Introduction

Top Abstract Introduction Results Discussion Appendix References

Pimobendan [4,5-dihydro-6-2-(4-methoxyphenyl)-1H-benzimidazole-5-yl-5-methyl-3(2H)-pyridazinone] (pimo)¹ is a new benzimidazole-pyridazinone derivative that exhibits both positive inotropic and vasodilator properties (Ruegg et al., 1984; Von Meel, 1985; Verdouw et al., 1986) by specific type III phosphodiesterase inhibition (Scholz and Meyer, 1986; Schmitz et al., 1989) and by directly increasing the calcium sensitivity of the cardiac myofilaments (Ruegg, 1986; Fujino et al., 1988). Oral and i.v. pimo has been shown to increase cardiac output (CO), heart rate (HR), ejection fraction (EF), left ventricular maximum rate of change in pressure over time (dP/dt), and decrease systemic vascular resistance, left ventricle filling pressure, and pulmonary capillary wedge pressure in a dose-dependent manner in animals and patients with chronic congestive heart failure (Walter et al., 1988; Renard et al., 1988; Hagemeijer et al., 1989a,b; Baumann et al., 1989; Remme et al., 1989). However, no hemodynamic data are available from healthy volunteers.

We have reported the pharmacokinetics of enantiomers of pimo in healthy humans (Chu et al., 1995a). The present report focuses on the pharmacodynamic (PD) effect of pimo. In addition, a simultaneous pharmacokinetic (PK)-PD modeling was established to describe and predict the hemodynamic effect of pimo after oral and i.v. administration.

Materials and Methods

Procedures. The study protocol was approved by the Human Subjects Institutional Review Board of the National Defense Medical Center and the Department of Health, Executive Yuan, Republic of China. The criteria for the selection of the eight volunteers, the laboratory tests, and the general procedure have been described previously. Briefly, three 2.5-mg (7.5 mg) capsules of pimo (Boehringer Ingelheim GmbH, Ingelheim, Germany) were administered in the oral study. Two weeks after the oral study, 5 mg of pimo was given i.v. Blood samples were analyzed for enantiomers of pimo by a coupled achiral-chiral normal-phase high-performance liquid chromatography (Chu et al., 1992). Data were fitted iteratively reweighted as 1/(predicted value) (Sheiner and Beal, 1985; Sheiner, 1985) to two- or three-compartment models by using a nonlinear regression program PCNONLIN (Weiner, 1986). A lag time was included in each model in the oral study. In all subjects the 2-compartment model gave the better fit to the data as assessed by the Akaike information criterion (Yamaoka et al., 1978). Blood pressure (BP) and HR were checked and the hemodynamic effects of pimo were evaluated by echocardiography at 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, and 10 h after drug administration. A Hewlett-Packard Ultrasound Imaging System-1500 (Hewlett-Packard Co., Palo Alto, CA) was used for echocardiographic recording.

Echocardiography Recording and Analysis. Cardiac function indices including cardiac index (CI), EF, mean velocity of circumferential fiber shortening (MVcf), systolic time intervals, and peak flow velocities were evaluated by a Hewlett-Packard SONOS ultrasound system with phased-array Doppler (model 1500). M-mode, two-dimensional echocardiographic images, and pulsed-wave Doppler velocity signals were recorded on a Panasonic standard VHS video cassette recorder (model AG7300) for a later playback and analysis. The subjects were examined in a slight left-lateral decubitus position during quiet respiration. Left parasternal long-axis, left parasternal short-axis, apical 4-chamber, and apical 5-chamber views were recorded. Mitral valve-flow peak velocity (MVPV) and aortic flow peak velocity (AOPV) were recorded using a pulsed-wave Doppler. Transmitral flow velocity recordings were obtained from an apical 4-chamber or apical 2-chamber view using a 2.5 MHz imaging transducer. The sample-volume cylinder, approximately 1 cm in axial length, was superimposed on the two-dimensional image at the level of the mitral valve orifice and oriented parallel to the apparent long-axis of blood flow. The transducer was then angled to record maximal transmitral flow velocities. Doppler tracings of aortic flow were obtained using the apical 5-chamber view, with the sample-volume placed just beyond the valve leaflets within the proximal aortic root. Slight adjustments in transducer angulation or sample-volume position were at times required to maximize the audio and graphic quality of the Doppler signal. The aortic valve motion on parasternal long-axis view and velocities of mitral and aortic flow were recorded over several cardiac cycles at a monitor sweep speed of 100 mm/s. M-mode echocardiographic measurements followed the standard method as recommended by the American Society of Echocardiography (Sahn et al., 1978). Left ventricle end-diastolic dimension (LVEDD) and left ventricle end-systolic dimension (LVESD) were measured in parasternal long-axis view at the level between the papillary muscle and mitral leaflet tips. Left atrium (LA) was measured at the aortic valve level. Left ventricle volume (V) was determined by the dimension (D) using Teichholz's equation (Teichholz et al., 1976) V = [7.0/(2.4 + D)] × D3. Stroke volume (SV) equals LVEDV  LVESV; EF equals SV/LVEDV and CI equals SV × HR/body surface area. Pre-ejection period (PEP) was the time duration measured from the onset of ventricular depolarization (onset of QRS complex on simultaneous electrocardiogram) to the opening of the aortic valve. Left ventricular ejection time (LVET) was measured from the opening to the closure of the aortic valve. All measurements were made with the aid of an off-line computerized analysis station equipped with internal calipers and a programmable graphic analyzer (Dextra 300, Dextra Medical Inc., Long Beach, CA). The recorded images were played back through a videocassette system equipped with a frame-by-frame bidirectional search module (Panasonic model AG7300).

Reproducibility and Validation. The variabilities of echocardiographic indices were evaluated in eight other normal subjects before conducting this study. Four levels of coefficient of variation (CV) were evaluated. Level-1 CV was determined by repeatedly measuring the cardiac function indices in one cardiac cycle 5 times to evaluate the error resulting from using the pointing device of a computer digitizer. Level-2 CV was evaluated by measuring the indices 5 times in five consecutive cardiac cycles to detect the error caused by sonographer, interpreter, and examined subject in addition to level-1 error. Level-3 CV was evaluated by examining each subject at four time points (9:00 AM, 11:00 AM, 2:00 PM, and 4:00 PM) in a day to identify the time-period variation in addition to the error of the previous two levels. Level-4 CV detected the interobserver variation by measurement of all indices by four experienced echocardiographers. The percentage of coefficient of variation (CV%) of levels-1 and -2 were less than 5% except in PEP (8.0% in level-2). Level-3 CV% was less than 10% except in MVPV (11.1%) and CO (12.0%). The interobserver CV% was less than 10% except in measuring LV posterior wall dimension (10.6%).

Simultaneous PK-PD Modeling. Plasma and red blood cell concentrations of (+)- and ()-pimo and changes in LVESD were used to characterize the simultaneous PK-PD model. Plots of (+)-pimo concentration versus percent changes in LVESD in time sequence showed counterclockwise hysteresis loops. Therefore, two different levels of effect may be seen at a single plasma concentration (C_p) and two different C_ps of pimo may produce the same effect, which seems against PD principles. However, this might happen in conditions in which the site of action of pimo is kinetically distinguishable from the plasma compartment. To collapse the hysteresis loop and determine the relationship between C_p and the effect, one may: 1) sample effect-site concentration (C_e), 2) perform multiple steady-state experiments, or 3) model the effect site as a kinetic "compartment" linked to the PK compartments. We postulated a hypothetical effect compartment linked to the plasma compartment by a first order process (link PK model; Fig. 1). A two-compartment model of pimo disposition was used when modeling the PK data. It was assumed that drug enters and leaves the effect compartment by a first order process with no appreciable reflux of drug back into the PK system. In this link PK model, the concentration and time course of drug at the effect compartment are determined by the elimination rate constant of the effect-compartment (k_e0). The greater the k_e0, the less is the time lag between the central and effect compartments. To determine an appropriate k_e0, a parametric approach with sigmoid maximal effect (E_max) PD model was included in the simultaneous PK-PD model.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   This diagram represents the link PK model defined as the link of the hypothetical compartment to the PK model. By assuming that the transfer of the mass of drug from the central compartment to the effect compartment is relatively small, the drug moving back to the central compartment is negligible and will not add another compartment to the original two-compartment PK model. GI: gastrointestinal compartment; C: central compartment; P: peripheral compartment; E: effect compartment.

Statistical Analysis. The changes in the PD data according to time were evaluated by repeated measured analysis of variances followed by Dunnett's multiple comparisons. A paired t test was used to compare the parameters of the simultaneous PK-PD model in oral and i.v. studies. Differences in the p values of < .05 (two-tailed) were considered significant. Results were reported as mean ± S.E..

	Results

Top Abstract Introduction Results Discussion Appendix References

Pharmacodynamics. After oral administration of 7.5 mg of pimo, the dimension of the heart decreased (Fig. 2 and Table 1). The LVESD and LVEDD significantly reduced 29.0% and 16.5% of their baseline values at 2 to 5 h, respectively. This effect lasted for at least 10 h. In a similar trend, the LA dimension, LVET, and PEP were significantly reduced 20 to 30%. The parameters of LV contractile force including MVcf, EF, and fractional shorting were significantly increased to 105.9%, 29.8%, and 46%, respectively. BP showed no significant change, whereas HR increased 38%. The AOPV, which was determined by the SV, aortic valve area, afterload, and contractility, was also increased significantly to 29.3% in 2 to 4 h. The stroke volume index (SVI) reduced insignificantly in 2 to 4 h and returned toward baseline after 5 h. The CI, calculated by SVI × HR, increased gradually but not significantly until 5 h. This increase in CI resulted from an increase in HR. The hemodynamic change after i.v. administration of 5 mg of pimo showed a similar change in the parameter but to a lesser extent (Fig. 3). Because the bioavailability of pimo was about 50% (Chu et al., 1995a), a greater effect was seen with a smaller available dose of pimo (3.75 versus 5 mg), which suggested that the demethylated metabolite formed in first pass metabolism was contributing to the overall hemodynamic effect.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   This plot represents the cardiovascular effects of pimo in eight normal healthy volunteers after oral administration of 7.5 mg of racemic pimo. SBP and DBP: systolic and diastolic blood pressure. Decrease in cardiac chamber dimensions and increase in ejection-phase indices, HR, and CI developed gradually toward their maxima in 2 to 5 h and lasted for 8 to 10 h. Data were expressed in mean ± S.E.. *represents p < .05 by using repeated measured ANOVAs followed by Dunnett's multiple comparisons.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   This plot demonstrates the cardiovascular effects of pimo in eight normal healthy volunteers after i.v. administration of 5 mg of racemic pimo. Legends are the same as in Fig. 1. The changes are similar to those in the oral study but to a lesser extent. Diastolic BP decreased significantly in 2 to 6 h. CI did not change significantly.

Simultaneous PK-PD Model. There was a delay between the peaking time of plasma pimo concentration and the E_max (changes in LVESD; Fig. 4). When the effects were plotted against C_ps in time sequence, large counterclockwise hysteresis loops were observed for oral and i.v. administration, respectively (Fig. 5A and B). After applying the simultaneous PK-PD model, the hypothetical C_es were obtained along with the estimated effect (E_est; Fig. 4). It is worth noting that observed effect (E_obs) is unable to fit or estimate unless the effect site concentration of drug is known. This figure demonstrates a good fitting between E_obs and E_est. When effects were plotted against the effect-compartment concentrations, the hysteresis loop collapsed (Fig. 5, C and D). This demonstrated that concentration-effect relationship or pharmacodynamics of pimo could be found by using the simultaneous PK-PD model.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Observed Cp of (+)-pimo and Eobs and Eest versus time curves after oral administration of 7.5 mg of racemic pimo in eight normal healthy volunteers. A simultaneous PK and PD model was established using observed Cps and Eobs. There is a good correlation between Eobs and Eest. The dotted line represents hypothetical effect compartment concentration of (+)-pimo. Data are presented as mean ± S.E.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Eobs and Eest versus Cps of (+)-pimo in plasma plotted in time sequence in eight normal healthy volunteers after oral (A) and i.v. (B) administrations of pimo showed counterclockwise hysteresis loops. The curves of Eest showed good fitting with the observed data. It was unable to fit the data without knowing the Ce, which was obtained by applying the simultaneous PK-PD model. When effect was plotted against the hypothetical effect compartment concentration (C, D), the hysteresis loop collapsed and the concentration-effect relationship could be found.

1

	Discussion

Top Abstract Introduction Results Discussion Appendix References

Pharmacodynamics. We have reported the hemodynamic effects of pimo in patients with congestive heart failure (Chu et al., 1995b). In patients, the HR, SV, and CO were increased and the BP and the filling pressure of the cardiac chambers were decreased. In our study of normal volunteers, the decrease in chamber size and increase in HR and CO were also significant, but BP and SV were decreased insignificantly. These different effects on SV in patients and in normal subjects might be due to different preload and inotropic response.

Simultaneous PK-PD Model. One of the major goals in clinical pharmacology is to have a drug given in a proper way, proper dose, and suitable duration to achieve and maintain an ideal therapeutic effect in patients of variable disease entities. To produce its characteristic effects, a drug must be present in appropriate concentrations at its sites of action. In most cases, the concentration of drug in the systemic circulation will be parallel to that at the sites of action. There is then a direct link between PK and PD models. For some drugs, there is no clear or simple relationship between pharmacological effect and concentration in plasma. Many factors, such as formation of active metabolites, enzyme induction, enzyme inhibition, receptor desensitization, indirect or irreversible pharmacological effect, delayed equilibrium in concentration between plasma and sites of action, and enantiomers with different elimination rates and potency (Holford and Sheiner, 1981; Mehvar, 1992) may account for the lack of correlation. Although the growth of pharmacodynamics is less than that in pharmacokinetics because difficulties exist in the quantitative determination of meaningful responses, it is possible to use mathematical models to describe the effect-concentration relationship. The simultaneous PK-PD model is first proposed by Segre (1968) and later by Galeazzi et al. (1976), Hull et al. (1978), and subsequently elaborated by Sheiner et al. (1979). These modeling techniques have been applied to various drugs including neuromuscular blocking agents (Evans et al., 1984), anesthetics (Stanski et al., 1984), bronchodilators (Oosterhuis et al., 1984), antihypertensives (Reid and Meredith, 1990), benzodiazepines (Buhrer et al., 1990), antibiotics (Garrison et al., 1990), steroids (Kong et al., 1989), and diuretics (Hammarlund et al., 1985). We herein applied this integrated model to a cardiotonic agent.