Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

In the development of new short-acting neuromuscular blocking agents (NMBAs), it is essential to know the factors governing the time course of action of this class of drugs (Wierda et al., 1993). It has become evident that the time course of action (onset, duration, recovery rate, accumulation) is determined, at least partly, by the time course of the plasma drug concentration (Donati, 1988; Wierda et al., 1993). Therefore, pharmacokinetic (PK) analysis is an essential step in the development of NMBAs with a desired time profile, and may help to understand the underlying mechanisms. It is well recognized that physicochemical properties play an important role in the PK as well as the pharmacodynamic (PD) behavior of drugs, which in turn are related to the chemical structure (Rekker, 1977; Seydel and Schaper, 1982; Neef and Meijer, 1984; Noy and Zakim, 1993). Therefore, chemical structure, physicochemical properties, and PK and PD should be studied simultaneously to give a rational basis for the developmental research to new drugs (Seydel and Schaper, 1982; Peck et al., 1992). NMBAs represent a unique group of model compounds for PK/PD studies because their PD effect can be monitored accurately with well defined baseline and maximal effects.

The liver plays an important role in the distribution and elimination of more bulky cationic drugs, such as aminosteroidal neuromuscular blocking agents, and as a result, in the time course of action of NMBAs, as has been demonstrated in humans for vecuronium (Bencini et al., 1986). These bulky cations are supposed to be taken up via a carrier-mediated mechanism (Meijer et al., 1997; Koepsell, 1998; Smit et al., 1998; Zhang et al., 1998). Because the distribution, elimination, and the effect of drugs are related to the unbound concentration, protein binding may affect the potency and time course of action of drugs, as well as their hepatobiliary transport rates.

To obtain more insight in the relationship between chemical structure, physicochemical properties of NMBAs, and their PK and PD characteristics, a series of 12 aminosteroidal NMBAs, supplemented with data on five related NMBAs from the literature (pancuronium, dacuronium, Org 6368, pipecuronium, and vecuronium), was investigated in a screening program. In a previous article, the relationships between chemical structure and physicochemical properties (lipophilicity and plasma protein binding) and their behavior in the isolated perfused rat liver have been described (Proost et al., 1997). The PK (plasma concentration-time profile and excretion into urine and bile of the parent compounds and their potential metabolites) and PD (neuromuscular blocking effect on tibialis anterior muscle) of these compounds were studied in vivo in anesthetized cats after administration of an i.v. bolus administration of two times the 90% blocking dose (ED₉₀). The present article describes the PK and the relationship between their chemical structure and physicochemical and PK properties. The PD and the PK/PD relationship will be described in a separate article.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. The NMBAs (Fig. 1) and their putative metabolites, i.e., their 3-OH-, 17-OH-, and 3,17-di-OH-derivatives, were supplied by Organon Labs. Ltd., Newhouse, Scotland. All other chemicals were of analytical grade and were obtained from commercial sources.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Chemical structures, including steroidal skeleton and list of compounds.

Experimental Procedure. With approval of the committee for care and use of laboratory animals of the University of Groningen, adult male cats, weighing 3 to 6 kg, were anaesthetized with pentobarbital sodium 40 mg · kg¹ i.p. Anesthesia was maintained by pentobarbital sodium (4.8 mg · kg¹ · h¹) i.v. via a catheter in the left vena cephalica. After intubation ventilation was maintained with room air. Blood pressure was measured via a catheter in the arteria femoralis. Blood pressure and heart rate were monitored and the body temperature was measured rectally and maintained at 37°C.

Determination of Compounds and Their Hydroxy-Derivatives in Plasma, Urine, Bile, and Liver Homogenate. The determination of the NMBAs and their putative metabolites (3-OH, 17-OH, and 3,17-di-OH analogs) in plasma, urine, bile, and liver homogenate was carried out by HPLC with postcolumn ion-pair extraction and fluorimetric detection, as described elsewhere (Kleef et al., 1993; Proost et al., 1997), with the following modifications.

1

PK Analysis. Plasma concentration data were analyzed by iterative nonlinear regression with the program MultiFit (developed in our department). For each individual animal, the parameters of a two- and three-exponential equation were fitted to the logarithm of the plasma concentration-time data pairs, assuming a constant relative error. The correctness of this assumption was tested by visual inspection of the graphs of the residuals plotted against time and against the concentration. Moreover, it is known that the relative error of the bioanalysis was almost independent of the concentration over the entire concentration range (Kleef et al., 1993).

10

5

Correlation Analysis. The correlation between various parameters was determined by standard linear regression analysis. Correlations were considered significant if P < .05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The potency of the compounds was estimated in pilot experiments in cats by assessment of the dose needed to reach a 90% neuromuscular block (ED₉₀). Two times the estimated ED₉₀ was used in the experiments (Table 1).

Plasma Concentration Profile. After i.v. injection of each of the compounds, the plasma concentration decreased very rapidly, showing a biphasic pattern; an example is shown in Fig. 2). In some cases, the individual plasma concentration data could not be fitted satisfactorily due to the small number of data points above the limit of quantification. Therefore, the PK parameters derived from the pooled data for each compound, are listed in Table 2. The mean values of the PK parameters obtained from the data of each individual experiment were close to the values in Table 2; for no parameters did the difference exceed 10%. For Org 9273 the biexponential equation could not be fitted to the data; therefore, the one-compartment model was used for this compound. For the pooled data of Org 9489 and Org 9487, the three-compartment model fitted significantly better to the data than the two-compartment model; however, the goodness-of-fit for the two-compartment model was better than that for the other compounds. Therefore, the PK parameters for the two-compartment model (except for Org 9273) have been given in Table 2, allowing a fair comparison with the other compounds. The goodness-of-fit was satisfactory in all cases, with a mean residual coefficient of variation of 12%, ranging from 4 to 34%.

View larger version (9K): [in this window] [in a new window]   Fig. 2.   Plasma concentration of Org 9487 in the cat (data from one experiment). The line represents the calculated plasma concentration profile obtained by PK analysis.

1

1

1

1

1

1

1

Urinary Excretion. The urinary excretion was invariably found to be low; the total of the parent drug and metabolites excreted via the urine was <10% for each of the compounds (Table 3). Only small amounts (<0.5%) of the 3-OH metabolites were found, with exception of Org 9616 where 2% of the 3-OH metabolite were excreted into urine. Other metabolites were not found, or only in trace amounts (<0.1%).

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Urinary excretion rate of Org 9487 () and its 3-OH metabolite () in the cat (data from the same experiment as shown in Fig. 2). The lines represent the calculated urinary excretion rate profiles obtained by PK analysis.

Hepatic Uptake and Biliary Excretion. The parent drugs were excreted in large amounts into bile, besides smaller amounts of the 3-OH metabolites in the case of parent compounds containing a 3-acetyl group. Only trace amounts of the 17-OH metabolites were found (Table 3). Significant amounts of the parent drug, and of the 3-OH metabolite of 3-acetyl compounds, were found in the liver after termination of the experiment (Table 3).

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Biliary excretion rate of Org 9487 () and its 3-OH metabolite () in the cat (data from the same experiment as shown in Fig. 2). The lines represent the calculated biliary excretion rate profiles obtained by PK analysis.

Correlation between Lipophilicity and PK Parameters. In Table 5, the relationship between lipophilicity and the PK parameters has been summarized. Lipophilicity was expressed as the logarithm of the partition coefficient octanol/Krebs (Proost et al., 1997). The data of pipecuronium were not included because the numerical value of P could not be determined reliably; the estimated value was <.02 (Proost et al., 1997). The data for Org 9273 were not included because the two-compartment model could not be applied to these data. It was confirmed that exclusion of these data did not affect the general conclusions.

View this table: [in this window] [in a new window]   TABLE 5 Correlation between lipophilicity and PK parameters Correlation between lipophilicity [expressed as logarithm of partition coefficient (Proost et al., 1997)] and PK parameters, calculated by linear correlation analysis. Data excluding pipecuronium (no value for partition coefficient available) and Org 9273 (no data for two-compartment model available).

View larger version (9K): [in this window] [in a new window]   Fig. 5.   Relationship between lipophilicity (log P) and unbound clearance in the cat.

View larger version (9K): [in this window] [in a new window]   Fig. 6.   Relationship between lipophilicity (log P) and steady-state volume of distribution in the cat.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Each of the investigated compounds disappeared very rapidly from plasma; both the initial and the second half-life were significantly shorter than for the older compounds pancuronium, pipecuronium, and vecuronium as a result of a higher plasma clearance and a smaller volume of distribution (Table 2).

The PK of three of the investigated compounds, rocuronium (Org 9426), Org 9616, and Org 7268 (3-desacetylvecuronium), have been studied in cats previously (Khuenl-Brady et al., 1990; Segredo et al., 1991). Khuenl-Brady et al. (1990) reported a longer initial half-life and a higher central volume for rocuronium and Org 9616; these differences are likely to be related to the less frequent blood sampling shortly after injection because we obtained markedly higher values with only the data at sample times used in that study. Values of terminal half-life and steady-state volume of distribution reported by Khuenl-Brady et al. (1990) were larger than in our study; these differences may related to the higher dose used in that study (four to seven times the doses used in our study), allowing plasma concentration measurement over a longer period of time. The plasma clearance of rocuronium and Org 9616 was similar in both studies. In contrast, Segredo et al. (1991) reported for Org 7268 a mean plasma clearance of 14 ml · min¹ · kg¹; we found a value of 58 ml · min¹ · kg¹. The values for renal excretion of rocuronium and Org 7268 reported by Khuenl-Brady et al. (1990) and Segredo et al. (1991) were markedly higher than in our study. The reasons for these differences are not well understood; it cannot be excluded that nonlinear PK may become apparent at the higher dose used in these studies.

We found a significant negative correlation between lipophilicity and both half-lives, V_ss, and MRT (Table 5). Previously, it was found that protein binding of NMBAs and other organic cations was strongly correlated to lipophilicity (Van der Sluijs and Meijer, 1985; Proost et al., 1997). The correlation between lipophilicity and V_ss (Fig. 6) may be at least partly explained by the influence of protein binding because an increase of protein binding results in a decrease of the volume of distribution. As a result, the half-lives and MRT will decrease.

In the isolated, perfused rat liver, we found previously a significant correlation between lipophilicity and hepatic uptake clearance (Proost et al., 1997). Because the liver is the major organ of elimination for most of the investigated compounds, such a relationship was expected in the present study. However, no obvious correlation was found between the total body clearance and the lipophilicity (Table 5). Yet, we found a significant correlation of lipophilicity with unbound clearance and unbound initial plasma clearance (Table 5; Fig. 5). It can be inferred that the increase of unbound clearance with increasing lipophilicity is counteracted by the concurrent increase in protein binding, resulting in a much less pronounced change of plasma clearance. Similar patterns can be seen for the initial plasma clearance and initial unbound clearance. This finding demonstrates that studying the primary parameters (e.g., protein binding and clearance of the unbound fraction) may reveal details that are not noticed if only secondary parameters (e.g., total body clearance) are compared. Also, it emphasizes the importance of protein binding for the PK and time course of action of drugs (Proost et al., 1996).

Lipophilicity cannot be responsible for the difference in PK between vecuronium and rocuronium because the lipophilicity of both compounds is similar (Proost et al., 1997). In the isolated, perfused rat liver, it was found that rocuronium is extracted more efficiently by the liver than vecuronium (Proost et al., 1997), and this also was observed in isolated human hepatocytes (Sandker et al., 1994). Therefore, it is likely that the higher unbound clearance for rocuronium compared with vecuronium in the present study (41 versus 27 ml · min¹ · kg¹) is due to a more efficient liver uptake.

The half-life of the compounds in urine was found to be somewhat longer than that observed in plasma. Obviously, the true half-life in plasma is longer than the observed values and could not be determined adequately because the plasma concentration rapidly dropped below the limit of quantification. The half-life of the biliary excretion was even longer, in the range of 120 to 480 min, and it is not unlikely that these values are more close to the true terminal half-life in plasma. However, this very long half-life is reached at plasma concentration far below the effective concentrations, and even far below the limit of quantification. This demonstrates the limited significance of the concept of half-life (Hughes et al., 1992). Moreover, the clinical significance of half-lives in terms of duration of drug action is complicated and can be misleading (Shafer and Stanski, 1992). If the true half-life is much longer than the values listed in Table 2, clearance will be overestimated, and steady-state volume of distribution will be underestimated significantly. In fact, the measurable range of the plasma concentration reflects distribution and elimination only partly and cannot discriminate between elimination and distribution into deep compartments.

The PK analysis of the plasma concentration data was performed under the usual assumption that elimination takes place from the central compartment. However, considering the results obtained in the isolated perfused rat liver (Proost et al., 1997), showing that equilibration between liver and plasma is not extremely rapid, the assumption that hepatic elimination, i.e., both metabolism and biliary excretion, takes place from the central compartment seems questionable. If metabolism and biliary excretion are assumed to take place from the peripheral compartment, V_ss and MRT will be significantly larger than the values listed in Table 2; the values of the half-lives, CL_init, CL, and V₁ are independent of the route of elimination.

For each of the compounds, the renal clearance was low (Table 4), without a correlation to lipophilicity (Table 5). The unbound renal clearance ranged from 0.8 (rocuronium) to 9.7 ml · min¹ · kg¹ (Org 9991). Assuming a glomerular filtration rate of ~2.5 ml · min¹ · kg¹ in the cat (Russo et al., 1986), this would imply that several of the compounds (vecuronium, Org 9489, Org 9487, Org 9453, Org 9991, Org 20297, Org 9955) are actively secreted in the kidney. However, a hypothesis of such active secretory process is not supported by other studies. More likely, the relatively high unbound renal clearance support the aforementioned hypothesis that the observed values for the plasma clearance are overestimated.

Each of the compounds containing an acetyl-ester group at position 3 was partly metabolized to the 3-OH derivative (Table 3). In general, the fraction of the dose converted to the 3-OH derivative was markedly smaller than in the isolated, perfused rat liver (Proost et al., 1997). Also, hydrolysis of the 17-OH ester group hardly occurred in the cat, in contrast to the rat. It should be noted that the formation of 3-OH derivatives is not necessarily a purely enzymatic process. It is known that hydrolysis of the 3-OH ester bond occurs rapidly in aqueous solutions at 37°C and pH 7.4; the 17-O-esters are more stable against hydrolysis (J.H.P. and J.R., unpublished data).

Significant amounts of the parent drug and the 3-OH metabolite were found in the liver after termination of the experiment (Table 3), and biliary excretion was still measurable 480 min after administration. However, in all experiments, the estimated amount to be excreted after the sampling time extrapolated to infinity was markedly lower than the measured amount in the liver 480 min after administration, ranging from 4 to 49%. This finding suggests that a considerable amount of the drug in the liver is not directly available for biliary excretion, and that a certain fraction is excreted even more slowly. Persistent storage in lysosomes and mitochondria may lead to a deep PK compartment in the liver (Mol and Meijer, 1990; Meijer et al., 1997).

The recovery of the administered doses in urine, bile, and liver homogenate was found to be incomplete, ranging from 39 to 93%. In principle, metabolic pathways other than hydrolysis of the esters at positions 3 and 17 cannot be excluded, and these metabolites may be undetectable in our HPLC analysis (Proost et al., 1997). Alternatively, the low recovery may indicate that the parent drug or its metabolites are effectively stored in other tissues, e.g., those rich in acid mucopolysaccharides (Waser, 1973). Binding to such macromolecules may attribute to a very slow elimination of a fraction of the administered dose. Therefore, it is hypothesized that a fraction of the dose of the investigated compounds remains in the body for a prolonged time, both in the liver and in other tissues.

It would be interesting to know to what extent the PK data of these compounds in cats are predictive for their behavior in humans. Because several of the tested compounds have been investigated in PK studies in humans, a limited interspecies comparison can be made. These data will be reported in a separate article, together with the dynamic data obtained in the present study.