Rocuronium bromide (ORG 9426) is a new steroidal nondepolarizing neuromuscular blocking drug that has a faster onset of action than other nondepolarizing neuromuscular blocking drugs . In animals, rocuronium appears to be eliminated primarily via the liver. For example, in cats, 54% of an intravenous (IV) dose is found in the bile and 21% in liver homogenate . In addition, when the liver is excluded from the circulation (via a portal vein-to-inferior vena cava shunt), rocuronium's clinical duration of action increases almost threefold . Finally, only 10% of an IV dose is excreted in urine [2,3]. These findings suggest that rocuronium's pharmacokinetics, and possibly its time course of neuromuscular effects, might be altered by liver disease. To examine this possibility, we determined the pharmacokinetics and the onset and clinical duration of rocuronium in patients with liver disease and a group of patients with normal hepatic and renal function.
With approval from our Committee on Human Research, we obtained written informed consent from 20 surgical patients--10 healthy patients (ASA class I-II) and 10 with hepatic dysfunction (ASA class III). Hepatic dysfunction was defined as the presence of one or more of the following: 1) hepatic cirrhosis diagnosed by biopsy or ultrasound; 2) ascites and history of ethanol abuse; and, 3) alcohol-related increase of serum transaminases (five times normal aspartate aminotransferase or alanine aminotransferase) within 36 h of surgery. Patients also were assigned the appropriate Child's classification based on their laboratory data and physical findings . Exclusion criteria were evidence of neuromuscular disease, use of medications known to influence neuromuscular function, or obesity (>30% above ideal body weight). Although the control and hepatic disease patients were studied concurrently, pharmacokinetic data for the control patients have been reported previously .
After administration of midazolam, 0.02-0.05 mg/kg IV, an intraarterial or IV catheter was placed to permit sampling of blood for determination of plasma rocuronium concentration. Anesthesia was induced with thiopental, 1-5 mg/kg IV, followed by inhalation of 1%-4% isoflurane in oxygen. The trachea was intubated without administration of a muscle relaxant. Anesthesia was maintained with 60% nitrous oxide and isoflurane (end-tidal concentrations of 1.0% +/- 0.1% in hepatic patients, 1.2% +/- 0.5% in control patients), monitored by mass spectrometry. Ventilation was controlled to maintain end-tidal PCO2 between 30 and 40 mm Hg. Esophageal temperature was maintained at 35-37 degrees C. Routine monitoring consisted of electrocardiography, automated blood pressure, and pulse oximetry.
Prior to administration of rocuronium, the mechanical evoked response of the adductor pollicis muscle to the first stimulus (T1) in a train-of-four sequence was determined. A Grass S88 nerve stimulator (Grass Medical Instruments, Quincy, MA) delivered supramaximal, train-of-four square-wave impulses of 0.2-ms duration (2 Hz) via surface electrodes near the ulnar nerve at the wrist. The stimulus sequence was repeated at intervals of 12 s and the evoked mechanical response of the adductor pollicis muscle was quantified by a force-displacement transducer (APMX Registered Trademark; Professional Instruments, Houston, TX) and recorded. When the amplitude of T1 was stable for > 15 min, this peak amplitude response was used as the baseline for comparison of all subsequent responses.
After anesthetic conditions and twitch tensions were stable for >or=to 15 min, rocuronium, 0.6 mg/kg [twice the ED95[5,6]], was injected rapidly. Twitch tension was recorded until the end of surgery or until additional muscle relaxation was required and provided by atracurium (which does not interfere with measurement of rocuronium). Residual neuromuscular blockade at the end of surgery was antagonized by administration of neostigmine and glycopyrrolate (40-70 micro gram/kg and 8-15 micro gram/kg, respectively). The times from injection of rocuronium to complete ablation of T1 (onset) and T1 return to 25% of baseline (clinical duration) were recorded.
Venous or arterial blood samples (4 mL each) were drawn from the contralateral arm before and at 2, 4, 6, 8, 10, 15, 20, 25, 30, 40, 50, 60, 75, 90, 120, 150, 180, 210, 240, 300, and 360 min after administration of rocuronium. In the first seven control patients, samples were obtained for 240 min only. Blood samples were anti-coagulated with heparin, iced, and centrifuged within 1 h of collection. Plasma was mixed with an equivalent volume of 0.8 M sodium dihydrogen phosphate and was stored at -30 degrees C until analysis. After methylene chloride extraction and derivatization, plasma concentrations of rocuronium were determined using gas chromatography with nitrogen-phosphorus detection . This method is sensitive to 10 ng/mL with a coefficient of variation of < 10% at 20 ng/mL. Chromatograms also were examined for the presence and magnitude of peaks consistent with ORG 9943 (17-desacetylrocuronium), a putative metabolite of rocuronium.
A population pharmacokinetic analysis was performed using NONMEM  to estimate "typical values" of the pharmacokinetic variables, standard errors of these estimates, and interindividual variability within the population. Two- and three-compartment pharmacokinetic models were fit to the rocuronium plasma concentration data for all patients; models were parameterized in terms of volume of the central (V1), second (V2), and, if appropriate, third compartments (V3); clearance (Cl); rapid distributional clearance (Clrapid, equal to V1 centered dot k12 where k12 is the intracompartmental rate constant for drug movement from the central to the second compartment); and, if appropriate, slow distributional clearance--Clslow, the intracompartmental rate constant for drug movement from the central to third compartment (equal to V1 centered dot k13). Volume of distribution at steady state was calculated as the sum of V1, V2, and V3. Distribution half-lives (t 1/2 pi for the three-compartment model, t 1/2 alpha for both models) and elimination half-life (t 1/2 beta) were determined iteratively (Excel Solver User's Guide, Version 3.0a, Redmond, WA, Microsoft Corporation, 1991).
Variability among patients in Cl, V1, V2, and V3 was modeled by assuming that individual pharmacokinetic values for each variable may be expressed as the sum of the typical value for the population and a factor for the individual. Because interindividual variability tends to be skewed (i.e., log-normally distributed), interindividual variability for Cl was modeled as: Equation 1 where Cli is the estimate for clearance for the ith individual, Cl is the typical value for the population, and eta is a random variable with a mean value of 0.0. Equation 1 can be rewritten as: Equation 2 Interindividual variability for V1, V2, and V3 was modeled similarly, but interindividual variability for V2 and V3 was assumed to be the same.
The effect of liver disease on rocuronium pharmacokinetics was determined by comparing pharmacokinetic models having the same or different "typical values" for Cl, V1, V2, and V3 (alone or in combination) in patients with and without liver disease. These models were compared by their effects on the objective function (NONMEM's equivalent of the residual sum of squares for traditional nonlinear regression analysis), the residual differences between observed and predicted plasma concentrations, and interindividual variability. Liver disease was entered into the model as a dichotomous variable, i.e., absent or present. The influence of liver disease on Cl, V1, V2, and V3 (alone or in combination) was included in the model if it improved the objective function significantly (P < 0.05, i.e., the larger model improved the objective function by 3.8) and by the appearance of the residual differences between measured plasma concentrations and those predicted by the model. Whether the addition of a third pharmacokinetic compartment was justified  and whether estimates of pharmacokinetic variables were improved by normalizing for weight were determined.
Physical characteristics and neuromuscular variables in the two groups were compared by Mann-Whitney U-test. Unless otherwise identified, results are expressed as mean +/- SD.
There were no differences between groups in age, body weight, or gender Table 1. Nine of 10 patients in the hepatic group had alcoholic liver disease (cirrhosis) and the remaining patient, had a diagnosis of hepatoma with >2 L of ascites. Patients underwent general, gynecologic, or orthopedic procedures lasting 3.2 +/- 2.0 h. For technical reasons, blood samples could not be obtained in one patient with liver disease, onset could not be determined in a different patient, and clinical duration could not be determined in three patients. Onset times were similar in the control and hepatic patients. T1 was completely abolished in all patients. Clinical duration tended to be longer in patients with liver disease (P = 0.06).
"Average" plasma concentrations [determined] using a smoother, Supersmoother ; Statistical Sciences, Inc., Seattle, WA] were similar for control and hepatic patients for the first 30 min after rocuronium administration; thereafter, plasma concentrations declined more slowly in patients with liver disease Figure 1 and Figure 2. When a two-compartment pharmacokinetic model was fit to the plasma concentration data, residual differences between the observed and predicted plasma concentrations suggested that this model failed to describe the data adequately. A three-compartment model markedly improved the objective function (P < 0.005), and examination of the residuals indicated no model-misspecification. Neither weight-normalization of pharmacokinetic values nor the influence of liver disease on Cl further improved the objective function. However, objective function was improved by permitting V2 and V3 to differ between patients with normal liver function and those with liver disease; permitting V1, V2, and V3 to differ between groups further improved the fit Table 2. Therefore, "optimal" modeling of the effect of liver disease on rocuronium pharmacokinetics permitted V1, V2, and V3 (and, thus, the derived variables, volume of distribution at steady state, t 1/2 pi, t 1/2 alpha, and t 1/2 beta) to differ between control and hepatic patients. Cl, Clrapid, and Clslow were identical for control and hepatic patients Table 3 and Table 4.
The putative metabolite of rocuronium, 17-desacetylrocuronium (ORG 9943), was not detected.
We found that hepatic dysfunction increased the volume of distribution of rocuronium but did not alter its clearance. Although rocuronium's onset of action was not affected by liver disease, its clinical duration tended to be longer in those patients.
Khalil et al.  reported a similar clinical duration of rocuronium, 0.6 mg/kg, in cirrhotic patients compared to patients with normal liver function (42 +/- 16 vs 41 +/- 7 min, respectively). However, their recovery index (time for twitch tension to recover from 25% to 75% of control) was longer for cirrhotic patients (35 +/- 14 min) than for control patients (17 +/- 5 min). We observed a longer clinical duration in cirrhotic patients compared to control patients (median values of 62 and 42 min, respectively), although these differences did not attain statistical significance (presumably a function of the small sample size). Although our results differ from those of Khalil et al., findings from both studies suggest that initial recovery after a usual clinical dose of rocuronium (which is presumably a function of distribution rather than elimination) is not affected by liver disease. However, later recovery after usual clinical doses, or recovery after larger initial doses or repeated administration (which are presumably a function of elimination rather than distribution), may be prolonged in patients with liver disease. This is a result of the longer elimination half-life of rocuronium in cirrhotic patients (as well as the longer mean residence time in these patients reported by Khalil et al.).
Our findings regarding onset differ from those of Khalil et al.: we observed similar onset times in both groups whereas Khalil et al. reported that onset was slower in patients with cirrhosis. In addition, their onset times were markedly longer than ours. This latter finding may be explained by differences in anesthetic conditions in the two studies--we administered isoflurane to all subjects whereas they used an opioid-based anesthetic. Because isoflurane potentiates rocuronium relative to opioid-based anesthesia  and the same absolute dose of rocuronium, 0.6 mg/kg, was administered in the two studies, our dose represents a larger multiple of the ED95, and this relatively larger dose should result in a faster onset of paralysis . We are unable to explain why Khalil et al. observed a longer onset time in cirrhotics although we saw no difference between groups.
The pharmacokinetic analyses suggest that the volume of distribution of rocuronium is larger in patients with liver disease than in control patients but that clearance is the same in both groups. The larger volume of distribution may result from an increase in the extracellular fluid volume that occurs in patients with liver disease , particularly those with ascites, as in our subjects. Whether altered protein binding in liver disease also contributes to the larger volume of distribution is unknown. However, for drugs of low protein binding (rocuronium's protein binding is approximately 30% [personal communication, Mitchell Weinberger, PhD, Organon, Inc., 1992]), protein binding minimally influences volume of distribution . Based on previous reports that rocuronium is eliminated predominantly by the liver and is excreted intact in bile, we expected clearance of rocuronium to be smaller in patients with liver disease than in control patients. In contrast, clearance was similar in the two groups. There are two possible explanations for this finding. First, clearance of rocuronium is markedly less than hepatic plasma flow suggesting that rocuronium [like vecuronium ] has a low extraction ratio; this may lessen the effect of liver disease. Second, biliary excretory capacity may be preserved despite extensive hepatocellular disease .
The findings of the present study can be compared with those for vecuronium, another steroidal muscle relaxant that is eliminated by the liver [18-22]. In cirrhotic patients, vecuronium's onset is delayed , its clearance is decreased , and the duration of large doses is prolonged compared to controls [20-22]. The present findings suggest that rocuronium is affected by liver disease differently than vecuronium.
We did not model the relationship between concentration and effect as has been done in previous studies of muscle relaxants . This decision was based on the variability of isoflurane concentrations in the control group. Although we could have determined the concentration-effect relationship in patients with liver disease, the absence of appropriate control data limits the utility of that estimate. In addition, plasma samples from all control patients and approximately half of the patients with liver disease were venous rather than arterial. The only time at which venous and arterial concentrations differ significantly is during onset, the time during which optimal estimates (i.e., arterial concentrations) are needed.
In summary, liver disease increases the volume of distribution of rocuronium and its elimination half-life, but not its clearance. The longer elimination half-life in patients with liver disease should prolong rocuronium's duration of action, particularly with larger initial doses or prolonged administration. However, onset time is not altered by the presence of liver disease, preserving a desirable property of rocuronium. These results suggest that rocuronium may be used cautiously in patients with impaired liver function. Further studies will indicate whether rocuronium's clinical effects are markedly influenced by liver disease.
The authors wish to thank James Arden, MD, for his critical review of the manuscript and substantive comments and also Winifred von Ehrenburg for her editorial expertise and enthusiasm in the preparation of this manuscript.
1. Magorian T, Flannery KB, Miller RD. Comparison of rocuronium, succinylcholine, and vecuronium for rapid-sequence induction of anesthesia in adult patients. Anesthesiology 1993;79:913-8.
2. Khuenl-Brady K, Castagnoli KP, Canfell PC, et al. The neuromuscular blocking effects and pharmacokinetics of ORG 9426 and ORG 9616 in the cat. Anesthesiology 1990;72:669-74.
3. Szenohradsky J, Fisher DM, Segredo V, et al. Pharmacokinetics of rocuronium bromide (ORG 9426) in patients with normal renal function or patients undergoing cadaver renal transplantation. Anesthesiology 1992;77:899-904.
4. Child CG, ed. The liver and portal hypertension, major problems in clinical surgery. Philadelphia: WB Saunders, 1963.
5. Quill TJ, Begin M, Glass PS, et al. Clinical responses to ORG 9426 during isoflurane anesthesia. Anesth Analg 1991;72:203-6.
6. Foldes FF, Nagashima H, Nguyen HD, et al. The neuromuscular effects of ORG 9426 in patients receiving balanced anesthesia. Anesthesiology 1991;75:191-6.
7. Furata T, Canfell PC, Castagnoli KP, et al. Quantitation of pancuronium, 3-desacetylvecuronium, vecuronium, pipecuronium, and 3-desacetylpipecuronium in biological fluids by capillary gas chromatography using nitrogen-sensitive detection. J Chromatogr 1988;427:41-53.
8. Beal SL, Sheiner LB. NONMEM User's guide. San Francisco: University of California, 1979.
9. Bates DM, Watts DG. Nonlinear regression analysis and its applications. New York: Wiley, 1988:200-31.
10. Modern regression methods. S-Plus user's manual, Version 3.0. Seattle: Statistical Sciences, Inc, 1991:1-46.
11. Khalil M, D'Honneur G, Duvaldestin P, et al. Pharmacokinetics and pharmacodynamics of rocuronium in patients with cirrhosis. Anesthesiology 1994;80:1241-7.
12. Shanks CA, Fragen RJ, Ling D. Continuous intravenous infusion of rocuronium (ORG 9426) in patients receiving balanced, enflurane, or isoflurane anesthesia. Anesthesiology 1993;78:649-51.
13. Ginsberg B, Glass PS, Quill T, et al. Onset and duration of neuromuscular blockade following high-dose vecuronium administration. Anesthesiology 1989;71:201-5.
14. Wilkinson GR, Shand DG. A physiological approach to hepatic drug clearance. Clin Pharmacol Ther 1975;18:377-90.
15. Blaschke TF. Protein binding and kinetics of drugs in liver diseases. Clin Pharmacokinet 1977;2:32-44.
16. Goldfarb G, Ganneau P, Ang ET, et al. Hepatic extraction and clearance of vecuronium in humans. Anesthesiology 1988;69:A480.
17. Mol WEM, Fokkema GN, Meijer DKF, et al. Mechanisms for the hepatic uptake of organic cations. Studies with the muscle relaxant vecuronium in isolated rat hepatocytes. J Pharmacol Exp Ther 1988;244:268-75.
18. Bencini AF, Mol WEM, Scaf AHJ, et al. Uptake and excretion of vecuronium bromide and pancuronium bromide in the isolated perfused rat liver. Anesthesiology 1988;69:487-92.
19. Arden JR, Lynam DP, Castagnoli KP, et al. Vecuronium in alcoholic liver disease: a pharmacokinetic and pharmacodynamic analysis. Anesthesiology 1988;68:771-6.
20. Bencini AF, Scaf AHJ, Sohn YJ, et al. Hepatobiliary disposition of vecuronium bromide in man. Br J Anaesth 1986;58:988-95.
21. Lebrault C, Berger JL, D'Hollander AA, et al. Pharmacokinetics and pharmacodynamics of vecuronium (ORG NC 45) in patients with cirrhosis. Anesthesiology 1985;62:601-5.
22. Hunter JM, Parker CJR, Bell CF, et al. The use of different doses of vecuronium in patients with liver dysfunction. Br J Anaesth 1985;57:758-64.
© 1995 International Anesthesia Research Society
23. Sheiner LB, Stanski DR, Vozeh S, et al. Simultaneous modeling of pharmacokinetics and pharmacodynamics: application to d-tubocurarine. Clin Pharmacol Ther 1979;25:358-71.