Materials and Methods

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.

• Download figure
• Open in new tab
• Download powerpoint

Figure 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 · kg⁻¹ i.p. Anesthesia was maintained by pentobarbital sodium (4.8 mg · kg⁻¹ · h⁻¹) 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.

The bile duct and urinary bladder were cannulated according to the procedure described in Bencini et al. (1985). Laparotomy was performed through a cranial midline incision extending from the xiphisternum to the umbilicus. The cystic duct was ligated and the common bile duct was cannulated. After a caudal midline incision, the urethra was exposed and cannulated. The abdominal wound was carefully closed.

The twitch response of the left musculus tibialis anterior elicited by supramaximal square wave stimuli of 0.2 ms duration applied to the common peroneal nerve at 0.1 Hz was recorded throughout the experiment. The NMBA was administered as a bolus injection in a dose of two times the estimated ED₉₀ via a catheter in the right vena saphena, after dissolving in an appropriate amount of an isotonic aqueous solution buffered at pH 4 (Organon Labs. Ltd.). The maximum neuromuscular block (twitch height depression expressed as percentage of control twitch height), onset time (defined as the time elapsed between the end of the injection and the maximal block, or to 100% block if the neuromuscular block was complete), recovery index (defined as the time between recovery from 25 to 75% of the control value), and the duration of action (defined as the time elapsed between the end of the injection and 90% recovery) were calculated from the recorded twitch height.

Blood samples were taken from the arteria femoralis catheter before dosing, and at 1, 2, 3, 4, 5, 7, 10, 15, 20, 30, 40, 60, 90, 120, 240, 360, and 480 min after administration. Blood samples were cooled in ice immediately, and plasma was obtained by centrifugation. To prevent hydrolysis of the drug, the plasma samples were acidified with a measured volume of 1 M sodium hydrogenphosphate to pH ±5. Blank urine and bile samples were collected for 30 min. After administration of the NMBA, urine and bile were collected in a measured volume of 1 M sodium hydrogenphosphate at 30-min intervals for 120 min, and at 60-min intervals for up to 480 min.

After 480 min, the experiment was terminated. The liver was excised, weighed, and a 5-g sample was homogenized in 45 ml of 1 M sodium hydrogenphosphate, cooled in ice. The acidified samples of plasma, urine, bile, and liver homogenate were kept frozen at −20°C until analysis. It was checked that the compounds were stable at the conditions of storage.

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.

The internal standard was a quaternary aminosteroidal compound, selected for each compound separately. Liver homogenates were extracted after mixing 100 to 1000 μl with 1 ml of KI-glycine buffer (prepared by solving 6.4 g potassium iodide in a mixture of 4 ml of 0.1 N sodium hydroxide and 6 ml of a solution containing 45.4 mg of glycine and 35.1 mg of sodium chloride). After addition of the internal standard and dichloromethane, the samples were treated in the same way as plasma and bile samples. Urine samples were injected into the HPLC system without pretreatment.

The limit of quantification was ∼10 ng in the prepared sample. The lower limit of quantification in plasma was in the range of 10 to 50 ng · ml⁻¹, depending on the compound to be analyzed. Concentrations and amounts of the NMBAs and their metabolites have been expressed in molar units.

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).

The minimization procedure of the residual sum of squares was performed with the Marquardt algorithm (Press et al., 1986), with initial estimates obtained by an automated curve-stripping procedure. The fitting procedure stopped when the relative improvement of the residual sum of squares was <10⁻¹⁰, and the relative change of each parameter was <10⁻⁵.

Goodness-of-fit was evaluated from visual inspection of the measured and calculated data points and of the residuals plotted against time and against concentration. Also, the residual coefficient of variation and the standard errors of the estimated model parameters, determined from the variance-covariance matrix (Veng Pedersen, 1977; Draper and Smith, 1981), were used as measures of goodness-of-fit. The choice between the two- and three-exponential equation was based on theF test (Boxenbaum et al., 1974), accepting a more complex model as significantly better fitting if P < .05.

The volume of the central compartment (V₁) and peripheral compartments (V₂, V₃), steady-state volume of distribution (V_ss), plasma clearance (CL), distribution clearance to peripheral compartments (CL₁₂, CL₁₃), initial plasma clearance (CL_init, defined as the sum of elimination and distribution clearance at time zero), half-lives (t_1/2 1,t_1/2 2,t_1/2 3), and mean residence time (MRT) were calculated with standard equations, assuming that elimination takes place from the central compartment (Wagner, 1975;Cutler, 1987). The unbound clearance (CLu) was estimated by dividing the clearance by the fraction unbound (Proost et al., 1997).

In several cases, the number of data points in the individual animals was insufficient to obtain reliable estimates of the PK parameters. Therefore, the data also were analyzed as described above, after pooling all available plasma concentration measurements for each compound.

The fraction of the dose excreted in urine unchanged (f_e) and in the form of one of the putative metabolites was calculated for each animal. Also, the urinary excretion data were evaluated by iterative nonlinear regression as described above. Instead of the plasma concentration, the fraction excreted in each sampling interval was used. An additional parameter was added to allow for a time lag between drug administration and appearance at the site of sampling. From the parameters of the exponential equation, the half-lives and the total amount excreted (calculated as the sum of the measured amounts excreted and the calculated amount to be excreted after the last sampling time, extrapolated to infinity, expressed as a percentage of the dose) were obtained. Biliary excretion data were analyzed by the same procedure. Renal clearance (CL_renal) and biliary clearance (CL_bile) were calculated as the product of the plasma clearance and the fraction excreted in urine and bile, respectively.

Correlation Analysis.

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

Results

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 (Table1).

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%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 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.

The initial half-life ranged from 0.4 to 1.4 min, and the half-life was between 3 and 10 min for the second phase; these values were much shorter than for the traditional compounds pancuronium, pipecuronium, and vecuronium (Table 2). For each of the investigated compounds, the clearance was in the range from 24 to 58 ml · min⁻¹ · kg⁻¹; these values were much higher than that for the older compounds pancuronium, pipecuronium, and vecuronium. The distribution volume of the central compartment was in the range from 36 to 124 ml · kg⁻¹, which is equal to or larger than the plasma volume. The volume of distribution at steady state varied widely between 71 and 247 ml · kg⁻¹. Pancuronium and vecuronium exhibit a V_ss at the upper side of this range, whereas V_ss of pipecuronium is even higher. The mean residence of the investigated compounds was very short, in the range of 1.5 to 8 min; these values were much shorter than that of the older compounds. With the three-compartment model, the plasma half-lives of Org 9489 and Org 9487 were larger (13.8 and 14.4 min, respectively) than for the two-compartment model (Table 2). For both compounds, clearance (29 and 22 ml · min⁻¹ · kg⁻¹, respectively) and V_ss (166 and 108 ml · kg⁻¹, respectively) for the three-compartment model were slightly lower than for the two-compartment model (Table 2), and MRT (5.7 and 5.0 min for Org 9489 and Org 9487, respectively) was almost similar for both models. The 3-OH derivatives were present in plasma only in very low concentrations, close to or below the limit of quantification; other metabolites were not found in plasma.

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 (Table3). 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%).

The urinary excretion rate profiles of the parent compounds and the 3-OH metabolites exhibited a monophasic or biphasic pattern (Fig.3). In many cases, it was not possible to obtain a satisfactory fit; for Org 20059, Org 7268, and Org 9991 due to the low number of data points (less than four), and in other cases because of an irregular excretion profile. In none of the profiles could a reliable estimate of a second half-life be obtained. The values for the half-life of the parent compound of the initial phase are listed in Table 4; the half-life of rocuronium was 40 min; for the other compound the values ranged from 11 to 17 min. The half-life of the 3-OH metabolites was somewhat larger than that of the parent compound (Fig. 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 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 (Table3). 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).

The biliary excretion rate profiles of both the parent compounds and the 3-OH metabolites exhibited a biphasic pattern (Fig.4). In all experiments, a satisfactory fit could be obtained for a two-exponential equation; however, the residual coefficient of variation and the relative standard errors of the parameters were relatively large. Considering only the parameters that could be estimated with a relative standard error not exceeding 40%, the time lag between administration and appearance of the parent compound at the sampling site ranged from zero to 17 min, the rapid half-life from 8 to 23 min, and the terminal half-life from 119 to 489 min (Table 4). For the 3-OH metabolites these values were comparable, i.e., 6 to 20 min, 13 to 23 min, and 158 to 529 min, respectively. The estimated amounts of drug to be excreted after the last time point at 480 min were low, ranging from 0.3 to 5% of the dose for the parent compound, and from 0.4 to 3.9% for the 3-OH metabolites; however, for Org 7268 the estimated amount to be excreted was 13.4% of the dose.

• Download figure
• Open in new tab
• Download powerpoint

Figure 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 Table5, 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.

No significant correlation with lipophilicity was found for plasma clearance, initial plasma clearance, and V₁. There was a significant positive correlation between lipophilicity and unbound plasma clearance (Fig. 5) and unbound initial plasma clearance, and a significant negative correlation between lipophilicity and both plasma half-lives, V_ss (Fig. 6), and MRT. The fraction excreted unchanged into urine was found to be negatively correlated to the lipophilicity of the compounds. However, renal clearance and unbound renal clearance were not correlated to lipophilicity. Also, the fraction excreted unchanged into bile, time lag, half-lives of biliary excretion, biliary clearance, and unbound biliary clearance were not correlated to lipophilicity.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Relationship between lipophilicity (log P) and unbound clearance in the cat.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Relationship between lipophilicity (log P) and steady-state volume of distribution in the cat.

Discussion

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 byKhuenl-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 · min⁻¹ · kg⁻¹; we found a value of 58 ml · min⁻¹ · kg⁻¹. The values for renal excretion of rocuronium and Org 7268 reported byKhuenl-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 · min⁻¹ · kg⁻¹) 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 · min⁻¹ · kg⁻¹(Org 9991). Assuming a glomerular filtration rate of ∼2.5 ml · min⁻¹ · kg⁻¹in 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.