic sphygmomanometer (model 1225), and
`corresponding heart rates were measured from precordial
`electrocardiographic leads displayed on a Grass poly-
`graph. Blood pressure was measured in duplicate after at
`least 10 min of rest supine, and then on standing for up
`to 5 min. These measurements were made at intervals up
`to 24 h after drug administration, and at corresponding
`times venous blood samples were withdrawn from an in-
`dwelling cannula for subsequent measurement of whole
`blood drug concentrations. Prazosin (10) and doxazosin
`(11) were assayed by high-pressure liquid chromatog-
`raphy with fluorescence detection. A similar assay for
`
`Prazosin
`
`Doxazosin
`
`Trimazosin
`
`1-Hydroxy
`
`Trimazosin
`
`FIG. 1. Chemical structures of the quin-
`azoline derivatives prazosin, doxazosin,
`and trimazosin.
`
`trimazosin (I2) simultaneously detected both parent drug
`and its major metabolite CP 23445 (Fig. 1).
`
`Pharmacokinetic analysis
`The pharmacokinetic analyses were carried out by
`computer-assisted nonlinear least squares fitting.
`In all
`subjects the disposition of the three drugs was most ap-
`propriately fitted, as assessed by the general linear test,
`to a two-compartment model described by the following
`equation:
`
`Cd“) : Ae—u(Il + BCABII)
`
`For trimazosin, in which metabolite (CP 23445) concen-
`trations were measured at the same time as the parent
`drug, the two-compartment model was extended to in-
`clude a third compartment to describe the profile of the
`metabolite disposition. The parent drug and metabolite
`were thus fitted simultaneously to Eqs.
`1 and 2. respec-
`tively.
`
`Cm“) =
`
`V.
`
`__A_
`(kmo " 00
`
`(e—um _ e—k,m,(/))
`klm [
`L —B<n _ —km.,u>]
`+ (km _ B) (e
`e
`)
`
`2
`
`)
`
`(
`
`The parameters derived from this approach were as fol-
`lows: the coefficients (A and B); the hybrid first-order
`rate constants for drug disposition ((1 and B); the first-
`order rate constant describing metabolite elimination
`(km); and the constant km, where VC and V", are the
`volumes of central and metabolite compartments. re-
`spectively.
`
`J Cardiovasr Plmrmural, Vol. 7, No. 3, I985
`
`EX. 1053-0002
`
`Ex. 1053-0002
`
`

`
`534
`
`P. A. MEREDITH ET AL.
`
`Pharmacodynamic modelling
`Using the effect model proposed by Sheiner et al. (3).
`the pharmacokinetic data were related to the placebo-
`corrected fall in systolic blood pressure (i.e.. the differ-
`ence in the systolic blood pressure after 5 min of standing
`following each active treatment and placebo administra-
`tion). The blood pressure was recorded after standing for
`5 min. With this method the standard pharmacokinetic
`model is augmented by an effect compartment that is
`deemed small enough not to influence the pharmacoki-
`netics and is governed by first-order processes, with rate
`constants kle and km. The measured effect—in this case,
`change in blood pressure—is then related to the concen-
`tration of drug in the effect compartment by nonlinear
`least squares fitting with equal weighting of the points.
`The three parameters derived from this procedure are m,
`i, and keq; where m represents the sensitivity to the drug
`(i.e., effect per unit of concentration in the effect com-
`partment), i is the intercept term from the equation re-
`lating drug concentrations to effect, and km is the first-
`order rate constant which characterises the concentra-
`tion—effect disequilibrium. In the case of trimazosin, the
`effect model was extended to include an effect compart-
`ment associated with the metabolite (8).
`
`Statistical analysis
`Comparison of results was made by application of Stu-
`dent’s t test for paired data with Bonferroni correction.
`
`RESULTS
`
`Pharmacokinetics
`
`From the same subject. representative whole
`blood c0ncentration—time profiles of prazosin, dox-
`azosin, and trimazosin (along with its metabolite CP
`23445) are shown in Figure 2. Table l summarises
`the parameters derived from the pharmacokinetic
`analysis, including the terminal elimination half-life
`(Bt.,,), the drug clearance from whole blood, and the
`steady-state volume of distribution. The Bt.,.s ob-
`tained in the six individuals, as detailed in Table 1,
`illustrate that the data from the subject shown in
`Figure 2 are representative of the group as a whole.
`The mean Btv, for prazosin was 2.0 1- 0.4 h; that
`for trimazosin was not statistically different at 3.1
`: 0.3 h. However, for doxazosin the mean (311,, was
`
`(ng/ml)
`Drug/MetaboliteConcentration
`
`
`o
`
`100
`
`200
`
`300
`
`400
`
`500
`
`600
`
`700
`
`Time (min)
`
`FIG. 2. Pharmacokinetic profiles of intravenous prazosin (1
`mg). doxazosin (1 mg), and trimazosin (100 mg) in a repre-
`sentative subject (subject 4).
`
`significantly (p < 0.005) longer, in all cases, at 9.4
`1 1.5 h. The elimination half-life (kmot./,) of 1.5 :
`0.5 h for trimazosin’s metabolite (CP 23445) was
`shorter, in all subjects, than the Bt.,. of the parent
`drug trimazosin. However, as shown by the con-
`centration—time profiles in Figure 2, the slope of
`the elimination phase for trimazosin’s metabolite
`appears to be parallel to that of trimazosin itself,
`indicating that the apparent elimination of CP 23445
`is dependent on its formation, not its elimination.
`In all subjects the clearance of prazosin was signif-
`icantly greater (p < 0.001) than that of the other
`two drugs, with a mean of 327 : 78 ml/min, com-
`pared with 139 : 30 and 67 : 29 ml/min for dox-
`azosin and trimazosin, respectively.
`
`Pharmacodynamics
`The differences in the profiles of the mean hy-
`potensive activity are illustrated in Figure 3 by the
`standing systolic pressures following administration
`of each of the drugs and placebo. With prazosin,
`the fall in blood pressure had a rapid onset—within
`
`TABLE 1. Pharma('()kineti(‘ parameters ofintravenoux prazosin. doxazosin, and trimazosin
`
`BI, ‘
`(hf
`
`Subject
`
`Prazosin
`
`Doxazosin
`
`Trimazosin
`
`1
`2
`3
`4
`5
`6
`Mean
`4.
`SD
`
`1.63
`2.18
`2.52
`1.63
`2.07
`2.18
`2.03
`..
`0.35
`
`7.67
`8.40
`9.53
`12.15
`9.52
`8.92
`9.37
`+
`1.53
`
`2.97
`2.27
`1.87
`2.()3
`4.43
`5.02
`3.10
`+
`1.33
`
`Clearance
`(ml/min)
`
`VJ“
`
`Prazosin
`
`Doxuzosin
`
`Trimuzusin
`
`Prazosin
`
`Doxazosin
`
`Trimazosin
`
`348
`201
`433
`338
`284
`356
`327
`+
`78
`
`185
`I34
`139
`94
`I28
`151
`139
`+
`30
`
`100
`83
`91
`58
`25
`44
`67
`+
`29
`
`42.6
`35.2
`77.7
`45.3
`46.8
`59.0
`51.1
`+
`15.1
`
`121.5
`94.8
`114.5
`96.4
`105.3
`115.2
`108.0
`:
`10.9
`
`19.3
`10.6
`12.2
`9.6
`8.2
`13.9
`12.3
`I
`4.0
`
`/~'.....h
`(h) '
`
`0.72
`1.38
`1.75
`1.43
`1.90
`1.95
`1.52
`+
`0.48
`
`[31, ‘. terminal elimination half-life: kmntl ‘. elimination half-life of metabolite; VA“. steady-slulc volume of distribution.
`" Trimazosin.
`
`J Curdi0vu.rr P/’1(1N7lll('0i, Vol. 7, No. 3, 1985
`
`Ex. 1053-0003
`
`Ex. 1053-0003
`
`

`
`ot-BLOCKERS AND EFFECT MODELLING
`
`535
`
`.
`FIG. 3. Mean systolic blood pressures
`(B_P) after 5 min of standing fol|owing_ ad-
`ministration of prazosin, doxazosin, trima—
`zosin, and placebo.
`
`Systolic BP
`(mm HQ)
`
`120
`
`110
`
`100
`
`.0
`
`
`
`,
`
`l
`
`/”
`
`\/
`
`I
`
`\
`l
`
`\/
`
`b
`I
`g
`:_‘- 52:23:: (mg)
`.,d°xaz°sin(1mg)
`._4 trimazosin (1oomg)
`
`
`1 20
`240
`360
`480
`600
`720
`
`Time (min)
`
`0.5 h—and was maximal within the first 2 h, with
`a return to baseline and placebo levels after 4 h.
`Following administration of doxazosin, there was a
`gradual onset of hypotensive effect, reaching a max-
`imum at ~6 h and persisting beyond 8 h. Following
`trimazosin treatment, there appeared to be a bi-
`phasic response: The maximum effect was again de-
`layed, to between 4 and 6 h, but there was a tran-
`sient initial blood pressure reduction within the
`1st h.
`
`Concentration—effect analysis
`For all three drugs, there was no simple linear
`relationship between drug concentrations in blood
`and fall in blood pressure. However, concentra-
`tion—effect analysis described a significant corre-
`lation between the placebo—corrected fall
`in
`standing systolic blood pressure and the concentra-
`tion of drug in an “effect compartment” or, in the
`case of trimazosin, two effect compartments—one
`associated with the parent drug and the other with
`its major metabolite. The parameters derived from
`this analysis are shown in Table 2, whilst fitted data
`for each of the three drugs in the same represen-
`tative subject are shown in Figure 4. In all individ-
`
`uals, satisfactory fits were obtained for each drug,
`as confirmed by the high values for the correlation
`coefficients. The intercept term i was included to
`achieve optimal fits, but the mean value for each
`drug was not significantly different from zero. Es-
`timates of the effect model parameter In or slope,
`which characterises the responsiveness to the drug,
`are shown in Table 2. As reflected in the doses ad-
`
`ministered, trimazosin was the least potent drug in
`terms of blood pressure fall per unit drug concen-
`tration in blood, with a mean value for m of -2.8
`-_r 1.8 mm Hg/p.g/ml. Using the same criterion, the
`responsiveness to prazosin and doxazosin was sim-
`ilar, with mean values for m of -2.0 : 0.7 and
`-2.3 1 0.4 mm Hg/ng/ml, respectively. In all sub
`jects, the parameter m for the metabolite of trima-
`zosin (CP 23445) was greater than that for the parent
`drug (mean value of metabolite was -23 : 20 mm
`Hg/pg/ml, compared with -2.8 mm Hg/pig/ml for
`parent drug). The rate constant keq characterises the
`time course of the effect following any rapid change
`in whole blood concentration of drug, or metabo-
`lite, and thus reflects the onset and offset of drug
`effect. In all subjects, the value of the rate constant
`keq was greater for prazosin (mean, 2.1 : 0.6 l/h)
`
`Prazosin
`
`TABLE 2. E/,‘f'm'I lI1()(/(’l [7(lI‘(llHt‘fl’I'.\'
`Doxazosin
`Trimazosin
`
`Subject
`
`Slope (rn)
`(mm Hg/ng/ml)
`
`I
`2
`3
`4
`5
`6
`Mean
`:
`SD
`
`-2.2l
`- l.5()
`-3.26
`— l.40
`— |.75
`- |.64
`- L96
`:
`0.70
`
`kw
`(l/
`
`l.50
`|.86
`|.74
`2.46
`3.|2
`l.74
`2.07
`:
`0.61
`
`Slope (m)
`(mm Hg/ng/ml)
`
`2.33
`2.36
`— l.84
`2.52
`|.95
`-
`— 2.79
`2.30
`:
`0.35
`
`kcq
`(I/h)
`
`0.55
`0.66
`0.84
`0.78
`0.57
`0.50
`0.65
`:
`0. I4
`
`Slope (m)
`(mm/pg/ml)
`
`-3.6|
`- |.28
`-3.52
`- |.94
`-0.94
`-5.62
`- 2.8
`:
`|.8
`
`km
`(I/h)
`
`0.48
`6.42
`6.42
`2.82
`2l.48
`6.66
`7.38
`:
`7.20
`
`km. first-order rate constant characterising Concentration—effect disequilibrium.
`
`CP 23445
`
`Slope (m)
`(mm Hg/pig/ml)
`
`-69
`— l0
`-8.3
`- 38.4
`-2.7
`-8.0
`- 23.0
`:
`20.0
`
`keq
`(1/h)
`
`0.22
`0.83
`1.03
`0.052
`2.33
`2.46
`l.l5
`:
`0.96
`
`J C(lI‘dl0|'(l.\‘(‘ PIizii'niziroI. Vol. 7, No. 3, 1985
`
`Ex. 1053-0004
`
`Ex. 1053-0004
`
`

`
`536
`
`P. A. MEREDITH ET AL.
`
`|.V. PRAZOSIN ( 1 mg)
`
`0
`
`was associated with the parent drug, whilst the
`later, sustained fall in blood pressure was associated
`with the metabolite CP 23445.
`
`DISCUSSION
`
`Modelling techniques used to relate pharmaco-
`logical effect to drug levels have now been de-
`scribed for a number of drugs (13). In the routine
`management of hypertension, little attempt is made
`to titrate the dosage for individual patients ac-
`cording to blood levels. There is some justification
`for this, as many hypertensive drugs have flat dose-
`response relationships, and in most cases no direct
`correlation has been consistently identified between
`fall in blood pressure and plasma drug level. The
`present study illustrates that the effect modelling
`technique may be applied to ouadrenoceptor antag-
`onists used in antihypertensive management. How-
`ever, it remains to be demonstrated whether or not
`the technique is appropriate for describing the ef-
`fects in hypertensive patients. In this application of
`effect modelling, the assumption was made that
`there was a linear relationship between the effect
`and the concentration of drug and metabolite.
`It
`might be predicted that the concentration—effect re-
`lationship would be more appropriately fitted to a
`sigmoid form such as would be described by the
`Hill or Langmuir equations appropriate for phar-
`macological studies. However, within the ranges of
`drug concentrations and hypotensive effects ob-
`served in this study of normotensive human sub-
`jects, there were no systematic deviations in the
`mathematical fits to suggest that the linear model
`was inappropriate. Furthermore, although the in-
`tercept parameter was included to achieve statisti-
`cally satisfactory fits, the mean values were not sig-
`nificantly different from zero.
`This study has also demonstrated that the mod-
`elling approach can be used to characterise differ-
`ences in the response profiles of structurally related
`drugs. The fall in blood pressure following admin-
`istration of prazosin was maximal within 1.5 h of
`intravenous administration, and by 6 h had returned
`to baseline values. This is consistent with a direct,
`simple plasma concentration—effect relationship, as
`has been identified in some studies (5.14). Such a
`direct, simple relationship cannot readily be applied
`to doxazosin. Following intravenous doxazosin, the
`maximal fall did not occur until 4 and 6 h after ad-
`
`ministration, and a significant blood pressure re-
`duction was sustained up to 10 h. Despite the rel-
`atively delayed onset of the hypotensive effect and
`the long duration of action, there is no evidence to
`suggest that an active metabolite contributes to the
`pharmacodynamic profile of doxazosin. With intra-
`venous trimazosin there was a biphasic hypotensive
`response—an initial modest fall within the 1st h and
`then a later and maximal fall at 4-6 h—which could
`
`EX. 1053-0005
`
`
`
`3‘ -5
`
`IEEA
`
`QU
`
`-10
`
`-15
`
`‘7;
`1U)
`4 -20
`
`-25
`
`o
`
`200
`
`400
`
`we
`
`we
`
`Time (min)
`
`l.V. DOXAZOSIN llmgl
`
`o
`
`-5
`
`-10
`
`-15
`
`-20
`
`o
`
`-5
`
`~10
`
`-15
`
`‘N
`
`U EE
`
`‘I
`A.Q
`.3
`
`E‘
`
`>'.
`U1
`<1
`
`f'\
`U’
`EE
`x.
`R$
`.2
`
`E'
`
`9.
`ll)
`<l
`
`0
`
`200
`
`400
`
`600
`
`800
`
`Time (min)
`
`l.V. TRIMAZOSIN (100mg)
`
`viio
`
`O
`
`200
`
`400
`
`600
`
`800
`
`Ti me ‘(mi n)
`FIG. 4. Fitted pharmacodynamic profiles of prazosin (A),
`doxazosin (B), and trimazosin (C) in a representative subject
`(subject 4). BP, blood pressure.
`
`than for doxazosin (mean, 0.66 : 0.14 1/h) (Table
`2). This reflects the more rapid onset and shorter
`duration of action of prazosin compared with dox-
`azosin. For trimazosin’s metabolite CP 23445 the
`
`mean keq was 1.15 : 0.96 1/h, larger than that for
`the parent trimazosin (mean, 7.4 1- 7.2 l/h). This
`suggests that the early fall
`in blood pressure fol-
`lowing the intravenous administration of trimazosin
`
`J Cardiovaxc Pharmacol, Vol. 7, No. 3. 1985
`
`Ex. 1053-0005
`
`

`
`ot-BLOCKERS AND EFFECT MODELLING
`
`537
`
`not readily be correlated with trimazosin blood
`levels. The effect modelling approach strongly sug-
`gests that the rapid onset, initial effect is associated
`with the parent drug, whilst the subsequent max-
`imal fall is mainly attributable to the metabolite CP
`23445.
`
`For each drug, a fairly wide range of values was
`derived for the effect model parameters. Whilst this
`obviously reflects the errors implicit
`in the mea-
`surements, it also reflects and characterises the an-
`ticipated interindividual variability of response.
`However,
`it
`is noteworthy that the derived slope
`parameter for prazosin in the individual subjects in
`this study was very similar to that interpolated from
`the mean data presented in the studies of Bateman
`et al. (5).
`Whilst this study has highlighted the value of an
`integrated approach to pharmacokinetic and phar-
`macodynamic modelling for comparing and con-
`trasting the properties of related drugs, such an ap-
`proach has potentially greater value in its extrapo-
`lation to steady-state oral
`therapy.
`In theory,
`parameters derived for the concentration—effect re-
`lationship from a single acute dose can be used to
`predict the effect at steady state. In practice, during
`long-term vasodilator therapy.
`there may be
`changes in the intensity of effect; however, the po-
`tential value of this approach in the management of
`individual patients with hypertension warrants fur-
`ther study.
`
`Acknowledgment: We thank Pfizer Central Research
`(U.K.) for support, and Mrs. S. Thomson for typing the
`manuscript.
`
`REFERENCES
`
`l. Whiting B, Holford NHG, Sheiner LB. Quantitative analysis
`of the disopyramide concentration effect relationship. Br J
`Clin Pharmuml l980;9:67—75.
`
`. Holford NHG, Coates PE, Guentest TW. Riegelman S,
`Sheiner LB. The effect of quinidine and its metabolites on
`the electrocardiogram and systolic time intervals: concentra-
`tion response relationships. Br J Clin Pharmm-ol
`l98l:ll:l87—95.
`. Sheiner LB, Starski DR. Vozeh S, Miller RD. Ham J. Si-
`multaneous modelling of pharmacokinetics and dynamics:
`application of d-tubocurarine. Clin Pharmacol Ther
`l979'.25:358—7l.
`. Whiting B, Kelman AW, Barclay J. Addis GJ. Modelling of
`theophylline response in individual patients with chronic
`bronchitis. Br J Clin Pharmacol I981 ;l2:48l—7.
`. Bateman DN, Hobbs DC, Twomey TM, Stevens EA, Raw-
`lins MD. Prazosin: pharmacokinetics and concentration ef-
`fect. Eur J Clin Phurmacol l979:l6:l77—8l.
`. Elliott HL. McLean K, Sumner DJ, Meredith PA. Reid IL.
`Immediate cardiovascular responses to oral prazosin: effect
`of concurrent beta blockers. C/in Plzurmm-al Thur
`1981 ;29:303—9.
`. Larochelle P, Souich P. Hamet P. Larocque P, Armstrong J.
`Prazosin plasma concentration and blood pressure reduc-
`tion. Hypertension l982;4:93— l0l.
`. Meredith PA, Kelman AW, Elliott HL. Reid IL. Pharmaco-
`kinetic and pharmacodynamic modelling of trimazosin and
`its major metabolite. J Pharmuvokinet Biopharm l983;||:
`323-5.
`Vincent J. Elliott HL. Meredith PA. Reid JL. Doxazosin. an
`alpha, adrenoceptor antagonist: pharmacokinetics and con-
`centration effect relationships. Br J Clin Pharmat-ul
`l983;l5:7l9-25.
`Yee YG, Rubin PC, Meffin P. Prazosin determination by high
`pressure liquid chromatography using fluorescence detec-
`tion. J Chromatogr l979;l72:3l3—8.
`. Rubin PC, Brunton J, Meredith PA. The determination of
`the vasodilator UK33274 by HPLC using fluorescence de-
`tection. J Chromatogr l980;22l:l93—5.
`Hughes MA, Meredith PA. Elliott HL. The determination of
`trimazosin and its metabolite, CP23445,
`in whole blood by
`high performance liquid chromatography using fluorescence
`detection. J Pharmacol Methods l984;l2:29—34.
`Whiting B, Kelman AW. The modelling of drug response.
`Clin S('1'l98O:59:3ll—5.
`Seidemann P, Grahnen A. Haglund K. Lindstrom B. von
`Bahr C. Prazosin dynamics in hypertension: relationship to
`plasma concentration. Clin Pharmaml Tlzer l98l;30:
`447-51.
`
`I0.
`
`12.
`
`13.
`
`14.
`
`J Cardiovaxc Plmrmacol, Vol. 7. No. 3, 1985
`
`Ex. 1053-0006
`
`Ex. 1053-0006