Pharmacokinetics and Pharmacodynamics of Mivacurium in Patients Phenotypically Homozygous for the Atypical Plasma Cholinesterase Variant: Effect of Injection of Human Cholinesterase

Eleven patients with American Society of Anesthesiologists physical status I or II and phenotypically homozygous occurrence of the atypical gene who were scheduled to undergo elective surgery were enrolled in the study during a period of 5 yr. All patients had previously been issued warning cards by the Danish Cholinesterase Research Unit.10Patients with a history of neuromuscular, cardiovascular, renal, or hepatic disorders were excluded from the study, as were females of childbearing potential and patients receiving drugs that might affect neuromuscular transmission. The patients gave informed consent, and the Ethics Committee in Copenhagen County, Denmark, approved the study.

Premedication consisted of 0.1–0.2 mg/kg oral diazepam. The actual pChe activity was determined from a blood sample taken before induction of anesthesia11(table 1). Anesthesia was induced with 1–4 μg/kg fentanyl, 0.05–0.15 mg/kg diazepam, and 3–6 mg/kg thiopentone and maintained with 66% nitrous oxide in oxygen and supplementary doses of fentanyl, and in a few patients, with 5–10 mg · kg⁻¹· h⁻¹propofol.

During surgery, the patients were monitored continuously using electrocardiography, pulse oximetry, and capnography. Blood pressure was measured every minute for the first 3 min after administration of mivacurium and every 5 min thereafter. The blood pressure cuff and the intravenous line used for administration of anesthetics and fluid were on the same arm. Ventilation was adjusted to maintain normocapnia (end-tidal carbon dioxide pressure 34–42 mmHg). The rectal and peripheral skin temperatures were measured and maintained above 35° and 32°C, respectively.12After induction of anesthesia, a second intravenous catheter was inserted for blood sampling in the arm used for monitoring.

At the start of the study, neither the duration of action of normal tracheal intubation doses of mivacurium nor the effectiveness of cholinesterase were known in these patients. Therefore, the first patient received 0.075 mg/kg, and the remaining 10 patients (patients 2–11) were given 0.15 mg/kg mivacurium, which represents a clinically relevant dose for induction of anesthesia.

To evaluate the effect of mivacurium after the injection of cholinesterase and after recovery of neuromuscular block following the first dose of mivacurium, one patient (patient 4) received two supplementary doses of 0.03 mg/kg mivacurium, and another patient (patient 5) received 0.03 and 0.06 mg/kg mivacurium.

The mechanical twitch was recorded using a Myograph 2000 (Biometer International, Odense, Denmark). The ulnar nerve was stimulated at the wrist using surface electrodes and 1-Hz single-twitch stimulation. After supramaximal stimulation was achieved, the stimulation pattern was changed to TOF stimulation every 12 s. When the response to stimulation was stable for 5 min, a single bolus dose of mivacurium was given over 30 s.

Tracheal intubation was performed at maximum T¹suppression, where T¹is the first response in the TOF response.

Onset (time from beginning of injection of mivacurium to 95% T¹suppression) and recovery data were determined using start control values, and monitoring was continued at least until 90% T¹recovery and a TOF ratio of 0.80 were obtained.12The time to first response to TOF stimulation and the durations to 10, 25, 50, and 90% T¹recovery and a TOF ratio of 0.80 were determined. The interval from 25% to 75% twitch height recovery was also calculated. In patients given neostigmine for reversal of neuromuscular block (see Injection of Neostigmine), the times from administration of neostigmine to 90% T¹recovery and to a TOF ratio of 0.80 were measured.12 

We used serum cholinesterase, Centeon® (Centeon Pharma, Marburg, Germany), a highly purified concentrate of the enzyme. The content of each vial, 55 mg (27–83 mg), is equivalent in activity to 500 ml fresh human plasma.

Six patients with expected short-lasting surgeries received cholinesterase 30–36 min after the injection of mivacurium (patients 1–6), and five patients with expected long-lasting surgeries received cholinesterase after 82–143 min (patients 7–11).

In the first seven patients given cholinesterase, a dose of 4.0–5.5 mg/kg was administered (patients 1, 2, 3, 7, 8, 9, and 10; table 1), doses that increased the effective enzyme activity to within normal levels in all patients. To evaluate whether an even higher dose of cholinesterase would further increase the metabolic rate of mivacurium and thus shorten the recovery, four patients (patients 4, 5, 6, and 11; table 1) received 6.6–10.0 mg/kg cholinesterase.

Plasma cholinesterase activity was measured using benzoylcholine as a substrate, and the previously found phenotype was confirmed using several inhibitors.11However, only the results regarding the dibucaine number are given.

Blood samples for measurement of pChe activity were collected immediately before induction of anesthesia, before and 5–15 min after administration of human cholinesterase, and before and 5 min after administration of neostigmine. In four patients, samples for measurement of pChe were also collected 30 and 60 min after administration of human cholinesterase. The amount of active (or efficient) enzyme (pChe^efficient) in blood was calculated as: pChe^efficient= pChe^after− pChe^before, where pChe^afteris the activity measured after the injection of human cholinesterase, and pChe^beforeis the activity before the injection.

In principle, neuromuscular block was left to recover “spontaneously” after injection of cholinesterase (patients 1, 3, 4, 6, 9, and 10). However, when surgery ended before sufficient recovery of neuromuscular function after injection of cholinesterase was achieved, 0.04 mg/kg neostigmine preceded by atropine was given (patients 2, 5, 7, 8, and 11).

Venous blood samples (5 ml) were collected immediately before administration of mivacurium; 1, 2, 5, 10, 15, 20, 30, and 60 min after the start of injection; and 1, 2, 5, 10, 15, 20, 30, 60, and 120 min after injection of human cholinesterase.13In less than 10 s, the blood was transferred into a vacutainer containing a cholinesterase inhibitor (phospholine iodide). The samples were centrifuged, and the plasma was separated and frozen at −70°C. The ratio of cis-cis, cis-trans, and trans-trans isomers in the clinical trial material used were approximately 5.8, 35.5, and 58.8%, respectively (data from certificate of analysis, GlaxoSmithKline [formerly GlaxoWellcome, Beckenham, United Kingdom]). The concentration of each isomer of mivacurium was determined by a stereospecific high-performance liquid chromatographic method with fluorometric detection and a stepped gradient. The coefficient of variation was less than 10% at all concentrations except for the lowest level of quantification (5 ng/ml) (14%). Extraction efficiency was 75%. Calibration was linear over the range 5–1,000 ng/ml.13The relative error was less than 6% for the isomers.

A two-compartment model was fitted to the plasma concentrations of each isomer separately. The fitting was performed using numerical integration of differential equations and using the inverse square of the calculated plasma concentration of mivacurium as a weight factor.14After the administration of cholinesterase, an extra k′^ewas numerically added to the elimination rate constant. The parameters V¹, k^e, k¹², k²¹, and k′^ewere fitted to the data, and the fitting for the elimination rate constant K was as follows: before cholinesterase K = k^e; after cholinesterase K = k^e+ k′^e. The extra k′^ewas removed at the time of the administration of neostigmine. The goodness of fit was evaluated by visual inspection and from plots of residuals. The following secondary parameters were calculated before and after administration of cholinesterase: clearance, CL = V¹* K; volume of distribution at steady state, V^ss= V¹* (k²¹+ k¹²)/k²¹; and the terminal rate constant λ²in the two-compartment model was calculated from the primary parameters as

The clearance of each isomer was calculated from the time of injection of mivacurium until the administration of cholinesterase. The clearance after the injection of human cholinesterase was calculated from the time of the injection until infinity.

Pharmacokinetic–pharmacodynamic modeling was performed in two steps. The first step was fitting the described pharmacokinetic model to concentrations calculated as concentration of trans-trans isomer plus 0.9 concentration of the cis-trans isomer. The contribution of the less potent cis-cis isomer was ignored. Second, a link model between the effect and the central compartment concentrations was fitted to the effect data using fixed parameters for the pharmacokinetic part of the model. The sigmoid Emax model was fitted to the effect compartment concentrations: E = (E^max· C^γe)/EC^γ50+ C^γe, where E is the intensity of pharmacologic effect (percent depression of T¹), Emax is the maximum effect, C^eis the concentration in the hypothetical effect compartment in the k^eomodel described by Holford and Sheiner,15EC⁵⁰is the plasma concentration at 50% T¹depression, and γ (the Hill coefficient) is the parameter determining the sigmoidicity of the relation between concentration and effect. Fitting and evaluation of the goodness of the fit were performed as described in the Pharmacokinetics section.

All pharmacokinetic and pharmacodynamic results are presented as median and range. Data before and after administration of cholinesterase were compared using the nonparametric Mann–Whitney U test. P < 0.05 was considered significant.

Two males and nine females were included in the study. The median age was 48 yr (range, 15–73 yr). All patients except one were within 20% of their ideal body weight (patient 3 was 30% above the ideal body weight), body weight being 65 kg (55–100 kg) and height being 167 cm (156–176 cm).

All patients had dibucaine numbers below 30 at baseline (table 1). After injection of human cholinesterase, pChe activity increased within 2–10 min in a dose-dependent manner.

Median onset times after 0.075 and 0.15 mg/kg mivacurium were 1.5 and 2.0 min (range, 1.0–2.3 min), respectively. The neuromuscular block was very deep at the time of injection of cholinesterase as indicated by no reaction to TOF.

Table 2shows times the recovery data after injection of cholinesterase. In all patients, the administration speeded up recovery. No significant relation was found between the dose of injected cholinesterase and the rate of recovery. For example, the median times to reappearance of T¹after the low and high doses of cholinesterase were 12.8 min (3.7–35.5 min) and 18.2 min (13.3–44.2 min), respectively (not significant).

With respect to the timing of the administration of cholinesterase, the data indicate a faster recovery of neuromuscular function after the late injection (patients 7–11) as compared with the early injection (patients 2–6). The median times to reappearance of T¹and to T¹10%, respectively, were 10.2 min (3.7–14.7 min) and 13.2 min (8.8–19.6 min) after the late injection, as compared with 21.7 min (11.7–44.2 min; not significant) and 27.3 min (18.1–56.7 min; P < 0.05) after the early injection of cholinesterase.

Table 3shows recovery data following injection of neostigmine administered after cholinesterase. The injection of neostigmine caused a fast recovery in three of four patients. The 25–75% interval was 7.5 min (4.0–9.3 min), which was shorter than in patients who did not receive neostigmine 15.1 min (7.4–19.6 min). In one patient (patient 2), recoveries after injection of cholinesterase and after neostigmine were both very slow (table 3). One patient (patient 5) received neostigmine after the second supplementary dose of mivacurium at a T¹of 42%, and a TOF ratio of 0.8 was obtained in 6 min.

Recovery data following injection of the two supplementary doses of mivacurium after administration of cholinesterase (patients 4 and 5) are presented in table 4.

Figure 1shows the individual plasma concentrations of each of the three isomers over time after late injection of cholinesterase and neostigmine in one of the patients. The pharmacokinetic data for the three isomers are summarized in table 5. The clearances of all isomers and the elimination rate constants were decreased as compared with phenotypically normal patients indicating longer elimination half-lives (data from a previous study in phenotypically normal patients is presented in table 5for comparison).16The clearances and the rate constants of the active isomers were increased more than the cis-cis isomer. The volume of distribution was small for all isomers as in phenotypically normal patients.

Immediately after injection of cholinesterase, the concentration of the active isomers decreased rapidly, whereas the concentration of the cis-cis isomer was less affected (fig. 1). After administration of neostigmine, the concentration of the active isomers increased moderately (fig. 1).

The clearances and the elimination rate constants of all three isomers increased by a factor of approximately 10. The half-lives decreased correspondingly.

Figure 2illustrates the relation between the estimated effect compartment concentration and time. The observed plasma concentration decreases rapidly after injection of cholinesterase. Despite this, recovery from the neuromuscular block was slow due to slow elimination from the effect compartment.

To evaluate whether the concentration–effect relation differed from that of phenotypically normal patients, curve fitting of the data describing the concentration–effect model was performed and compared with data previously obtained in phenotypically normal patients. The pharmacokinetic–pharmacodynamic data from this latter group is derived from data presented in Østergaard et al.  16Also, estimations of the concentrations of the combined active isomers correlating to 50% T¹depression were performed (EC⁵⁰). Table 6shows the fitted parameters in the Emax model for the combined active isomers. No significant differences in k^eoor t^1/2keowere seen in phenotypically normal and phenotypically homozygous atypical patients for the combined effect of the two active isomers.

Our main findings are that the clearances of mivacurium were significantly reduced and the elimination half-lives of the three isomers of mivacurium were prolonged in patients phenotypically homozygous for the atypical enzyme. This caused prolonged duration of action and recovery time of mivacurium. Injection of cholinesterase caused a normal or even high effective pChe activity, increased the clearance significantly, shortened the elimination half-lives, and enhanced recovery from neuromuscular block. However, in some patients, neuromuscular block recovered slowly. Injection of neostigmine increased the rate of recovery in the majority of patients. No difference in concentration–effect relation was found compared with phenotypically normal patients.

After administration of cholinesterase, pChe activity increased significantly in all patients, and in all but one, the effective pChe activity (= pChe^after− pChe^before) obtained was within the normal range used by the Danish Cholinesterase Research Unit.11The increase in pChe activity seen in our study is consistent with the findings of Naguib et al.  17 

As expected, the duration of action of mivacurium was prolonged. T¹had not reappeared in any of the patients at the time when cholinesterase was given 120 min after the administration of mivacurium. This is in accordance with spontaneous recovery data in a patient homozygous for the silent enzyme, in whom the time to recovery of T¹was 274 min.18 

As expected, injection of cholinesterase increased the rate of recovery. However, no relation was found between the amount of injected cholinesterase and the rate of recovery of block, most probably because all patients obtained effective pChe activities above 550 U/l. Enzyme activities above this level do not seem to speed up the hydrolysis of either mivacurium or succinylcholine in phenotypically normal patients.16,19However, the number of patients included in the study was small, and that the three patients receiving the highest dose of cholinesterase received cholinesterase early after administration of mivacurium.

Judging from the pharmacokinetic data, a faster recovery from the late injection was to be expected (see Pharmacokinetics section). The time to T¹10% was significantly shorter after the late injection of cholinesterase, but no significant difference was seen in time to reappearance of T¹, probably because of the small number of patients and the large variation in time to reappearance of T¹(table 2).

There was a considerable delay from the injection of cholinesterase to the reappearance of T¹, especially after the early injection. This might be explained by the pharmacokinetics of mivacurium in those patients. When a mivacurium-induced block recovers spontaneously in phenotypically normal patients, mivacurium has been eliminated from plasma and tissues due to the rapid hydrolysis by pChe. In phenotypically homozygous atypical patients, the drug is not hydrolyzed in plasma but distributed to the different body compartments. At the time of the administration of cholinesterase, the amount of mivacurium in plasma represents represent only a small fraction of the total amount in the body. It takes some time before mivacurium, in other compartments including the neuromuscular endplate, is hydrolyzed—either after redistribution into plasma or by injected cholinesterase diffused out of plasma to the tissues. This might explain not only the initial slow recovery seen in this study after cholinesterase, but also that there is a difference in recovery time, depending on the time of injection of cholinesterase. This is illustrated in figure 2with data derived from one of our patients. Thirty minutes after administration of mivacurium, the calculated concentration of mivacurium in the effect compartment (C^e) is significantly higher than 80 or 120 min later (upper curve). The effect of cholinesterase injected 30 and 80 min after mivacurium on C^eis illustrated (two lower curves). The time to minimal effective concentration (100 μg/l) is longer after an early injection of cholinesterase compared with a late injection.

Injection of neostigmine after administration of cholinesterase caused a fast recovery in the majority of patients with a shorter 25–75% interval. This decrease in 25–75% interval is in accordance with findings in phenotypically normal patients.20Several case reports have described unsuccessful administration of neostigmine at intense levels of block in phenotypically homozygous atypical patients,5,7whereas neostigmine seems to be effective after cholinesterase.8,9Our results document that neostigmine given after injection of cholinesterase at less intensive block is effective in increasing the rate of recovery. This is in accordance with the findings in phenotypically normal patients, in whom administration of neostigmine has different effects depending on the level of block. At intense block, neostigmine does not speed up recovery, a dual effect is seen when injected at 1–5% T¹, and recovery is increased when injected at greater than 10% T¹.21–23 

In the two patients receiving supplementary doses of mivacurium after administration of cholinesterase, recovery of T¹was almost identical after the first and second doses of 0.03 mg/kg mivacurium, as well as after the 0.03 and 0.06 mg/kg doses. This is in accordance with the findings in phenotypically normal patients. However, the recovery time was longer, and the 25–75% interval was twice that seen in phenotypically normal patients.

As expected, spontaneous recovery from 0.15 mg/kg mivacurium is prolonged in patients phenotypically homozygous for the atypical variant, but recovery can be facilitated considerably by an injection of cholinesterase. The timing of the administration of cholinesterase is not crucial. Cholinesterase can be given both early and late after mivacurium. However, it often may take some time before the effect is seen, and the time to sufficient recovery (TOF ratio = 0.8) may be 30–60 min. Finally, in most patients with slow recovery of block after injection of cholinesterase, neostigmine speeds up recovery when injected at greater than 8% T¹. In some patients, however, the rate of recovery may still be slow.

The clearances of the two active isomers (cis-trans and trans-trans) were lower than in phenotypically normal patients with normal or drug-induced low pChe activity16or low activity due to renal or hepatic failure,24–26whereas the clearance of the cis-cis isomer was less influenced.2,17,24,25Most probably, the clearance of the cis-cis isomer is predominantly renal, because a 50% reduction in clearance is seen in patients with end-stage renal failure.24 

The elimination half-lives of the active isomers were significantly longer than in patients with normal pChe and also longer than in patients with drug-induced low pChe or disease.16,24–26Only data in patients receiving human cholinesterase late were used because valid data cannot be obtained with a sampling period of 30 min.

The initial volume of distribution in phenotypically homozygous patients was comparable to our findings in phenotypically normal patients using the same methodology and higher than that reported by Laurin et al. , who used arterial sampling.16,27 

Administration of cholinesterase caused an immediate decrease in plasma concentrations of all three isomers. After neostigmine, there was a slight increase in the concentration of the active isomers. This is consistent with previous findings and is most probably due the inhibiting effect of neostigmine on pChe and the diffusion of mivacurium back to plasma.28 

The postcholinesterase clearance of the cis-trans isomer was higher than the clearance of the trans-trans isomer. This is in accordance with data from phenotypically normal patients.16,24,25This has also been observed in vitro .29The clearances of the cis-trans and trans-trans isomers were approximately 15 and 9 times higher, respectively, than before administration of cholinesterase and comparable to clearances in phenotypically normal patients previously studied using the same technique.16The clearances are slightly lower than after both short- and a long-lasting infusions of mivacurium in phenotypically normal patients.3,24,25The clearance of the cis-cis isomer increased 12-fold after injection of cholinesterase. This finding was unexpected because clearance of the cis-cis isomer is predominantly renal.24The resulting clearance is higher than reported earlier in phenotypically normal patients, where values from 3.8 to 5.4 ml · kg⁻¹· min⁻¹have been found.3,16,24,25 

The elimination rate constants of the active isomers increased 9- to 15-fold after administration of cholinesterase. The postcholinesterase values are comparable to reported data in phenotypically normal patients.16The elimination half-lives of the active isomers decreased, compared with the half-lives before administration of cholinesterase, whereas that of the cis-cis isomer decreased only 50%. The initial volume of distribution did not differ from parameters in normal patients.16 

The estimated Emax, EC⁵⁰, k^eo, and γ (Hill coefficient) values did not differ from those found by us in phenotypically normal patients. The k^eovalue is in accord with the findings of Schiere et al.  30but higher than the EC⁵⁰and k^eovalues reported by Laurin et al.  31In the latter study and our study, a central compartment approach was used, but the methods used were not quite comparable. We cannot exclude that our estimations of these pharmacokinetic–pharmacodynamic parameters, at least in phenotypically normal patients, are flawed by the use of venous sampling as opposed to arterial, frequent sampling. In phenotypically normal patients, the hydrolysis of mivacurium in plasma is fast and starts immediately after injection. Measured plasma concentrations based on relatively sparse venous sampling may therefore not be representative of the changes in arterial concentrations, especially in the initial distribution phase.32In homozygous, atypical patients, however, hydrolysis is slow, and the plasma concentration remains high for a longer period of time. The data in these patients are therefore less flawed by the use of venous sampling. The major strength of this study is that the same approach was used in both groups of patients, indicating that if any differences are observed, they are the result of differences in the two groups of patients. Laurin et al.  31have presented both a central approach and a peripheral link model, and the latter method seems to enable a more precise pharmacokinetic–pharmacodynamic modeling. Also, Schiere et al.  30has presented a model with an interstitial space compartment interposed between plasma and the effect compartment. However, we decided to use the conventional method to make comparisons with data in phenotypically normal patients. All in all, we think that the pharmacokinetic–pharmacodynamic analyses support our pharmacokinetic findings and our conclusion that the longer duration of action of mivacurium in patients phenotypically homozygous for the atypical enzyme is due to pharmacokinetic changes caused by a less efficient pChe enzyme.

In conclusion, in patients phenotypically homozygous for the atypical variant, clearance of mivacurium is reduced and the elimination half-lives are prolonged, causing a markedly prolonged duration of action of mivacurium. Injection of human cholinesterase increases pChe activity, leading to a 9- to 15-fold increased clearance and a shorter elimination half-life of mivacurium. This in turn shortens the duration of action of mivacurium three to four times. Although the plasma concentration decreases rapidly, there is a delay from the time of injection of cholinesterase to the reappearance of T¹, most pronounced after early injection of cholinesterase. Neostigmine administered after cholinesterase increases the rate of recovery further, although recovery may be slow in some patients. The elimination rate constant from the effect compartment (k^eo) seems to be the same in patients phenotypically homozygous for the atypical variant and phenotypically normal patients.

The plasma cholinesterase determinations and the biochemical analyses were performed by Bodil Mathiesen (Lab Technician, Danish Cholinesterase Research Unit and Department of Pharmacology, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark). The plasma concentrations of mivacurium were determined by Bodil Mathiesen and by Johanne Couture, R.T. (Senior Technical Assistant, Faculty of Pharmacy, University of Montreal, Montreal, Canada), under the supervision of France Varin, B.Pharm., Ph.D. (Professor, Faculty of Pharmacy, University of Montreal).