J. vet. Pharmacol. Therap. 26, 213–224, 2003.
RENAL AND CARDIOVASCULAR DRUGS
Pharmacokinetic/pharmacodynamic modelling of the disposition and effect of benazepril and benazeprilat in cats J. N. KING* M. MAURER & P. L. TOUTAINà *Novartis Animal Health Inc., CH-4002, Basel, Switzerland; Novartis Animal Health Inc., Centre de Recherches Sante´ Animale, CH-1566, St Aubin, Switzerland; àEcole Nationale Ve´te´rinaire de Toulouse, INRA, UMR181 de Physiopathologie et Toxicologie Experimentales, Toulouse, France
King, J. N., Maurer, M., Toutain, P. L. Pharmacokinetic/pharmacodynamic modelling of the disposition and effect of benazepril and benazeprilat in cats. J. vet. Pharmacol. Therap. 26, 213–224. The disposition and effect of benazepril and its active metabolite, benazeprilat, were evaluated in cats using a pharmacokinetic/pharmacodynamic model. Cats received single 1 mg/kg doses of intravenous 14C-benazeprilat and oral 14 C-benazepril.HCl, and single and repeat (eight daily) oral administrations of 0.25, 0.5 and 1.0 mg/kg nonlabelled benazepril.HCl. The pharmacokinetic endpoints were plasma concentrations of benazepril and benazeprilat, and recovery of radioactivity in faeces and urine. The pharmacodynamic endpoint was plasma angiotensin-converting enzyme (ACE) activity. Benazeprilat data were fitted to an equation corresponding to a singlecompartment model with a volume equal to the blood space (Vc ¼ 0.093 L/kg). Within this space, benazeprilat was bound nonlinearly to ACE, which was mainly tissular (89.4%) rather than circulating (10.6%). Free benazeprilat was eliminated quickly from the central compartment (t1/2 1.0 h; Cl 0.125 L/kg/h), elimination being principally biliary (85%) rather than urinary (15%). Nevertheless, inhibition of ACE was long-lasting (t1/2 16–23 h) due to high affinity binding of benazeprilat to ACE (Kd 3.5 mmol/L, IC50 4.3 mmol/L). Simulations using the model predict a lack of proportionality between dose of benazepril, plasma benazeprilat concentrations and effect due to the nonlinear binding of benazeprilat to ACE. For example, increasing the dose of benazepril (e.g. above 0.125 mg/kg q24 h) produced only small incremental inhibition of ACE (either peak effect or duration of action). (Paper received 15 November 2001; accepted for publication 23 December 2002) J. N. King, Novartis Animal Health Inc., Werk Rosental, Postfach, CH-4002, Basel, Switzerland. Tel.: +41 61 697 5126; fax: +41 61 697 7080; E-mail:
[email protected]
INTRODUCTION Benazepril is a nonsulphydryl inhibitor of the angiotensinconverting enzyme (ACE). Benazepril is an ester pro-drug which, after absorption from the gastrointestinal tract, is metabolized into its active metabolite, benazeprilat, which is a highly potent and specific inhibitor of ACE (Balfour & Goa, 1991; Webb et al., 1990). Angiotensin-converting enzyme inhibitors (ACEIs) may be useful in the management of several feline diseases, notably chronic renal insufficiency, congestive heart failure and hypertension. Data have been published describing the effect of single and repeated oral doses of benazepril on plasma ACE activity in cats (King et al., 1999). In that study, only conventional pharmacokinetic techniques were used, employing total plasma concentrations and fitting of decline phases to exponential 2003 Blackwell Publishing Ltd
equations. However, classical compartmental models are not directly relevant for ACEIs, whose kinetics can only be described optimally using models which take into account the (nonlinear) binding of the active moiety to ACE (both tissue and plasma components) as described previously (Lees et al., 1989; Toutain et al., 2000a). In the present paper, we describe the use of a ‘physiological pharmacokinetic/pharmacodynamic’ model in cats. First, data were fitted to an equation corresponding to the model using new data after single administration of intravenous (i.v.) benazeprilat and oral benazepril, and previously published results after single and repeated oral administration of different doses of benazepril (King et al., 1999). Secondly, we used the model to make simulations of plasma benazeprilat and ACE activity profiles after different dosage regimens for benazepril in cats in order to illustrate practical uses of the model. 213
214 J. N. King et al.
Faecal and urine samples were collected separately 8 h after administration of the test treatment, and then once daily in the morning prior to feeding (at 24, 48, 72, 96 and 120 h). In part II, cats received single oral doses of benazepril.HCl at dosages of 0.25 mg/kg (n ¼ 5), 0.5 mg/kg (n ¼ 5) and 1.0 mg/kg (n ¼ 5). The target dosages cover the clinically used range of benazepril.HCl in cats (0.25–1 mg/kg once per day). Blood samples were taken the day before ()24 h) and at 2, 4, 6, 8, 12, 24, 36, 48 and 72 h after administration of the test treatments. In a second phase, 3 months after the first, cats received oral benazepril.HCl once daily for eight consecutive days at doses of 0.25 mg/kg (n ¼ 6), 0.5 mg/kg (n ¼ 6) and 1.0 mg/ kg (n ¼ 6). Blood samples were taken prior to the trial (day 5) and at 2, 4, 6, 8, 12, 24, 36, 48, 72 and 96 h after administration of the last (eighth) dose. A total of 13 cats were used in both phases, allowing comparison of parameters with a single dose and at steady-state. Benazepril was administered as film-coated tablets containing 5 mg of benazepril.HCl (Fortekor 5, Novartis Animal Health Inc., Basel, Switzerland). The test treatment was dosed manually orally each morning to the cats after they had been offered 100 g of canned meat.
MATERIALS AND METHODS The biological phase of the study was conducted in two parts (I and II), performed under Swiss Federal permits after approval by ethical committees. Animals Domestic short hair cats were used. They were housed individually in cages in a climate controlled environment for the duration of the trials, and fed canned meat and dry food twice daily. In part I, the cats were housed in open metabolism cages to permit collection of faeces and urine. Water was available ad libitum. Health checks were made at the start, at least once per day during, and at the end of each part of the study. In part I, two female and two castrated male cats, weighing 2.8–4.2 kg and aged 1.5–2 years were used. In part II, nine female and 11 castrated male cats, weighing 3.1–6.3 kg and aged 1–4 years were used. Study design In part I, each cat received i.v. benazeprilat and oral benazepril (as the hydrochloride salt) as a two-part crossover with a washout period of 2 weeks. Benazepril and benazeprilat had been labelled previously with 14C (Fig. 1). 14C-benazeprilat (Mol Wt 396.45) was dissolved in 0.7% saline and 1% NaHCO3 (approximately 5 mg/g) and administered as a bolus i.v. injection into the left cephalic vein using a 0.8 mm-diameter needle. The dose administered was 2.7–4.2 mg, equivalent to a dosage of 0.94–1.03 mg/kg (1.62–1.77 MBq/kg). 14C-benazepril.HCl (Mol Wt 460.96) was diluted in lactose, placed into gelatine capsules and dosed manually via the oral route. The dose administered was 2.8–4.1 mg, equivalent to a dosage of 0.97–1.02 mg/kg (1.44–1.52 MBq/kg). The target dosage of 1 mg/kg benazepril.HCl is at the upper end of the clinically used range (0.25–1 mg/kg once per day). The morning feed was given approximately 30 min after i.v. benazeprilat and a few minutes after oral benazepril. Volumes of approximately 3 mL of blood were taken from either a left or right cephalic vein using needles and placed into glass tubes containing 143 USP units of lithium heparin (Vacutainer, Becton Dickinson, Meylan, France). The samples were taken pretest, and 3, 10 and 30 min and 1, 2, 4, 8, 12, 24, 36, 48, 72 and 96 h after i.v. benazeprilat. Samples were taken pretest, and 10 and 30 min and 1, 2, 4, 8, 12, 24, 36, 48, 72 and 96 h after oral benazepril.
(S) H * N
COOCH 2CH3
NH H N
Determination of plasma ACE activity In part II, the tubes containing blood were placed onto ice and centrifuged within 2 h at 2500 · g for 10 min, the plasma collected and stored at below )20 C for a maximum of 3 months before analysis. Plasma ACE activity was measured using a commercial kit based on the artificial substrate 3H-HipGly-Gly (ACE activity radioassay kit, Ventrex Labs, Portland, ME, USA), and validated in our laboratory (King et al., 1997). COOH (S)
O
COOH
HCl
C - labelled Benazepril.HCl
In both parts I and II, the tubes containing blood were placed onto ice and centrifuged within 2 h at 2500 · g for 10 min, the plasma collected and stored at below )20 C for a maximum of 3 months before analysis. Benazepril and benazeprilat are stable at this temperature for at least 1 year (Kaiser et al., 1987). A gas chromatography method with mass selective detection (GCMS) was used as previously described and validated in our laboratory (Kaiser et al., 1987; King et al., 1997). The limit of quantification of the assay (coefficients of variation for both intra- and inter-assay precision 80% at 5 ng/mL, 95.8% at 50 ng/mL and 95.6% at 500 ng/mL. The measured total binding of benazeprilat consists of the sum of specific binding to ACE and nonspecific binding to albumin and is therefore not a parameter but a concentrationdependent variable. Binding could not be determined more
218 J. N. King et al. Table 2. Pharmacokinetic parameters describing the disposition of benazeprilat after single and multiple (eight daily) oral administrations of benazepril.HCl at dosages of 0.25, 0.5 and 1 mg/kg (respectively, equivalent to 0.54, 1.08 and 2.17 lmol/kg as benazeprilat) Single administration (n ¼ 5)
Repeated administration (n ¼ 6)
Ratio repeat/single
Mean ± SD
Mean ± SD
Geometric mean (95% confidence interval) n ¼ 15 0.25–1 mg/kg P value
Parameters (units)
0.25 mg/kg
Vc/F (L/kg) Cl/F (L/kg/h) T1/2K10 (h) Bmax (ng/mL) Bmax (nmol/L) Pmax (nmol/L) Kd (ng/mL) Kd (nmol/L) Fcirc (%) T1/2Ka (h) AUCfree(0-inf) (ng h/mL) AUCtot(0-inf) (ng h/mL)
1.09 2.02 0.44 104 263 25.1 1.09 2.75 10.4 1.65 152 409
± ± ± ± ± ± ± ± ± ± ± ±
0.5 mg/kg
1.0 mg/kg
0.25 mg/kg
0.66 0.975 ± 0.412 1.30 ± 1.60 1.54 1.04 ± 0.24 0.825 ± 0.290 0.26 0.65 ± 0.20 0.90 ± 0.76 31 74 ± 11 71 ± 30 77 186 ± 28 178 ± 77 6.3 17.2 ± 5.3 18.2 ± 10 0.24 1.17 ± 0.31 1.39 ± 0.70 0.61 2.95 ± 0.78 3.50 ± 1.77 4.3 9.4 ± 2.8 9.9 ± 2.5 0.52 2.17 ± 0.99 2.81 ± 1.65 85 434 ± 108 1139 ± 354 128 647 ± 90 1351 ± 446
2.65 0.94 2.09 56 141 12.3 2.48 6.2 8.75 1.90 273 356
± ± ± ± ± ± ± ± ± ± ± ±
0.5 mg/kg
1.0 mg/kg
0.66 1.74 ± 0.67 1.95 ± 1.61 0.36 0.826 ± 0.280 0.622 ± 0.134 0.48 1.62 ± 0.75 2.07 ± 1.52 6.8 50 ± 13 61 ± 25 17 127 ± 32 155 ± 62 6.1 18.4 ± 10.9 15.3 ± 9.8 0.99 1.73 ± 0.43 1.78 ± 0.61 2.5 4.35 ± 1.07 4.48 ± 1.55 4.0 15.7 ± 11.6 8.9 ± 4.1 1.66 2.53 ± 1.65 2.55 ± 1.48 152 575 ± 200 1443 ± 363 147 715 ± 179 1563 ± 414
2.39 0.65 3.67 0.59 0.59 0.62 1.51 1.51 1.05 0.90
(1.6, 3.6) (0.49, 0.87) (2.3, 5.9) (0.49, 0.71) (0.49, 0.71) (0.35, 1.1) (1.2, 1.9) (1.2, 1.9) (0.64, 1.7) (0.57, 1.4)
0.0005 0.0071 0.0001 0.0001 0.0000 0.09 0.0019 0.0019 0.84 0.62
Plasma concentration (ng/mL)
Data were fitted to a monocompartmental model in all cases. Symbols as in Table 1.
Benazeprilat Benazepril
120
accurately at 5 ng/mL for both molecules because of limits of quantification of the assay.
100 80 60 40 20 0 0
12
24
36
48
Time (h)
Fig. 2. Mean ± SEM total plasma benazepril and benazeprilat concentrations following oral administration of 1.0 mg/kg benazepril.HCl at time 0 to four cats.
Binding parameters There were no significant (P > 0.2) differences in estimates for binding parameters after i.v. benazeprilat and oral benazepril (Table 1). The total binding capacity for benazeprilat to ACE (Bmax), expressed after scaling for volume of distribution (Vc), was similar after i.v. benazeprilat (202 ± 30 nmol/L) and oral benazepril (196 ± 53 nmol/L). Most of the ACE binding capacity was located in the tissues, the nontissular (termed circulating) binding capacity (Fcirc) was only 10.6 ± 2.6% of the total after i.v. benazeprilat and 10.2 ± 2.8% after oral benazepril. The absolute tissular-binding capacity (Bmax) was 18.2 ± 4.0 nmol/kg and the nontissular binding capacity (Pmax) was 1.89 ± 0.39 nmol/kg.
M1
F1
M2
F2
Mean ± SD
Faeces 0–8 h 8–24 h 24–48 h 48–72 h 72–96 h 6–120 h 96–120 h (corrected)* Urine 0–8 h 8–24 h 24–48 h 48–72 h 48–72 h (corrected)*
0 56.5 56.5 82.7 82.9 82.9 86.9
0 12.2 46.6 46.6 47.2 47.2 74.6
0 0 78.3 85.6 86.1 86.1 87.6
0 0 80.5 81.5 81.7 81.7 84.5
0 17.2 65.5 74.1 74.5 74.5 83.4
± ± ± ± ± ± ±
0 13.4 8.3 9.2 9.1 9.1 3.0
7.6 12.1 12.5 12.5 13.1
14.8 15.2 16.0 16.1 25.4
11.7 12.0 12.1 12.2 12.4
14.0 14.5 14.8 15.0 15.5
12.0 13.5 13.9 13.9 16.6
± ± ± ± ±
1.6 0.8 0.9 1.0 3.0
Total faeces + urine (0–120 h)
95.5
63.3
98.3
96.6
88.4 ± 8.4
Table 3. The cumulative excretion of radioactivity, expressed as a percentage of the administered dose, recovered in faeces and urine in four cats after bolus i.v. injection of 1 mg/kg 14C-benazeprilat (M ¼ male, F ¼ female)
*Corrected so that total recovery in faeces and urine from 0–120 h in each cat was 100% of the administered dose. 2003 Blackwell Publishing Ltd, J. vet. Pharmacol. Therap. 26, 213–224
PK/PD modelling of benazepril in cats 219 Table 4. The cumulative excretion of radioactivity, expressed as a percentage of the administered dose, recovered in faeces and urine in four cats after oral administration of 1 mg/kg 14C-benazepril.HCl (M ¼ male, F ¼ female)
M1
F1
M2
F2
0 0 97.1 98.2 99.0 99.0 96.0
0 0 61.7 62.8 63.2 63.2 96.3
0 82.3 82.3 104.6 105.5 105.5 96.6
0 60.4 60.4 137.1 137.6 137.6 97.0
0 35.7 75.4 100.7 101.3 101.3 96.5
± ± ± ± ± ± ±
0 21.1 8.8 15.2 15.3 15.3 0.2
0 3.6 4.1 4.1 4.0
1.9 2.2 2.4 2.4 3.7
1.8 3.4 3.6 3.7 3.4
0 3.9 4.1 4.2 3.0
0.9 3.3 3.5 3.6 3.5
± ± ± ± ±
0.5 0.4 0.4 0.4 0.2
103.1
65.6
109.2
141.9
Faeces 0–8 h 8–24 h 24–48 h 48–72 h 72–96 h 96–120 h 96–120 h (corrected)* Urine 0–8 h 8–24 h 24–48 h 48–72 h 48–72 h (corrected)* Total urine + faeces (0–120 h)
Mean ± SD
105.0 ± 15.6
*Corrected so that total recovery in faeces and urine from 0–120 h in each cat was 100% of the administered dose.
Assuming an even distribution in the plasma and in the entire volume of distribution, the estimated plasma ACE molar concentration (Pmax) was 21.5 nmol/L (i.v. benazeprilat) and 19.2 nmol/L (oral benazepril). The measurable plasma benazeprilat concentration corresponding to a free plasma concentration equal to Kd was determined from the following equation: Measurable concentration ¼ Kd in vivo + fcirc (Bmax/2) From Table 1, the Kd in vivo ¼ 1.40 ng/mL, fcirc ¼ 0.106 and Bmax ¼ 80 ng/mL. Thus, the measurable concentration for which half ACE is saturated was about 5.6 ng/mL i.e. the
concentration observed at approximately 24 h after benazepril administration (see Fig. 2). The amount of benazeprilat necessary to saturate half of the ACE pool is: Amount ¼ Vc(Free + (BmaxFree/(Kd + Free))) with Free, the free concentration equal to Kd thus: Amount ¼ Vc(Kd + Bmax/2) Using the mean values for Vc (0.093 L/kg), Kd (1.40 ng/mL) and Bmax (80.0 ng/mL) in Table 1 yields a value of 3.85 lg/kg benazeprilat to saturate half of the ACE pool.
Table 5. Pharmacodynamic parameters describing the inhibition of plasma ACE in cats with benazeprilat after single and repeated (eight daily) oral administrations of benazepril.HCl at dosages of 0.25, 0.5 and 1.0 mg/kg. Results are n¼4 or 5 (single) or n¼6 per group (multiple). Data from both single and repeated administrations in the same animal were available for 13 cats
Dose Benazepril.HCl (mg/kg)
Parameter
Units
Parameter
Single Mean ± SD
Repeated Mean ± SD
0.25
E0 IC50 n
Units/mL Ng/mL
E0 IC50 n
61 ± 2.2 0.21 ± 0.057 1.86 ± 0.53
49 ± 2.1 1.05 ± 0.40 3.69 ± 2.51
0.5
E0 IC50 n
Units/mL Ng/mL
E0 IC50 n
68 ± 4.8 0.33 ± 0.075 1.72 ± 0.24
54 ± 7.7 0.65 ± 0.42 2.05 ± 0.79
1.0
E0 IC50 n
Units/mL Ng/mL
E0 IC50 n
65 ± 5.7 0.32 ± 0.11 1.62 ± 0.52
48 ± 6.9 0.82 ± 0.44 2.3 ± 0.25
E0 IC50 n
Units/mL Ng/mL
E0 IC50 n
Ratio single/repeat Geometric mean (95% confidence interval)
P value
65 ± 5.7 48 ± 6.9 0.78 (0.73, 0.83) 0.0001 0.32 ± 0.11 0.82 ± 0.44 2.27 (1.29, 4.01) 0.0088 1.62 ± 0.52 2.3 ± 0.25 1.44 (1.10, 1.89) 0.012 The data were fitted with an Imax inhibitory model: E ¼ E0 1 ðCn =ðICn50 þ Cn ÞÞ . E0 is the control activity; IC50 is the free benazeprilat concentration giving half the maximum inhibition; n is the Hill factor (slope). IC50 values expressed in terms of measurable (total) plasma benazeprilat concentration were respectively: 1.76 ± 0.57, 1.80 ± 0.43 and 1.51 ± 0.33 ng/mL for 0.25, 0.5 and 1.0 mg/kg dosages of benazepril.HCl respectively after single administration; and 2.48 ± 1.07, 2.14 ± 0.75, and 2.19 ± 0.38 ng/mL for 0.25, 0.5 and 1.0 mg/kg dosages after repeated administration. For benazeprilat, 1 ng/mL ¼ 2.52 mmol/L. 0.25–1, n¼13
2003 Blackwell Publishing Ltd, J. vet. Pharmacol. Therap. 26, 213–224
220 J. N. King et al.
The amount of benazeprilat necessary to saturate all of the ACE pool can be calculated as: Amount ¼ Vc (Free + Bmax) with a free plasma benazeprilat concentration largely higher than Kd (e.g. 14 ng/mL i.e. 10 times Kd). Using the mean values for Vc (0.093 L/kg) and Bmax (80.0 ng/mL) in Table 1 yields a value of 8.7 lg/kg benazeprilat to saturate the entire ACE pool. Therefore, even with a bioavailability of 3–5%, a single oral dose of 250 lg/kg (0.25 mg/kg) benazepril.HCl in cats will still lead to saturation of the entire pool of ACE with benazeprilat. Part II Single administration The calculated clearance values (Cl/F) were lower in this part of the study (0.83–2.0 L/kg/h) compared with those in part I (4.7 L/kg/h). Therefore, the AUC values (both free and total) for plasma benazeprilat were higher in part II than in part I (Tables 1 and 2). Repeated administration After repeated administration of benazepril, the clearance (Cl/F) was significantly lower and the elimination rate of the free fraction (t1/2K10) was longer as compared with after single administration (Table 2). There was moderate but significant bioaccumulation of benazeprilat with repeated administration, with an R value (assessed from the ratio of AUC[0 fi 24 h] of total benazeprilat after repeat/single application) of 1.3 (1.2, 1.5). Several of the binding parameters for benazeprilat changed with repeated administration. Repeated administration led to significantly lower values for Bmax and significance was approached (P ¼ 0.09) for Pmax (Table 2). Values for IC50 and n were also significantly larger, and E0 was lower with repeated compared with single administration (Table 5). Simulations Simulations of plasma benazeprilat concentrations and ACE activity after single oral benazepril.HCl dosages (0.0625–1 mg/kg) are shown in Fig. 3. The simulations illustrate the lack of direct correlation between total plasma concentrations and effect, and the lack of dose proportionality between dose of benazepril administered and plasma benazeprilat concentrations or effect. In particular, there was minimal increase in effect (on plasma ACE) at dosages above 0.125 mg/kg, although plasma concentrations continued to increase. Figures 4, 5 and 6 illustrate the predicted pharmacodynamic effect of 0.0315–0.5 mg/kg benazepril.HCl administered for 12 days either once daily, every 12 h, or every 48 h. With once daily dosing (Fig. 4), there was negligible increase in effect at steady-state with dosages above 0.125 mg/kg, although the time to reach steady-state effect was slower with 0.125 mg/kg (72 h) compared with 0.25 mg/kg (48 h) or 0.5 mg/kg (24 h). Lower dosages (0.0315 and 0.0625 mg/kg) produced dose-related lesser inhibition of plasma ACE at steady-state and longer times to reach steady-state.
Fig. 3. Dose–effect relationship for benazepril in cats. Data were simulated for single oral dosages of 0.0625, 0.125, 0.25, 0.5 and 1 mg/kg benazepril.HCl using mean parameters in six cats after a single dose of 0.5 mg/kg benazepril.HCl (see Tables 2 and 5). Upper panel: semilogarithmic plot of total benazeprilat plasma concentrations vs. time. Lower panel: effect of benazepril on plasma ACE activity. Effects were simulated using free plasma concentrations. Visual inspection of the figures reveals the lack of proportionality between dose, plasma concentrations and effect. For example, although plasma concentrations increase with higher dosages, the effect is similar for dosages of 0.125–1.0 mg/kg.
Fig. 4. Plots of predicted effects on plasma ACE activity in cats of 0.0315, 0.0625, 0.125, 0.25 and 0.5 mg/kg benazepril.HCl administered every 24 h for 12 days (12 administrations). Data were simulated using free plasma concentrations from mean data in six cats after repeated administration of 0.5 mg/kg benazepril.HCl (Tables 2 and 5).
With dosing every 48 h, there was a marked loss of effect at trough as compared with once daily dosing, and the efficacy at 2003 Blackwell Publishing Ltd, J. vet. Pharmacol. Therap. 26, 213–224
PK/PD modelling of benazepril in cats 221
Fig. 5. Plots of predicted effects on plasma ACE activity in cats of 0.0315, 0.0625, 0.125, 0.25 and 0.5 mg/kg benazepril.HCl administered every 48 h for 12 days (six administrations). Data were simulated using free plasma concentrations from mean data in six cats after repeated administration of 0.5 mg/kg benazepril.HCl (Tables 2 and 5).
Fig. 6. Plots of predicted effects on plasma ACE activity in cats of 0.0315, 0.0625, 0.125, 0.25 and 0.5 mg/kg benazepril HCL administered every 12 h for 12 days (24 administrations). Data were simulated using free plasma concentrations from mean data in six cats after repeated administration of 0.5 mg/kg benazepril HCL (Tables 2 and 5).
trough was not improved by increasing the dosage (Fig. 5). At steady-state with dosing every 48 h, there was minimal increase in effect with dosages above 0.25 mg/kg. Dosing every 12 h produced less variation in peak and trough effects compared with dosing every 24 or 48 h (Fig. 6). At steady-state there was minimal increase in effect with dosages above 0.0625 mg/kg, and marked inhibition of ACE was achieved with the lowest dosage (0.0315 mg/kg). However, the time to reach steady-state was optimal only at dosages at or above 0.25 mg/kg and was proportionally longer with lower dosages (0.0315–0.5 mg/kg).
DISCUSSION The principal limitations of the study were: first the recovery of radioactivity in excreta was variable (range 63–142%), and second blood samples were not taken with sufficient frequency at early time points to permit detailed analysis of the kinetics of 2003 Blackwell Publishing Ltd, J. vet. Pharmacol. Therap. 26, 213–224
benazepril (the pro-drug). Nevertheless, the results were sufficiently reliable to enable us to make several important observations. First, most radioactivity (approximately 85%) was recovered in the faeces after i.v. benazeprilat indicating that the elimination of this molecule is predominately via the bile in cats. Secondly, we were able to model the disposition and effect of benazeprilat using a physiological-based PK/PD model. This model takes into account the (nonlinear) binding of benazeprilat to ACE in both plasma and tissue compartments, and therefore may permit greater understanding of the pharmacology of benazeprilat in cats than would be possible with conventional compartmental models. In addition, the model can be used to predict kinetic and dynamic profiles with different dosage regimens of benazepril. A similar approach has already been described in dogs with benazeprilat (Toutain et al., 2000a,b). Nevertheless, it must be acknowledged that certain assumptions had to made in the model, and as these cannot be tested at present, the structural validity of the model is currently unknown. The most important assumption may have been that plasma and tissue ACE have similar binding characteristics. Circulating (plasma) ACE is the result of shedding of bound (tissular) ACE and because the latter are extracellular ectopeptidases bound to cell membranes rather than being transmembrane receptors, there is no reason to think that plasma and tissue ACE are fundamentally different. Nevertheless, the microclimate of the two ACEs might differ and change their binding characteristics. In fact no evidence appeared during the modelling that there might have been two subclasses of ACE. In addition, during the modelling process, it was impossible to distinguish between benazeprilat free in the plasma and benazeprilat not specifically bound to other plasma proteins since it was not possible to estimate separately K10, Kd and the nonspecific binding factor (NS) linking free and nonspecifically bound drug (Toutain et al., 2000a). Therefore, it was necessary to ignore NS so that free benazeprilat consisted of genuinely free drug and drug bound nonspecifically to other proteins. A similar simplification was made previously for cilazaprilat (Francis et al., 1987). The consequence of this process is that in vivo Kd is actually equal to Kd(1 + NS). Blood sampling was sufficiently frequent to allow us to profile basic pharmacokinetic parameters of the pro-drug, benazepril. Peak benazepril concentrations were reached rapidly (Tmax 0.3 h), and declined relatively quickly (t1/2 0.7 h). Elimination of benazepril may be via several mechanisms including transformation to the active metabolite (benazeprilat), excretion of free drug, or excretion after transformation to other, as yet unidentified, metabolites. In other species, de-esterification (hydrolysis) of benazepril to benazeprilat is the predominant pathway, and in man and dog this occurs exclusively in the liver (Waldmeier & Schmid, 1989; Webb et al., 1990). After oral administration of benazepril, peak benazeprilat concentrations were reached rapidly (Tmax 1.3 h). This value reflects the absorption of benazepril and metabolic transformation to benazeprilat. The pharmacokinetics of ACEIs cannot be interpreted fully using conventional compartmental models; the decline phases observed in the (total) plasma benazeprilat
222 J. N. King et al.
concentration–time profiles are not equivalent to the distribution and elimination phases of a standard multiexponential compartmental model (Lees et al., 1989; MacFadyen et al., 1993; Toutain et al., 2000a). The final or ‘terminal’ (k3) decline phase (t1/2 16–23 h) for benazeprilat in cats is controlled by the dissociation of drug that was bound to ACE, mainly in the tissues (Till et al., 1984). It is not an elimination phase per se, and is rather the t1/2 for interaction with ACE. The elimination rate of (free) benazeprilat is in fact rather rapid in cats; the t1/2K10 was 1.0 h after single i.v. benazeprilat. This value for the true elimination of free benazeprilat is very similar to the second decline phase (t1/2k2 of 1.3 h) observed for total benazeprilat after i.v. administration. This is because the ACE is saturated at this time of high concentrations, and therefore the decay rate of total benazeprilat reflects mainly clearance of the free drug. This finding emphasizes that a visual good fit of data to an equation does not necessarily mean that the corresponding model is correct; choice of model must also take into account the known behaviour of the drug in question. The t1/2 is a hybrid parameter depending on clearance and volume of distribution. The results indicate that the short elimination t1/2K10 of benazeprilat in cats is due to a small volume of distribution rather than rapid clearance. The value for Vc (93 mL/kg) we calculated in cats for benazeprilat is similar in magnitude to the volume of the blood compartment. We estimated the clearance of (free) benazeprilat to be 0.125 L/kg/h, which is in fact similar to the value of 0.11 L/kg/h obtained using the conventional approach (CL ¼ Dose/AUC[0 fi ¥], using total concentrations). This is because, at a dosage of 1 mg/kg benazeprilat, concentrations are high. Therefore binding to ACE is saturated, most of the measured total plasma concentration is due to free drug and thus there is linearity. The clearance of (free) benazeprilat in cats is approximately one half lower (0.125 L/kg/h) than previously described in dogs (0.216 L/kg/h). We estimated the overall extraction rate of benazeprilat in cats to be approximately 1.5% (cardiac output 140 mL/kg/min vs. clearance of 2 mL/kg/min), lower than the estimated value of 3.6% in dogs (cardiac output 100 mL/kg/min vs. clearance of 4 mL/kg/min). An important observation of this study is that the biliary excretion of benazeprilat is more extensive in cats (85%) than in dogs (52%), rats (45%) or man (15%) (Dieterle et al., 1989; Waldmeier & Schmid, 1989). In cats, the plasma clearance of the free fraction of benazeprilat should be mainly of hepatic origin, and therefore can be approximated by Clh ¼ fu · Clint, where Clh is the hepatic clearance, fu is the benazeprilat free fraction (free from any binding) and Clint is the benazeprilat intrinsic clearance. It is most probable that Clint is the most important factor differentiating cats from other species, because both the binding of benazeprilat to ACE and the extent of binding of benazeprilat to plasma proteins are similar in cats and dogs (data from this study and Toutain et al., 2000a). The results show that the bioavailability of orally administered benazepril is incomplete in cats, although we could not quantify this reliably. Use of the conventional technique of comparing AUC values for total drug (i.e. total benazeprilat after oral benazepril vs. i.v. benazeprilat) produced an F value of 5.3%.
However, this value is biased because calculation of bioavailability by comparison of AUC values with different routes of administration assumes equal rates of clearance (Gibaldi & Perrier, 1982). This assumption is not valid for ACEIs, as their clearance rates are nonlinear with respect to total concentration due to their nonlinear binding to ACE (Toutain et al., 2000a). A more accurate estimate of bioavailability (F ¼ 3.2%) should be obtainable from the model used in this study by calculation of the ratio of clearances of free benazeprilat after oral benazepril vs. i.v. benazeprilat, equivalent to the ratio of AUC values for free benazeprilat. The data on recovery of radioactivity in faeces and urine indicate that the fraction of absorbed benazepril (Fabs) was approximately 23% (Table 1). Therefore, the low bioavailability estimate for benazeprilat (F ¼ 3.2%) in part I of the study may be explained by a combination of incomplete absorption of benazepril from the gastrointestinal tract (23%) and low transformation rate of absorbed benazepril by the liver into benazeprilat (Fm ¼ 13%). This very low estimate of F (3.2%) may be unreliable, however, as the values for clearance (Cl/F) were markedly lower in this part of the study (in only four cats) compared with part II (15–18 cats). Using the clearance values from part II leads to calculation of F values of 10–20% as opposed to 3%. Nevertheless we can conclude that the bioavailability of oral benazepril is incomplete in cats. Low bioavailability is a general feature of pro-drug ACEIs (MacFadyen et al., 1993; Verme-Gibboney, 1997). Repeated oral administration of benazepril leads to moderate increases in benazeprilat concentrations, the calculated bioaccumulation ratio was 1.3. This value would not be predicted from the measured rapid elimination rate of benazeprilat (t1/2K10 1 h with single dosing). The slow terminal decline phase of plasma benazeprilat (t1/2k3 of 16–23 h using conventional line fitting and total concentrations) is due to the dissociation of benzeprilat bound to ACE, and in theory should not contribute to accumulation during repeated administration (Lees et al., 1989; Toutain et al., 2000a). There are two explanations available for the increased exposure to benazeprilat with repeated administration. First, it was noted in a previous study that the bioavailability of benazepril increases in cats with repeated administration (by a factor of 2), presumably due either to increased absorption or reduced first-pass metabolism (King et al., 1999). Secondly, we observed in this study that the clearance of benazeprilat was reduced (by a factor of approximately 2) with repeated administration, evidenced by lower values for Cl/F and higher values for t1/2K10. If clearance is actually decreased during repeated dosing, this is probably due to a time-dependent effect rather than related to nonlinearity (of the disposition of the free fraction) with respect to dose. One use of the model described in this paper is that it allows evaluation of the ACE system in the target animal. We calculated values of 18.2 nmol/kg for the absolute Bmax pool (the absolute molar amount of ACE in the cat’s body assuming 1 mol of benazeprilat bound 1 mol of ACE), 10.6 for the circulating fraction (i.e. the percentage of ACE circulating rather than in the tissues), and 3.53 nmol/L for the affinity constant (Kd). These values are very similar to those described previously for dogs 2003 Blackwell Publishing Ltd, J. vet. Pharmacol. Therap. 26, 213–224
PK/PD modelling of benazepril in cats 223
(Bmax pool 23.5 nmol/kg, circulating fraction 10.6%, and Kd 4.5 nmol/L, Toutain et al., 2000a) suggesting that the ACE system is similar in the two species. Therefore in both cats and dogs, most ACE is tissular (approximately 90%) rather than circulating (approximately 10%). The pharmacodynamics of benazepril were assessed by measurement of plasma ACE activity in part II of the study. After a single benazepril dose, the IC50 for benazeprilat was 0.2–0.3 ng/mL (as free drug) i.e. lower than the corresponding Kd (1–3 ng/mL). Values of the same order (IC50 0.1 ng/mL, Kd 1–2 ng/mL) were reported previously in dogs (Toutain et al., 2000a). The differences between IC50 and Kd values may be due, among others, to the (in vitro) assay conditions for ACE (Toutain et al., 2000a). Interestingly, the IC50 values in cats were higher after repeated administration (0.7–1 ng/mL) compared with single application (0.2–0.3 ng/mL) and the Bmax also changed (decreased) with repeated administration. These results suggest that the ACE system changes in cats with sustained ACE inhibition. One practical consequence of these findings is that single dose studies may not be predictive for steady-state. Expressed in terms of total benazeprilat plasma concentration, the IC50 values for benazeprilat in cats were approximately 1.8 ng/mL (single dose) and 2.1–2.5 ng/mL (repeated dose). These concentrations were observed between 48 and 72 h after single or repeated dosing with 0.25–1 mg/kg benazepril.HCl, consistent with the long-lasting inhibition of ACE observed with benazepril in cats (King et al., 1999). The slope factor (Hill coefficient) for the benazeprilat/ACE interaction was rather high (1.6–1.9 after a single dose, 2.1–3.7 after multiple dosing). This suggests that ACE will remain fully inhibited over a range of plasma concentrations, but below a critical plasma benazeprilat-free concentration, ACE activity will return relatively quickly to baseline values. A similar slope factor (1.5 after a single dose) was also reported for benazeprilat in dogs (Toutain et al., 2000a). In vitro, slope factors greater than 1 may be due to positive interactions between more than one molecule of drug binding to the enzyme, although there is no evidence that more than one molecule of benazeprilat binds to ACE. In dogs, similar values were obtained with the model (1.5) compared with conventional concentration-effect analysis (1.3) for which no model was used (King et al., 1995). In vivo, the Hill coefficient should not be interpreted from a mechanistic point of view, but rather considered only as an empirical shape factor (Toutain, 2002). A final use of the model is prediction of kinetic and dynamic profiles with different dosage regimens of benazepril. These simulations show that although marked inhibition of plasma ACE activity can be achieved at steady-state with low dosages of benazepril.HCl (e.g. 0.0625 mg/kg administered twice daily), the time to achieve steady-state inhibition is faster with higher dosages. In addition, the effect of benazepril at trough effect is determined mainly by the frequency of dosing and not the dose. This is because, above a minimum dosage, benazeprilat concentrations in the terminal phase are largely independent of dose due to the saturable binding of benazeprilat to ACE. From this finding, 2003 Blackwell Publishing Ltd, J. vet. Pharmacol. Therap. 26, 213–224
one can predict that increased duration of action is unlikely to be achievable simply by increasing the dosage. It must be emphasized, however, that as the relation between plasma ACE activity and clinical efficacy of ACEIs is unproven and probably not simple in any species, any modelled dosage regimens need to be confirmed in clinical efficacy and tolerability trials. For example, part of the efficacy of ACEIs may be due to reduced degradation of bradykinin and/or reduced aldosterone production, neither of which are related in a simply way to ACE activity (Brown & Vaughan, 1998). In addition it is not known whether complete blockade of ACE activity throughout the dosing interval is needed, or is even desirable, during long-term therapy. For example, transient recovery of ACE activity before administration of the next dose may be important for safety reasons. In conclusion, the use of a physiological pharmacokinetic/ pharmacodynamic model permitted us to study several aspects of the pharmacology of benazepril and benazeprilat in cats. The main differences between cats and dogs for this drug are: first, the clearance of (free) benazeprilat is lower (0.125 L/kg/h) in cats compared with dogs (0.215 L/kg/h), and secondly, the biliary excretion of benazeprilat is more extensive in cats (85%) than in dogs (50%).
ACKNOWLEDGMENTS We thank Dr M. Jung for performing the protein binding study, Drs R. Hotz and E. Humbert-Droz for assistance with the animal work and Dr G. Strehlau for conducting the statistical analyses.
REFERENCES Balfour, J.A. & Goa, K.L. (1991) Benazepril. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in hypertension and congestive heart failure. Drugs, 42, 511–539. Brown, N.J. & Vaughan, D.E. (1998) Angiotensin-converting enzyme inhibitors. Circulation, 97, 1411–1420. Dieterle, W., Ackermann, R. & Kaiser, G. (1989) Pharmacokinetics of benazeprilat after intravenous administration in healthy volunteers. European Journal of Clinical Pharmacology [Suppl], 36, 303. Francis, R.J., Brown, A.N., Kler, L., Fasanella d‘Amore, T., Nussberger, J., Waeber, B. & Brunner, H.R. (1987) Pharmacokinetics of the converting enzyme inhibitor cilazaprilat in normal volunteers and relationship to enzyme inhibition: Development of a mathematical model. Journal of Cardiovascular Pharmacology, 9, 32–38. Gibaldi, M. & Perrier, D. (1982) Pharmacokinetics. Marcel Dekker, New York. Kaiser, G., Ackermann, R., Dieterle, W. & Dubois, J.P. (1987) Determination of a new angiotensin converting enzyme inhibitor and its active metabolite in plasma and urine by gas chromatography-mass spectrometry. Journal of Chromatography, 419, 123–133. King, J.N., Mauron, C. & Kaiser, G. (1995) Pharmacokinetics of the active metabolite of benazepril, benazeprilat, and inhibition of plasma angiotensin-converting enzyme activity after single and repeated administrations to dogs. American Journal of Veterinary Research, 56, 1620–1628. King, J.N., Maurer, M., Morrison, C., Mauron, C. & Kaiser, G. (1997) Pharmacokinetics of the angiotensin-converting-enzyme inhibitor,
224 J. N. King et al. benazepril, and its active metabolite, benazeprilat, in dog. Xenobiotica, 27, 819–829. King, J.N., Humbert-Droz, E. & Maurer, M. (1999) Plasma angiotensin converting enzyme activity and pharmacokinetics of benazepril and benazeprilat in cats after single and repeated oral administration of benazepril.HCl. Journal of Veterinary Pharmacology & Therapeutics, 22, 360–367. Koeppe, P. & Hammann, C. (1980) A program for non-linear regression analysis to be used on desk-top computers. Computer Methods and Programs in Biomedicine, 12, 121–128. Lees, K.R., Kelman, A.W., Reid, J.L. & Whitting, B. (1989) The pharmacokinetics of an ACE inhibitor, S-9780, in man: Evidence of tissue binding. Journal of Pharmacokinetics & Biopharmaceutics, 17, 529–550. MacFadyen, R.J., Meredith, P.A. & Elliott, H.L. (1993) Enalapril clinical pharmacokinetics and pharmacokinetic-pharmacodynamic relationships. An overview. Clinical Pharmacokinetics, 25, 274–282. Picard-Hagen, N., Gayrard, V., Alvinerie, M., Smeyers, H., Ricou, R., Bousquet-Melou, A. & Toutain, P.L. (2001) A nonlabeled method to evaluate cortisol production rate by modeling plasma CBG-free cortisol disposition. American Journal of Physiology Endocrinology and Metabolism, 281, E946–E956. Till, A.E., Gomez, H.J., Hichens, M., Bolognese, J.A., McNabb, W.R., Brooks, B.A., Noormohamed, F., & Lant, A.F. (1984) Pharmacokinetics
of repeated single oral doses of enalapril maleate (MK-421) in normal volunteers. Biopharmaceuticals and Drug Disposition, 5, 273–280. Toutain, P.L., Lefebvre, H. & King, J.N. (2000a) Benazeprilat disposition and effect in dogs revisited with a pharmacokinetic/pharmacodynamic modeling approach. Journal of Pharmacology and Experimental Therapeutics, 292, 1087–1093. Toutain, P.L., Lefebvre, H. & Laroute, V. (2000b) New insights on effect of kidney insufficiency on disposition of angiotensin-converting enzyme inhibitors: case of enalapril and benazepril in dogs. Journal of Pharmacology and Experimental Therapeutics, 292, 1094–1103. Toutain, P.L. (2002) Pharmacokinetic/pharmacodynamic integration in drug development and dosage regimen optimization for veterinary medicine. AAPS Pharm Sci, 4, article 38 (http://www.aapspharmsci.org/scientificjournals/pharmsci/journal/ps040438.htm). Verme-Gibboney, C. (1997) Oral angiotensin-converting-enzyme inhibitors. American Journal of Health System Pharmacy, 54, 2689– 2703. Waldmeier, F. & Schmid, K. (1989) Disposition of [14C] - benazepril hydrochloride in rat, dog and baboon. Drug Research, 39, 62–67. Webb, R.L., Miller, D., Traina, V. & Gomez, H.J. (1990) Benazepril. Cardiovascular Drug Reviews, 8, 89–104.
2003 Blackwell Publishing Ltd, J. vet. Pharmacol. Therap. 26, 213–224
