Faculté de pharmacie, Université de Montréal, 2900 boul. Edouard Montpetit, C.P. 6128, Succursale Centre-ville, Montréal (Québec), H3C 3J7, Canada
* Corresponding author. E-mail: france.varin{at}umontreal.ca
Accepted for publication March 11, 2004.
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methods. The octanol/phosphate buffer distribution coefficients (logD) of various NMBA (vecuronium, rocuronium, mivacurium isomers (cis-cis, cis-trans and trans-trans), doxacurium, cisatracurium, atracurium, succinylcholine) were determined. The free fraction for each drug was measured using an ultrafiltration technique. PK-PD descriptors were obtained from selected clinical studies. Correlations between physicochemical parameters (including molecular weight) and PK-PD descriptors were assessed by linear or multiple linear regression.
Results. A wide range of log D (4.15 for succinylcholine to 0.75 for vecuronium) and free fraction (from 31% for vecuronium to 80% for succinylcholine) is observed for NMBA. Molecular weight combined with either lipid solubility (r2=0.70; P=0.001) or free fraction (r2=0.84; P<0.001) were highly correlated with potency, while for ke0 a greater degree of correlation was obtained when both lipid solubility and free fraction (r2=0.74; P=0.002) were included.
Conclusions. The basic characteristics of NMBAs, namely, molecular weight, lipid solubility and protein binding, are strongly associated with the kinetics of the drug response.
Keywords: neuromuscular block, agents ; pharmacokinetics, potency ; protein, binding ; solubility, lipid
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Pharmacokinetics (cardiac output, transcapillary rate of transfer, plasma clearance, and dose) as well as pharmacodynamics (potency and mechanism of action) are known to influence the profile of activity of neuromuscular blocking drugs.26 In turn, the physicochemical properties of these drugs would also be expected to have an impact on the estimates of pharmacokinetic and pharmacodynamic descriptors. Currently, few data concerning the lipid solubility of neuromuscular blocking drugs are available. For a series of aminosteroidal agents, Wierda and colleagues have reported a relation between lipid solubility or protein binding and various pharmacokinetic and pharmacodynamic (PK-PD) descriptors.7 8 However, more studies are required to enable generalization of these results to other families of compounds.
Our study aimed to characterize these physicochemical factors (lipid solubility, in vivo protein binding) for several neuromuscular blocking agents of differing chemical structure. A second objective was to assess the degree of association between these factors (including molecular weight), and either the potency (EC50) or the plasma and effect compartment concentration equilibration rate constant (ke0) obtained for these drugs in previous clinical studies.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The distribution coefficient of each drug was determined in octanol/buffer. Stock solutions of the test drugs were diluted in octanol-saturated buffer (referred to as the aqueous phase thereafter) to obtain a final concentration range of 100 µg ml1 (doxacurium, mivacurium isomers, rocuronium, vecuronium) or 200 µg ml1 (atracurium, cisatracurium, succinylcholine). Samples consisting of 400 µl of octanol and 100 µl of the aqueous phase were submitted to slow agitation for 1 h and centrifuged at 2000 g for 5 min. The octanol and aqueous phases were separated, evaporated (Speed Vac®, Savant Instruments Inc., Farmingdale, NY, USA) and resuspended into the HPLC mobile phase before analysis. For atracurium and cisatracurium, centrifugation was carried out at 10°C and the samples immediately stabilized with the HPLC mobile phase (pH 3). On the same day, quantitative determination of model drugs in both phases was carried out in replicate (n: 36). With the exception of succinylcholine, experiments were repeated on different days (n: 35). The distribution coefficient (log D) was calculated as the logarithm of the ratio obtained by dividing the total concentration (neutral and ionized forms) of the molecule in the octanol phase by that in the aqueous phase.9
Protein binding determination
Informed consent was obtained from each subject before blood donation. Five healthy adults, four women and one man, mean age 43 (5.5) (range 3548) yr, American Society of Anesthesiologists status I, who were not taking any medication were recruited. Volunteers fasted for at least 12 h before blood sampling. Blood samples (70 ml) were collected in sodium citrate enriched tubes and centrifuged (2000 g for 15 min). Plasma was separated into five aliquots of approximatively 6 ml, flash frozen on dry ice and stored at 70°C.
In vitro plasma protein binding was determined for rocuronium, cisatracurium, and succinylcholine. A method similar to that described by Cameron and colleagues10 was used on thawed samples. For succinylcholine, non-specific plasma esterases were inhibited by the addition of echothiophate (20 µl of echothiophate 0.2% per ml of plasma). Plasma samples were spiked extemporaneously by the addition of model drug at concentrations varying from 2000 ng ml1 (cisatracurium and rocuronium) to 20 000 ng ml1 (succinylcholine). These concentrations were representative of the in vivo exposure and fell within the range of the analytical assay. Spiked plasma samples were immediately flash frozen for determination of total concentration. Aliquots of blank plasma were also analysed. Free concentration (fu) of model drug was determined at room temperature in duplicate as follows: 1 ml plasma was applied onto microseparation devices having a molecular weight cut-off of 30 000 (Centrifree, Amicon Micropartition System, Beverly, MA). The microseparation devices were centrifuged at 1200 g for 5 min. A precise aliquot (50 or 100 µl) of the ultrafiltrate was then transferred into 1 ml pre-acidified blank plasma and flash frozen until HPLC analysis. The unbound fraction was calculated by dividing drug concentration in the ultrafiltrate by total drug concentration.
HPLC analysis
Drug concentrations were determined using HPLC-UV for doxacurium,11 HPLC-fluorescence for atracurium, cisatracurium, and mivacurium,12 13 HPLC-EC for rocuronium and vecuronium,14 and tandem mass spectrometry for succinylcholine.15
PD descriptors
In order to assess the degree of correlation between commonly used physicochemical parameters and the PK-PD relationship, the effect compartment concentration corresponding to 50% block (EC50), often referred to as the potency factor, and the equilibration rate constant between plasma and the effect compartment (ke0) were chosen as descriptors. In order to minimize the inherent variability of clinical studies, mean values published by our group for these model drugs, vecuronium,14 16 rocuronium,17 mivacurium,18 doxacurium,11 19 cisatracurium,12 20 atracurium,21 22 and succinylcholine23 were used to establish correlations (Table 1). In studies where isoflurane was used as the anaesthetic agent, a correction factor of 1.7 was applied to the EC50 value to account for its well-established potentiating effect.17 In six studies14 17 18 19 22 23 intensive arterial blood sampling was carried out for 2 min after the administration of a bolus dose of the NMBA and at frequent intervals thereafter. Monitoring of neuromuscular function was carried out using single twitch or train-of-four stimulation, in agreement with good clinical research practice standards.24
|
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
In Table 1, the free fraction of drug in plasma measured for cisatracurium, rocuronium, and succinylcholine is presented. The results given for the other neuromuscular blocking agents (atracurium, mivacurium, doxacurium, and vecuronium) were published previously by our laboratory using a similar approach.10
We tested if a correlation existed between ke0 and different characteristics of the neuromuscular blocking drugs (molecular weight, log D, or free fraction). For the ke0, log D, as a single independent variable, proved to yield the highest association for our model drugs. A poor correlation was found between ke0 and the free fraction (r2=0.07) or molecular weight (r2=0.16). In Figure 3, the linear dependence of ke0 on log D for a series of neuromuscular blocking agents is illustrated (r2=0.45, P=0.016). The addition of the free fraction as a covariable results in a significant improvement in the determination coefficient (r2=0.74, P=0.002). The Multiple Partial F-test and the Partial F-test gave P values smaller than 0.001, which supports the addition of the free fraction as a cofactor in our regression. The equation describing this multiple regression is:
![]() |
|
Molecular weight, as a single independent variable, proved to have the highest degree of association with the EC50 of our model drugs (Fig. 4). A poor degree of association was found between the EC50 (nmol ml1) and the free fraction (r2=0.11) or log D (r2=0.002). The linear dependence of the EC50 on molecular weight is illustrated (r2=0.63, P=0.002). The addition of log D as a covariable results in an improvement in the determination coefficient (r2=0.77, P=0.001, data not presented). The Multiple Partial F-test and the Partial F-test gave P values smaller than 0.05, which supports the addition of the log D as a cofactor in our regression. The equation describing this multiple regression is:
![]() |
|
![]() |
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The onset time reported for a neuromuscular blocking agent in a clinical study will vary according to methodological factors such as the type of anaesthesia, the dose administered, the type of stimulation (single twitch vs train-of-four response), and the clinical end-point used for onset time determination (peak effect or fixed degree of twitch depression) during neuromuscular monitoring.4 25 As the clinical protocols used to obtain our reference values were not identical, we elected to establish correlations between the drug's physicochemical parameters and the ke0 or EC50 values instead of the onset times, as the former parameters should be dose independent and can be adjusted to account for methodological differences (e.g. use of inhalation agents).
Circulatory factors are recognized to have a direct impact on ke0 values that may, in turn, alter the onset of action of a drug.11 26 27 These results suggest that, for a given NMBA and under identical clinical settings, inter-individual variations in ke0 become sufficiently large to distinguish the impact of ke0 on onset times from that of other confounding factors. Alternatively, onset times may vary without any detectable changes in ke0.19 26 Furthermore, an equilibration delay similar to that of other NMBAs was reported recently for succinylcholine.23 These findings suggest that the limiting factors for ke0 and onset times are different, and that their net effects may differ.
Three physicochemical properties of a drug (partition coefficient, protein binding, and molecular weight) affect its transfer from blood to tissue. As neuromuscular blocking agents are large ionized molecules, their transcapillary rate of transfer in muscle tissue is expected to be mostlypore-restricted (intercellular gaps of 1020 nm).28 However, the relative contribution of the large number of smaller pores is uncertain. The molecular span of many NMBAs is approximately 12 nm between both quaternary amines.29 Therefore, molecular size and weight are not expected to be limiting factors for the transcapillary filtration of NMBAs.30 Accordingly, no correlation was observed between the ke0 value and molecular weight for our series of NMBAs.
Diffusion across the endothelial membrane is not expected to be important for ionized drugs such as neuromuscular blocking agents.31 According to Fick's law, the rate of diffusion would depend on the solute's unbound concentration, its partition coefficient, and its degree of ionization. However, membrane permeability is also related to the surface occupied by the pores in different endothelia and will differ according to the size and shape of each molecule.32 33 These physicochemical factors are thus expected to have a direct impact on the onset times of NMBAs.
For many years, organic solvent/aqueous partitioning systems have been used as predictors for biological membrane partitioning of anaesthetic agents.34 The log D coefficient seemed more appropriate for neuromuscular blocking agents because these drugs are fully ionized at pH 7.4. The log P value would have required that the degree of ionization at a given pH be considered as it assumes that only the neutral form of the compound is the partitioning species.35 36 Recently, the lipid solubility of a series of aminosteroidal neuromuscular blocking agents and related investigational compounds was documented using the octanol-buffer solvent as the partitioning system.7 37 For vecuronium, the partition coefficient was similar to ours (0.77 vs 0.75) but not for rocuronium (0.80 vs 1.68). In our study, rocuronium was found to be more hydrosoluble than vecuronium and this would be compatible with its higher polarity (Fig. 1B).
Although NMBAs are intrinsically very polar, individual log D values vary considerably. Indeed, almost a four log scale difference in lipid solubility is observed between succinylcholine and vecuronium. Overall, benzylisoquinoliums were found to be largely more water soluble (log D values less than 2) than the aminosteroidal series, probably because of their double positive charge.
Wierda and Proost7 found a positive correlation (r2=0.866) between lipid solubility and the ke0 values of a series of aminosteroidal compounds administered to anaesthetized patients. Lipid solubility was found to be the primary variable responsible for the time course of action but this relation was significantly improved when the free fraction was also taken into account. Our findings are in agreement with their study. Moreover, these results suggest that this association between lipid solubility and ke0 could be applied to most neuromuscular blocking agents.
In our study, the free fraction varied considerably, that is from 31% (vecuronium) to 80% (succinylcholine). It is generally agreed that in a series of aminosteroidal compounds, protein binding will increase as lipid solubility increases.7 37 A definite relation between both factors was observed for our study drugs (r2=0.67, P=0.002), and that independently of the chemical class. Adding the free fraction as a co-factor greatly improved the degree of association between log D and ke0.
There is substantial evidence to confirm that drug potency plays a major role in the onset of action of a neuromuscular blocking agent. An inverse relationship between onset times and molar potency (ED50) has been established for a series of aminosteroidal neuromuscular blocking agents in anaesthetized patients.4 38 This observation was supported by in vitro animal preparations,39 and further explained by a kinetic-dynamic model.40 According to Lee,30 the inverse correlation between NMBA potency and lipid solubility (r2=0.812) reported by Wierda and Proost7 for the aminosteroidal series could be coincidental with other factors such as steric hindrance. Indeed, physicochemical characteristics such as the large molecular size and the quaternary ammonium structure of NMBAs may have a definite impact on the onset of these drugs. These observations are reinforced by the limited space available for diffusion within the synaptic cleft (3050 nm).41 In turn, the repetitive binding to the acetylcholine receptor is enhanced because the nerve terminal presents a physical barrier to diffusion out of the cleft.42 In our study, where molecules from different chemical families were tested, the best predictive factor for the molar potency was molecular weight. This finding is in support of Lee's hypothesis and the analysis made by Ramzan,1 who first proposed molecular weight as a determinant of the speed of onset for a series of NMBAs. Inclusion of succinylcholine in our series of model drugs allowed a greater range of molecular weight to be studied. Its deletion from the series had very little impact on the multiple linear regression for the descriptors ke0 (r2=0.77, P=0.003) and EC50 (r2=0.79, P=0.002). However, it is possible that its different mechanism of action could be a confounding factor.
In our study, the linear dependence of drug potency on molecular weight was improved by the addition of either lipid solubility or protein binding as a covariable. Indeed, a linear correlation was observed previously between the molecular weight and the partition coefficient of a series of organic cations.43 Thus, the added value of fu or log D to the predictive value of the EC50 would be consistent with Wierda and colleagues's observations.7 37 In their study, the range of molecular weight was too small to be considered as an independent variable and was therefore not tested. The fact that, in our study, lipid solubility was not found to be a major determinant of potency but a synergistic factor to molecular weight supports Lee's hypothesis.
In conclusion, the basic characteristics of NMBAs, namely, molecular weight, lipid solubility and protein binding, are strongly associated with the kinetics of drug response.
![]() |
Acknowledgments |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Swerdlow BN, Holley FO. Intravenous anaesthetic agents pharmacokinetic-pharmacodynamic relationships. Clin Pharmacokinet 1987; 12: 79110[ISI][Medline]
3 Hennis PJ, Stanski DR. Pharmacokinetic and pharmacodynamic factors that govern the clinical use of muscle relaxants. Semin Anesth 1985; 4: 2130
4 Kopman AF, Klewicka MM, Kopman DJ, Neuman GG. Molar potency is predictive of the speed of onset of neuromuscular block for agents of intermediate, short, and ultrashort duration. Anesthesiology 1999; 90: 42531[CrossRef][ISI][Medline]
5 Kopman AF, Klewicka MM, Neuman GG. Molar potency is not predictive of the speed of onset of atracurium. Anesth Analg 1999; 89: 10469
6 Beaufort TM, Nigrovic V, Proost JH, Houwertjes MC, Wierda JM. Inhibition of the enzymic degradation of suxamethonium and mivacurium increases the onset time of submaximal neuromuscular block. Anesthesiology 1998; 89: 70714[ISI][Medline]
7 Wierda JMKH, Proost JH. Structure-pharmacodynamic-pharmacokinetic relationships of steroidal neuromuscular blocking agents. Eur J Anaesthesiol 1995; 12(Suppl 11): 4554[ISI]
8 Wierda JMKH, Proost JH, Muir AW, Marshall RJ. Design of drugs for rapid onset. Anaesth Pharmacol Rev 1993; 1: 5768
9 Barbato F, La Rotonda MI, Quaglia F. Interactions of nonsteroidal antiinflammatory drugs with phospholipids: comparison between octanol/buffer partition coefficients and chromatographic indexes on immobilized artificial membranes. J Pharm Sci 1997; 86: 2259[CrossRef][ISI][Medline]
10 Cameron M, Donati F, Varin F. In vitro plasma protein binding of neuromuscular blocking agents in different subpopulations of patients. Anesth Analg 1995; 81: 101925[Abstract]
11 Gariepy LP, Varin F, Donati F, Salib Y, Bevan DR. Influence of aging on the pharmacokinetics and pharmacodynamics of doxacurium. Clin Pharmacol Ther 1993; 53: 3407[ISI][Medline]
12 Bergeron L, Bevan DR, Berrill A, Kahwaji R, Varin F. Concentration-effect relationship of cisatracurium at three different dose levels in the anesthetized patient. Anesthesiology 2001; 95: 31423[CrossRef][ISI][Medline]
13 Bryant BJ, James CD, Cook DR, Harrelson JC. High performance liquid chromatography assay for cisatracurium and its metabolites in human urine. J Liquid Chromatogr Relative Technol 1997; 20: 204151[ISI]
14 Ducharme J, Donati F. Pharmacokinetics and pharmacodynamics of steroidal muscle relaxants. Anesthesiol Clin North Am 1993; 11: 283307
15 Roy JJ, Boismenu D, Gao H, Mamer OA, Varin F. Measurement of succinylcholine concentration in human plasma by electrospray tandem mass spectrometry. Anal Biochem 2001; 290: 23845[CrossRef][ISI][Medline]
16 Alloul K, Whalley DG, Shutway F, Ebrahim Z, Varin F. Pharmacokinetic origin of carbamazepine-induced resistance to vecuronium neuromuscular blockade in anesthetized patients. Anesthesiology 1996; 84: 3309[CrossRef][ISI][Medline]
17 Dragne A, Varin F, Plaud B, Donati F. Rocuronium pharmacokinetic-pharmacodynamic relationship under propofol and isoflurane anesthesia. Can J Anaesth 2002; 49: 35360
18 Laurin J, Donati F, Nekka F, Varin F. Peripheral link model as an alternative for pharmacokinetic-pharmacodynamic modeling of drugs having a very short elimination half-life. J Pharmacokinet Biopharm 2001; 28: 725[Medline]
19 Zhu Y, Audibert G, Donati F, Varin F. Pharmacokinetic-pharmacodynamic modeling of doxacurium: Effect of input rate. J Pharmacokinet Biopharm 1997; 25: 2337[CrossRef][ISI][Medline]
20 Tran TV, Fiset P, Varin F. Pharmacokinetics and pharmacodynamics of cisatracurium after a short infusion in patients under propofol anesthesia. Anesth Analg 1998; 87: 115863[Abstract]
21 Donati F, Gill SS, Bevan DR, Ducharme J, Theoret Y, Varin F. Pharmacokinetics and pharmacodynamics of atracurium with and without previous suxamethonium administration. Br J Anaesth 1991; 66: 55761[Abstract]
22 Ducharme J, Varin F, Donati F. Pharmacokinetics and pharmacodynamics of a second dose of atracurium in anaesthetised patients. Clin Drug Invest 1995; 9: 98110[ISI]
23 Roy JJ, Donati F, Boismenu D, Varin F. Concentration-effect relation of succinylcholine chloride during propofol anesthesia. Anesthesiology 2002; 97: 108292[CrossRef][ISI][Medline]
24 Viby-Mogensen J, Engbaek J, et al. Good clinical research practice (GCRP) in pharmacodynamic studies of neuromuscular blocking agents. Acta Anaesthesiol Scand 1996; 40: 5974[ISI][Medline]
25 Bartkowski RR, Witkowski TA, Azad S, Lessin J, Marr A. Rocuronium onset of action: a comparison with atracurium and vecuronium. Anesth Analg 1993; 77: 5748[Abstract]
26 Wright PM, Brown R, Lau M, Fisher DM. A pharmacodynamic explanation for the rapid onset/offset of rapacuronium bromide. Anesthesiology 1999; 90: 1623[ISI][Medline]
27 Kuipers JA, Boer F, Olofsen E, Bovill JG, Burm AG. Recirculatory pharmacokinetics and pharmacodynamics of rocuronium in patients. Anesthesiology 2001; 94: 4755[CrossRef][ISI][Medline]
28 Renkin EM. Microcirculation and exchange. In: Patton HD, Fuchs AF, Hille B, Scher AM, Steiner R, eds. Textbook of Physiology. Circulation, Respiration, Body Fluids, Metabolism and Endocrinology, 21st Edn. Philadelphia: WB Saunders Company, 1989; Vol. 2: 86078
29 Bovet D, Bovet-Nitti F. Structure et Activité Pharmacodynamique des Médicaments du Système Nerveux Végétatif. Bâle: S.Karger, 1948
30 Lee C. Structure, conformation, and action of neuromuscular blocking drugs. Br J Anaesth 2001; 87: 75569
31 Perrault C, Guay J, Gaudreault P, Cyrenne L, Varin F. Residual curarization in the neonate after Caesarean section. Can J Anaesth 1991; 38: 58791[Abstract]
32 Van Bree JBMM, De Boer AG, Danhof M, Ginsel LA, Breimer DD. Characterization of an in vitro blood-brain barrier: effects of molecular size and lipophilicity on cerebrovascular endothelial transport rates of drugs. J Pharmacol Exp Ther 1988; 247: 12339[Abstract]
33 Lien EJ, Wang PH. Lipophilicity, molecular weight, and drug action: reexamination of parabolic and bilinear models. J Pharm Sci 1980; 69: 64850[ISI][Medline]
34 Overton CE. Studien uber die Narkose, Jugleich ein Betrag zue Allgemeinen Pharmakologie. Jena: Fisher,G., 1901
35 Barton P, Davis AM, McCarthy DJ, Webborn PJH. Drug-phospholipid interactions. 2. Predicting the sites of drug distribution using n-octanol/water and membrane/water distribution coefficients. J Pharm Sci 1997; 86: 10349[CrossRef][ISI][Medline]
36 Martin YC. Calculation of the Physical Properties of Compounds. Quantitative Drug Design. New York: Marcel Deker, 1978; 62114
37 Proost JH, Roggeveld J, Wierda JM, Meijer DK. Relationship between chemical structure and physicochemical properties of series of bulky organic cations and their hepatic uptake and biliary excretion rates. J Pharmacol Exp Ther 1997; 282: 71526
38 Kopman AF. Pancuronium, gallamine, and d-tubocurarine compared: is speed of onset inversely related to drug potency? Anesthesiology 1989; 70: 91520[ISI][Medline]
39 LawMin JC, Bekavac I, Glavinovic MI, Donati F, Bevan DR. Iontophoretic study of speed of action of various muscle relaxants. Anesthesiology 1992; 77: 3516[ISI][Medline]
40 Donati F, Meistelman C. A kinetic-dynamic model to explain the relationship between high potency and slow onset time for neuromuscular blocking drugs. J Pharmacokinet Biopharm 1991; 19: 53752[ISI][Medline]
41 Couteaux R. Structure and cytochemical characteristics of the neuromuscular junction. In: Cheymol J, ed. Neuromuscular Blocking and Stimulating Agents. Oxford: Pergamon Press, Ltd, 1972; 756
42 Armstrong DL, Lester HA. The kinetics of tubocurarine action and restricted diffusion within the synaptic cleft. J Physiol 1979; 294: 36586[Abstract]
43 Neef C, Meijer DK. Structure-pharmacokinetics relationship of quaternary ammonium compounds. Correlation of physicochemical and pharmacokinetic parameters. Naunyn Schmiedebergs Arch Pharmacol 1984; 328: 1118[ISI][Medline]
44 Donati F, Varin F, Ducharme J, Gill SS, Theoret Y, Bevan DR. Pharmacokinetics and pharmacodynamics of atracurium obtained with arterial and venous blood samples. Clin Pharmacol Therapeut 1991; 49: 51522[ISI][Medline]
|