Do the pharmacokinetics of vecuronium change during prolonged administration in critically ill patients?

British Journal of Anaesthesia 1998; 80: 715–719

CLINICAL INVESTIGATIONS

Do the pharmacokinetics of vecuronium change during prolonged administration in critically ill patients? V. SEGREDO, J. E. CALDWELL, P. M. C. WRIGHT, M. L. SHARMA, L. D. GRUENKE AND R. D. MILLER

Summary Neuromuscular blocking drugs may be administered over several days to patients in the intensive care unit (ICU), but their pharmacokinetics have been studied at only one point in time, or assumed to be constant throughout the period of administration. We sought to determine if, in individual patients, the pharmacokinetics of vecuronium changed over the course of its administration in the ICU. In six critically ill patients, we measured plasma vecuronium concentrations during two periods: first, during initial administration of vecuronium and second, after its administration continuously for 3–6 days. A pharmacokinetic model was fitted to these plasma concentration data, and its parameters permitted to vary between the periods to determine if they had altered. Individual clearance values during the study ranged from 1.4 to 4.4 ml kg91 min91. During prolonged administration, vecuronium clearance increased in three and decreased in two patients. This change ranged from a 61% decrease to a 58% increase, and was not linked to any clinical factor. The steady-state volume of distribution (range 368– 1765 ml kg91; median 494 ml kg91) did not change in any patient during the study. The change in clearance of vecuronium during its prolonged administration in critically ill patients suggests that future studies of neuromuscular blocking drugs in the ICU should take account of their changing pharmacokinetics over the course of administration. (Br. J. Anaesth. 1998; 80: 715–719) Keywords: intensive care vecuronium; neuromuscular block vecuronium; pharmacokinetics vecuronium

The administration of neuromuscular blocking drugs is a common practice in the intensive care unit (ICU), where these drugs may be given for days or weeks.1–3 Studies of the pharmacokinetics of neuromuscular blocking drugs in the ICU have been performed at only a single time during drug administration4 5 or, if the study was conducted over several days, it was assumed that the pharmacokinetics were constant during the period of study.6 Because neuromuscular blocking drugs are often given for several days, during which time the physiological status of the patients may alter, the

pharmacokinetics of the drugs may also change, and an estimate of pharmacokinetics at one point in time may not be accurate at a later time. Vecuronium has been used commonly, and for prolonged periods, in the ICU.4 7 8 In the present study we determined the pharmacokinetics of vecuronium in patients in the ICU, and investigated whether the pharmacokinetics in individual patients changed during prolonged administration of the drug. If the pharmacokinetics of vecuronium change during prolonged administration, the potential for changing pharmacokinetics should be taken into account in future studies of other neuromuscular blocking drugs in the ICU.

Patients and methods The study was approved by the Committee of Human Research of the University of California at San Francisco. After obtaining written informed consent from each patient’s next of kin and primary ICU physician, we consecutively studied six critically ill patients (three males, three females) aged 22–86 yr. Patients receiving assisted ventilation for adult respiratory distress syndrome (ARDS) were screened for possible enrolment in the study if the attending ICU physician decided that neuromuscular block was clinically indicated. Patients with a history of neuromuscular disease, renal failure requiring haemodialysis, or administration of a neuromuscular blocking drug in the 24 h before the start of the study were excluded. PHARMACOKINETIC STUDY

In all patients vecuronium was administered i.v. at a rate of 6 ␮g kg91 min91 until complete neuromuscular block was induced (approximately 10 min). Arterial blood was sampled before and at 2, 4, 6, 8, 10, 15, 30, 45, 60, 90, 120, 150, 180, 210, 240, 300 and 360 min after the beginning of the infusion. VERONICA SEGREDO, MD, JAMES E. CALDWELL, MB CHB, PETER M. C. WRIGHT*, MD, MANOHAR L. SHARMA, PHD, LARRY D. GRUENKE, PHD, RONALD D. MILLER, MD, Department of Anesthesia, University of California, San Francisco, CA 94143– 0648, USA. Accepted for publication January 14, 1998. *Present address: Department of Anaesthetics, Freeman Hospital, Newcastle-upon-Tyne NE7 7DN. Correspondence to J. E. C.

716

British Journal of Anaesthesia

Table 1 Patient characteristics, plasma creatinine, total bilirubin, and liver function grade during initial administration of vecuronium (I) and after stopping its administration (II) in critically ill patients. *Liver function was graded in four classes: 0 : normal liver function tests; 1 : increased plasma aspartate aminotransferase (AST) and alkaline phosphatase (ALK-P) levels without increased bilirubin concentration or prothrombin time; 2 : increased plasma AST, ALK-P and bilirubin concentration without increased prothrombin time; 3 : increased prothrombin time caused by altered liver function Weight (kg)

Creatinine (µmol l91)

Bilirubin (µmol l91)

Liver function*

Patient

Primary pathology

Sex

Age (yr)

I

II

I

II

I

II

I

II

1 2 3 4 5 6

Sepsis Liver failure Unknown Sepsis Pneumonia Sepsis

M F F M M F

58 32 39 86 27 22

58 61 69 64 67 64

56 60 65 67 74 67

88 150 80 159 106 80

115 168 71 248 124 71

0.7 31 0.3 0.6 10.4 20

0.9 41 0.1 0.6 12.5 18.1

1 3 1 0 2 2

1 3 1 0 3 2

Table 2 Reasons for mechanical ventilation and sedative drugs used Patient

Primary pathology

Aetiology of respiratory failure

Sedation and analgesia

1 2 3 4 5 6

Non-Hodgkin’s lymphoma Alcoholic cirrhosis and encephalopathy Postoperative urinary tract infection Repair of abdominal aortic aneurysm Testicular cancer and pancytopenia Malignant histiocytosis and pancytopenia

Pneumocystis pneumonia; systemic candidiasis Aspiration pneumonia Enterococcal sepsis Aspiration pneumonia Sepsis of unknown aetiology Staphylococcal sepsis

Midazolam, lorazepam, morphine Lorazepam, morphine Midazolam, morphine Midazolam, lorazepam, fentanyl Midazolam, morphine Midazolam, pethidine

Subsequently, vecuronium was given by continuous infusion or frequent intermittent boluses according to clinical requirements. This method of administration continued until the patient’s clinical status suggested that vecuronium therapy might be discontinued within the next 12 h. At this time, in all patients, vecuronium administration was changed to a continuous infusion at an hourly rate equal to the average hourly requirement over the preceding 24 h. This infusion was continued unchanged for 12 h, then discontinued. Three arterial blood samples were drawn during the 60 min just before discontinuing the vecuronium infusion and at intervals for up to 8 h after stopping the infusion. All samples were placed in ice and processed within 1 h. Plasma concentrations of vecuronium were determined by a capillary gas chromatographic assay specific for vecuronium, with a coefficient of variation of 15% at concentrations of 15 ng ml91 and a sensitivity of 10 ng ml91.9 If, after withdrawal of vecuronium, the patients did not tolerate mechanical ventilation, atracurium was given during the period of blood sampling because it does not interfere with the vecuronium assay. Data from each subject were analysed separately but all data from each subject were analysed together using nonlinear mixed-effects modelling (NONMEM).10 Subjects were assumed to have

reached a steady state before discontinuation of the final infusion. Basic parameters of the models were clearance (Cl), intercompartmental clearances (Q1 and Q2), and volumes of distribution V1, V2 and V3. Volume of distribution at steady state (Vss) was equal to V1 ; V2 ; V3. First two- and three-compartment mamillary models were fitted to the data from each subject. The three-compartment model was used subsequently only if it could be statistically justified. Next, additional (change) parameters were introduced, each permitting one basic parameter (Cl, V1 etc.) to vary from the first to the second infusion. These change parameters were accepted into the model only if they statistically and visually improved the fit of the model to the data. When all justified parameters had been introduced, the necessity for each one was tested by removing it and obtaining fresh parameter estimates. The necessity for a change parameter was used to evaluate whether or not a basic parameter had altered, and the magnitude of the change in the basic parameter was determined from the value of the change parameter. Statistical justification for the use of a threecompartment rather than a two-compartment model and for inclusion of a new parameter in the model was a decrease of 9.2 in the NONMEM objective function (P : 0.01).10

Table 3 Details of NONMEM pharmacokinetic analysis. *In the basic model, volumes of distribution and clearance are assumed to be the same throughout the administration of vecuronium. In the other models, volumes of distribution, or clearance or both are allowed to vary between the time of initial administration and the end of administration of vecuronium, several days later. A decrease in the objective function of 9.2 denotes a significant improvement over a simpler model (P:0.01). Bold type: values for the optimal model Patient

1

2

3

4

5

6

Vecuronium administration Days of administration Total dose (mg)

5 331

4 115

6 342

4 180

3 135

4 212

Value of objective function* Basic model Volume allowed to vary Clearance allowed to vary Volume and clearance allowed to vary

347.7 339.0 302.2 299.6

277.5 263.3 231.8 231.0

161.0 161.0 151.7 151.2

216.1 211.0 195.8 191.1

261.6 261.2 246.3 245.9

219.8 219.7 218.6 N/A

Vecuronium pharmacokinetics in intensive care

717 Table 4 Individual pharmacokinetic values during initial administration of vecuronium (I) and after its administration for several days (II) in critically ill patients. *Mean residence time was calculated as volume of distribution/plasma clearance Plasma clearance (ml kg91 min91)

Mean residence time* (min)

Patient

Steady-state volume of distribution (ml kg91)

I

II

I

II

1 2 3 4 5 6 Median

436 836 1765 368 496 492 494

3.6 1.9 3.0 2.5 1.8 3.0 2.8

1.4 3.0 4.4 2.1 2.2 3.0 2.6

123 452 586 146 273 166 219

304 277 397 174 224 166 251

rubin, aspartate aminotransferase (AST) and alkaline phosphatase (ALK-P) were measured. If the results suggested an abnormality, prothrombin time was also measured. Liver function was graded in four classes: 0 : normal liver function tests; 1 : increased plasma AST and ALK-P levels without increased bilirubin concentration or prothrombin time; 2 : increased plasma AST, ALK-P and bilirubin concentrations without increased prothrombin time; and 3 : increased prothrombin time because of altered liver function. The administration of vasoactive drugs, and of any drugs that might alter the pharmacology of vecuronium, was recorded.

Results CLINICAL DATA

Figure 1 Plasma concentration vs time curves for patient 2. Symbols represent actual plasma concentrations. The initial and final doses are separated by a period of 4 days (represented by //). (A) Fitted curve with clearance and volumes of distribution assumed to be constant throughout the period of vecuronium administration. (B) Fitted curve with clearance assumed to be constant and volumes of distribution allowed to vary between the initial and final administration vecuronium; there is no visual or statistical improvement over previous fit. (C) Fitted curve with volumes of distribution assumed to be constant and clearance allowed to vary between the initial and final administration vecuronium; this final fit is visually and statistically (P : 0.01) superior to the previous two fits.

CLINICAL DATA

Renal and liver function tests were performed for all patients on the first and final days of the study. Renal function was assessed using plasma creatinine concentrations. Plasma concentrations of total bili-

Patient characteristics, renal and hepatic function are detailed in table 1. For all but one patient (patient 4), plasma creatinine concentration changed less than 30 ␮mol l91 during the study period; in patient 4, the plasma creatinine concentration increased from 159 to 248 ␮mol l91. Liver function changed little in any patient during the study, and only for patient 5 did the grade change. In this patient, prothrombin time increased from a normal value of 13.1 s to a marginally elevated value of 14.2 s, resulting in his score changing from 2 to 3. In all patients, mechanical ventilation was instituted for respiratory failure. Vecuronium was administered to enable the patients to tolerate mechanical ventilation, after appropriate sedation had been given. The patients’ lungs were ventilated to achieve normocapnia. The patients’ underlying pathology and the aetiology of their respiratory failure are shown in table 2. Only one patient (patient 2) required inotropic support, and this was during the late phase only, using dopamine at an average rate of 6 ␮g kg91 min91; patients 4 and 6 received dopamine in renal doses of 1.6–3.2 ␮g kg91 min91. All patients except patient 2 received an aminoglycoside antibiotic during the course of vecuronium administration. Patients 1, 4 and 5 received frusemide during the early, but not the late, phase of the study. Patients 1, 5 and 6 received hydrocortisone throughout the study. Patient 6 received phenytoin throughout the study.

718 PHARMACOKINETIC STUDY

For all patients, the three-compartment model was superior. Allowing the Vss to vary between early and late phases did not improve the fit in any patient (table 3, figure 1). In contrast, for five of the six patients, there was a significant improvement in fit when Cl was allowed to change from early to late phases (table 3, figure 1). Allowing the intercompartmental clearances to vary from early to late phases did not improve the fit in any patient. The values of the pharmacokinetic variables are given in table 4. Vss was calculated by summing the volumes of each of the three compartments, and mean residence time was calculated as Vss/Cl. Vss (range 368–1765 ml kg91; median 494 ml kg91) did not change in any patient during the course of vecuronium administration. In contrast, the clearance of vecuronium (range 1.4–4.4 ml kg91 min91) increased in three and decreased in two patients. Only one patient (patient 4) had a significant deterioration in renal function, and this was accompanied by a small decrease in clearance from 2.5 to 2.1 ml kg91 min91. In patient 5 liver function deteriorated slightly, but clearance increased from 1.8 to 2.2 ml kg91 min91.

Discussion We found that the clearance of vecuronium changed in five of six patients over the course of vecuronium administration. There was no consistent pattern to this change; clearance increased in three patients and decreased in two. There are several potential reasons why clearance changed; among them are alterations in organ function, and the administration of drugs that might influence the biodisposition of vecuronium. The largest change in clearance occurred in patient 1, in whom it decreased from 3.6 to 1.4 ml kg91 min91. This was accompanied by an increase in creatinine from 88 to 115 ␮mol l91 and no change in liver function. Even if renal function had deteriorated to a greater extent than was reflected by the relatively small change in creatinine, we consider it unlikely that the deterioration in renal function produced the decrease in clearance, as even in anephric patients vecuronium clearance may not change.11 The haemodynamic status of this patient actually improved during the study (mean arterial pressure of 86 and 117 mm Hg during early and late phases respectively), without the administration of vasoactive drugs. We could identify no other factors in this patient’s course that would account for the decrease in clearance of vecuronium. Because vecuronium is eliminated by both kidney and liver we would have expected changes in function of these organs to have been reflected in changes in clearance.12 13 However, in patient 4, who had significant deterioration in renal function, clearance decreased only slightly (2.5 to 2.1 ml kg91 min91), and in patient 5 whose liver function deteriorated slightly, clearance actually increased from 1.8 to 2.2 ml kg91 min91. The reason that we saw little relationship between organ function and clearance may be that the changes in organ function were relatively small and, additionally, other factors were influencing

British Journal of Anaesthesia clearance, such as the administration of diuretics. Patients 1, 4 and 5 received frusemide during the early, but not the late, phase of the study and we could speculate that the resultant diuresis may have artificially increased vecuronium clearance in the early phase. The only data on the effect of an induced diuresis on the clearance of a neuromuscular blocking drug relate to mannitol, which had no influence on the clearance of d-tubocurarine.14 In the absence of data, we cannot draw any conclusions about the possible interaction between frusemide and vecuronium clearance. Patient 6 was receiving phenytoin, which significantly alters the neuromuscular response to vecuronium.15 Vecuronium clearance in this patient did not change during the study, however, possibly because she had been receiving long-term phenytoin and any effect of the drug on vecuronium clearance would have been present at the start of the study. Patients were receiving a variety of other drugs, for example aminoglycoside antibiotics and hydrocortisone, that interact with vecuronium and alter its pharmacodynamics, but have no known effect on its pharmacokinetics. Because patients in the ICU receive many drugs, interactions might occur that affect the pharmacokinetics of vecuronium, but as there are no data available we cannot speculate further on such possible interactions. During the pharmacokinetic analysis, we assumed linearity, namely that plasma concentrations of vecuronium were proportional to the dose administered and that clearance would not vary with vecuronium concentration. This basic assumption underlies the pharmacokinetic modelling of nondepolarizing neuromuscular blocking drugs, and is supported by experimental evidence.16 In a carefully controlled study, Wright and colleagues demonstrated that the pharmacokinetics of vecuronium were linear over an almost three-fold dose range.16 In our study, the plasma concentrations of vecuronium were within the range of those studied by Wright and colleagues,16 but our experimental conditions were different. Wright and colleagues studied healthy volunteers and vecuronium was administered for only brief periods; our subjects were critically ill and vecuronium was administered over several days. Thus, although we consider it unlikely that nonlinearity developed over the course of our study, we cannot rule out the possibility. As vecuronium clearance could decrease or increase with time in an individual patient, the mean values for the patients as a group would not have differed between the early and late phases of the study. For this reason, we estimated the clearance and its change over time for subjects as individuals, and not as a group. Because patients in the ICU are so heterogeneous with regard to their pathophysiological state, expressing their characteristics as a mean value for a group is potentially misleading; in this study it would have suggested a stability in clearance which did not exist. We found no change in the Vss of vecuronium during the study, but the interindividual values (368– 1765 ml kg91) varied over a four-fold range; the mean value was 732 and the standard deviation 531 ml kg91, so the coefficient of variation was approximately 70%. These values for mean and coefficient of varia-

Vecuronium pharmacokinetics in intensive care tion are similar to those from an earlier vecuronium ICU study (1100 ml kg91, 55%).4 This apparent large variability in Vss in patients in the ICU probably results from the heterogeneous nature of these patients, with regard in particular to their underlying disease processes and level of organ function. As we studied only six patients, we cannot preclude the possibility that, in other patients, the Vss for vecuronium might have changed over the course of the drug’s administration. The mean value for Vss, 732 ml kg91, is greater than the range (180–510 ml kg91) reported after acute administration in the operating theatre in patients with normal or with impaired renal or hepatic function.17 This apparently large estimate for Vss probably is a result of the chronic administration of vecuronium, and consequently revelation of its distribution into a large peripheral compartment with low blood flow. This finding is consistent with the results of Matteo and colleagues, who, because they measured plasma concentrations for 96 h, estimated the volume of distribution of d-tubocurarine as 3.4 1 kg91.14 This hypothesis of a large peripheral compartment with low blood flow is supported by the work of Waser and colleagues.18 Using whole-body autoradiography in rats, they demonstrated that vecuronium is distributed widely, including into tissues with low blood flow, such as connective tissue. After acute administration of vecuronium, and with sampling periods of 6 h, the amount of drug in this peripheral compartment would be minimal, would not be revealed in the plasma concentration vs time curve, and hence would not influence the calculation of pharmacokinetic variables. However, with continued administration, progressively greater amounts of vecuronium would accumulate in the slowly equilibrating peripheral compartment. After discontinuation of vecuronium, distribution of the drug from this slow peripheral compartment into the plasma would significantly influence the decay profile of the plasma concentration vs time curve, and the larger Vss would be revealed. Normal clinical practice in our ICU dictated the regimen for administration of vecuronium over the course of the study. This resulted in continuous infusion, or regular and frequent bolus doses of vecuronium. The dose of vecuronium was increased until the patients tolerated mechanical ventilation, and remained unchanged thereafter. This regimen allowed us to model the pharmacokinetics of vecuronium using the assumption of steady-state, during the period between the early and late phases of the study.10 Vecuronium has a 3-desacetyl metabolite that has significant neuromuscular blocking potency.19 It was not the aim of this study to investigate the metabolite, but rather to determine whether the pharmacokinetics of the parent compound, vecuronium, changed with time. Because of our results with vecuronium, it is a reasonable assumption that the pharmacokinetics of its metabolite also change with time.

Acknowledgment This study was supported, in part, by NIH grant ROI GM 26403– 0648.

719

References 1. Bion JF, Bowden MI, Chow B, Honisberger L, Weatherley BC. Atracurium infusions in patients with fulminant hepatic failure awaiting liver transplantation. Intensive Care Medicine 1993;19:S94–S98. 2. Hansen-Flaschen JH, Brazinsky S, Basile C, Lanken PN. Use of sedating drugs and neuromuscular blocking agents in patients requiring mechanical ventilation for respiratory failure. A national survey. Journal of the American Medical Association 1991; 266: 2870–2875. 3. Klessig HT, Geiger HJ, Murray MJ, Coursin DB. A national survey on the practice patterns of anesthesiologist intensivists in the use of muscle relaxants. Critical Care Medicine 1992; 20: 1341–1345. 4. Segredo V, Caldwell JE, Matthay MA, Sharma ML, Gruenke LD, Miller RD. Persistent paralysis in critically ill patients after long-term administration of vecuronium. New England Journal of Medicine 1992; 327: 524–528. 5. Boyd AH, Eastwood NB, Parker CJ, Hunter JM. Comparison of the pharmacodynamics and pharmacokinetics of an infusion of cis-atracurium (51W89) or atracurium in critically ill patients undergoing mechanical ventilation in an intensive therapy unit. British Journal of Anaesthesia 1996; 76: 382–388. 6. Sparr HJ, Wierda JMKH, Proost JH, Keller C, Khuenl-Brady KS. Pharmacodynamics and pharmacokinetics of rocuronium in intensive care patients. British Journal of Anaesthesia 1997; 78: 267–273. 7. Smith CL, Hunter JM, Jones RS. Vecuronium infusions in patients with renal failure in an ITU. Anaesthesia 1987; 42: 387–93. 8. Prielipp RC, Coursin DB, Scuderi PE, Bowton DL, Ford SR, Cardenas VJ jr, Vender J, Howard D, Casale EJ, Murray MJ. Comparison of the infusion requirements and recovery profiles of vecuronium and cisatracurium 51W89 in intensive care unit patients. Anesthesia and Analgesia 1995; 81: 3–12. 9. Furuta T, Canfell PC, Castagnoli KP, Sharma ML, Miller RD. Quantitation of pancuronium, 3-desacetylpancuronium, vecuronium, 3-desacetylvecuronium, pipecuronium and 3-desacetylpipecuronium in biological fluids by capillary gas chromatography using nitrogen-sensitive detection. Journal of Chromatography 1988; 427: 41–53. 10. Beal SL, Sheiner LB. NONMEM User’s Guide. San Francisco, University of California, 1979. 11. Meistelman C, Lienhart A, Leveque C, Bitker MO, Pigot B, Viars P. Pharmacology of vecuronium in patients with end-stage renal failure. European Journal of Anaesthesiology 1986; 3: 153–158. 12. Bencini AF, Scaf AH, Sohn YJ, Meistelman C, Lienhart A, Kersten UW, Schwarz S, Agoston S. Disposition and urinary excretion of vecuronium bromide in anesthetized patients with normal renal function or renal failure. Anesthesia and Analgesia 1986; 65: 245–251. 13. Bencini AF, Scaf AHJ, Sohn YJ, Uw K-K, Agoston S. Hepatobiliary disposition of vecuronium bromide in man. British Journal of Anaesthesia 1986; 58: 988–995. 14. Matteo RS, Nishitateno K, Pua EK, Spector S. Pharmacokinetics of D-tubocurarine in man: effect of an osmotic diuretic on urinary excretion. Anesthesiology 1980; 52: 335–338. 15. Platt PR, Thackray NM. Phenytoin-induced resistance to vecuronium. Anaesthesia and Intensive Care 1993; 21: 185– 191. 16. Wright PM, Hart P, Lau M, Sharma ML, Gruenke L, Fisher DM. Cumulative characteristics of atracurium and vecuronium. A simultaneous clinical and pharmacokinetic study. Anesthesiology 1994; 81: 59–68. 17. Miller RD, Savarese JJ, Lien CA, Caldwell JE. Pharmacology of muscle relaxants and their antagonists. In: Miller RD, eds. Anesthesia. New York: Churchill Livingstone, 1994. 18. Waser PG, Wiederkehr H, Sin RA, Kaiser SE. Distribution and kinetics of 14C-vecuronium in rats and mice. British Journal of Anaesthesia 1987; 59: 1044–1051. 19. Caldwell JE, Szenohradszky J, Segredo V, Wright PMC, McLoughlin C, Sharma ML, Gruenke LD, Fisher DM, Miller RD. The pharmacodynamics and pharmacokinetics of the metabolite 3-desacetylvecuronium (Org 7268) and its parent compound, vecuronium, in human volunteers. Journal of Pharmacology and Experimental Therapeutics 1994; 270: 1216– 1222.