Preliminary studies suggest that desflurane and isoflurane potentiate the action of muscle relaxants equally. However, variability between subjects may confound these comparisons. A crossover study was performed in volunteers on the ability of desflurane and isoflurane to potentiate the neuromuscular effect of vecuronium, to influence its duration of action, and on the magnitude and time course of reversal of potentiation when anesthesia was withdrawn.
Adductor pollicis twitch tension was monitored in 16 volunteers given 1.25 MAC desflurane on one occasion, and 1.25 MAC isoflurane on another. In eight subjects, vecuronium bolus dose potency was determined using a two-dose dose-response technique; the vecuronium infusion dose requirement to achieve 85% twitch depression also was determined. Also in these subjects, the magnitude and time course of spontaneous neuromuscular recovery were determined when the anesthetic was withdrawn while maintaining a constant vecuronium infusion. In the other eight subjects, the time course of action of 100 micrograms/kg vecuronium was determined.
Vecuronium's ED50 and infusion requirement to maintain 85% twitch depression were 20% less during desflurane, compared to isoflurane, anesthesia; vecuronium plasma clearance was similar during the two anesthetics. After 100 micrograms/kg vecuronium, onset was faster and recovery was longer during desflurane anesthesia. When the end-tidal anesthetic concentration was abruptly reduced from 1.25 to 0.75 MAC, twitch tension increased similarly (approximately 15% of control), and time for the twitch tension to reach 90% of the final change was similar (approximately 30 min) with both anesthetics. Decreasing anesthetic concentration from 0.75 to 0.25 MAC increased twitch tension by 46 +/- 10% and 25 +/- 7% of control (mean +/- SD, P < 0.001) with desflurane and isoflurane, respectively; 90% response times for these changes were 31 +/- 10 min and 18 +/- 7 min (P < 0.05), respectively.
Desflurane potentiates the effect of vecuronium approximately 20% more than does an equipotent dose of isoflurane.
Key words: Anesthetics, volatile; desflurane; isoflurane. Neuromuscular blocking agents, nondepolarizing: vecuronium. Potency: neuromuscular blocking agents.
VOLATILE anesthetic agents influence the bolus dose-response relationship for muscle relaxants,  a phenomenon known as potentiation. Similarly, volatile anesthetics reduce muscle relaxant infusion requirements.  Because volatile anesthetics usually do not alter the pharmacokinetics of a muscle relaxant, [2,3] this potentiation is presumed to result from increased neuromuscular sensitivity. These neuromuscular potentiating properties require characterization during the evaluation of new anesthetic agents such as desflurane. Results from several studies suggest that desflurane and isoflurane potentiate the neuromascular effect of vecuronium equally, [4,5] but large interindividual variability in vecuronium dose-response may have precluded the demonstration of clinically significant differences.
Just as volatile anesthetics potentiate the effects of muscle relaxants, reversal of potentiation occurs on withdrawal of a volatile anesthetic.  Consequently, volatile anesthetics may increase the safety margin at the end of anesthesia by reducing the muscle relaxant dose requirement (and plasma concentration). Yet, there is evidence that isoflurane impedes antagonism of neuromuscular block with anticholinesterase drugs, presumably by persistence of anesthetic at the neuromuscular junction. [7,8] Perhaps a less soluble volatile anesthetic, e.g., desflurane, might be eliminated from the neuromuscular junction more rapidly, resulting in more rapid reversal of potentiation.
To examine these issues, we compared the ability of desflurane versus isoflurane to potentiate the effect of vecuronium and to influence its duration of action. We also compared the effect of the withdrawal of these anesthetics on the magnitude and time course of reversal of potentiation of vecuronium. To minimize the effect of interindividual variability, we performed a paired crossover study in volunteers.
After obtaining approval from the institutional review board, we studied 16 volunteers aged 23–31 yr. We excluded potential subjects who were obese, had a concurrent medical condition, or were taking medication. Age, weight, height, gender, resting heart rate, and arterial blood pressure were recorded, and venous blood was obtained to confirm normal hematologic and biochemical indexes. The study was divided into three parts: in eight subjects, the potency and infusion dose requirements of vecuronium were measured (potentiation studies); in the same eight subjects, reversal of potentiation during volatile anesthetic withdrawal was observed (reversal of potentiation studies); and in the remaining eight subjects, the time course of 100 micro gram/kg vecuronium was determined (onset and duration of action studies).
Experimental conditions were similar for all 16 volunteers. Lactated Ringer's solution was slowly infused through an intravenous cannula in the right antecubital fossa. Anesthesia was induced using 2–3 mg/kg propofol and 15–30 micro gram/kg alfentanil and maintained using air/oxygen with desflurane (MAC = 7.25 ) or isoflurane (MAC = 1.28 ) at an end-tidal concentration equal to 1.25 MAC. The trachea was intubated after induction of anesthesia,  and the lungs were ventilated mechanically to maintain end-tidal PCO2at 30–35 mmHg (Datex Ultima, Helsinki). Electrocardiogram, SPO2, and core temperature (distal esophageal thermistor) were monitored continuously, and leg arterial blood pressure was measured noninvasively every 5 min (Dinamap, Critikon, Tampa). Core temperature was maintained at 36–37 degrees Celsius using forced air warming. The left ulnar nerve was stimulated using needle electrodes placed at the wrist. Supramaximal stimuli of 0.2 ms duration were delivered in a train-of-four at 2 Hz every 12 s. Preload was maintained at 250–400 g. The evoked twitch tension of the adductor pollicis muscle was measured with a calibrated force transducer (Myotrace, Houston, Texas) and amplified (DC Bridge Signal Conditioner, Gould Electronics, Valley View, OH). Twitch tension, end-tidal PCO2, and end-tidal anesthetic concentration were digitized (NB-MIO-16, National Instruments. Austin, TX), displayed (LabView, National Instruments), and recorded on a Macintosh IIci computer. These signals also were recorded on a strip-chart (TA240, Gould). The anesthetic concentration monitor was calibrated before each experiment. Anesthetic vapor was administered for > 60 min, and end-tidal carbon dioxide and isoflurane concentrations and adductor pollicis twitch tension were stable for > 15 min before vecuronium administration.
Each subject was studied twice, receiving desflurane on one occasion and isoflurane on the other. The order of administration of the anesthetic was varied by random allocation and the occasions were separated by at least 6 days. All neuromuscular responses were based on the first twitch of the train-of-four (twitch tension).
Potentiation Studies. In eight subjects, 10 micro gram/kg vecuronium was given. When twitch tension was maximally depressed (i.e., no change during three consecutive trains-of-four), the time was noted (time to peak effect) and a second dose of vecuronium was given. The size of the second dose was determined according to the peak response following the first dose, aiming to achieve 90% twitch depression after the second dose ; this value was reported by LabView immediately after the peak response. This two-dose dose-response technique provides accurate estimates of the potency of medium-duration muscle relaxants for which use of the traditional four- to five-dose cumulative dose technique yields biased estimates.  The peak twitch response after each of the two doses was used to estimate the ED50and ED sub 90 for each individual. .
After the response to the second dose peaked, 60–70% N2O was added to the inspired gas (maintaining unchanged the concentration of the potent agent), and a vecuronium infusion was initiated. The vecuronium infusion rate was adjusted to achieve stable twitch depression in the range 80–90%. When twitch tension varied < 1% for > 10 min and vecuronium infusion rate was unchanged for > 20 min, we recorded the infusion rate. On the second occasion (alternate anesthetic) for each volunteer, target twitch depression was within 1% of that observed on the first occasion. When twitch depression was stable, two venous blood samples were obtained 10 min apart.
Reversal of Potentiation Studies. In the same eight volunteers, the infusion of vecuronium was continued unchanged. The anesthetic vapor concentration was abruptly reduced to an end-tidal concentration of 0.75 MAC. This was achieved by increasing fresh gas flow to > 10 l/min and reducing the inspired concentration to zero. As the anesthetic end-tidal concentration approached 0.75 MAC, fresh gas flow was reduced and anesthetic vapor was reintroduced to maintain this value. Vapor concentration was recorded continuously, and the time for the end-tidal concentration to achieve 90% of the desired change was determined (anesthetic 90% response time). End-tidal vapor concentration was maintained at 0.75 MAC until twitch tension stabilized (varying < 1% for > 10 min), and two venous blood samples again were obtained 10 min apart. The plateau in twitch tension, the change from the previous plateau, and the time for the twitch tension to reach 90% of this change (twitch tension 90% response time) were determined. The anesthetic vapor concentration then was abruptly reduced to an end-tidal concentration of 0.25 MAC using the same procedure as before. The target end-tidal concentration again was maintained until twitch tension stabilized, and two additional blood samples obtained. Plateau twitch tension, change in twitch tension from the previous plateau, twitch tension 90% response time, and anesthetic 90% response time again were determined.
Blood samples were promptly centrifuged and the plasma was removed, buffered with NaH2PO4, and frozen for subsequent analysis of vecuronium and 3-desacetylvecuronium (gas chromatography with nitrogen phosphorus detection ). The assay is sensitive to 5 ng/ml and has a CV of 15% at 10 ng/ml.
Onset and Duration of Action Studies. In the remaining eight volunteers, 100 micro gram/kg vecuronium was administered, and twitch tension was recorded continuously until it recovered to a plateau (typically 80–100% of control twitch tension). Onset (12 s after the last detectable twitch response) and times until initial reappearance of a twitch response and until 10%, 25%, 75%, and 90% of the final recovery of twitch tension were determined.
Data Analysis and Statistics
To ensure that vecuronium concentrations for each anesthetic did not deviate systematically from steady-state during the infusion, we used repeated measures analysis of variance with Dunnett's test to compare the mean of the two vecuronium concentrations at each anesthetic dose (1.25, 0.75, and 0.25 MAC). This comparison was repeated for 3-desacetylvecuronium concentrations and for “neuromuscular blocking activity”(a measure expressing the combined neuromuscular activity of vecuronium and its metabolite, equal to vecuronium concentration +0.8 *symbol* 3-desacetylvecuronium ). For each experiment, the six vecuronium concentrations (two at each of three anesthetic states) were averaged to calculate vecuronium plasma clearance (assuming that at steady-state, infusion rate = amount eliminated = clearance *symbol* vecuronium concentration).
Peak twitch depression after 10 micro gram/kg vecuronium, potency estimates for vecuronium (ED50, ED90, and infusion rate), vecuronium's onset and duration of action (time to initial, 10%, 25%, 75%, and 90% recovery, and 25%-75% recovery time), steady-state vecuronium plasma concentration during the infusion, and vecuronium plasma clearance were compared between anesthetics using Student's paired t test. Similarly, the magnitude of changes in twitch tension and twitch tension 90% response time during the two anesthetic dose reductions were compared between anesthetics using Student's paired t test. Time to peak effect after 10 micro gram/kg vecuronium and onset time after 100 micro gram/kg vecuronium were compared using the Wilcoxon's signed-rank test (a nonparametric equivalent of the paired t test).
After 10 micro gram/kg vecuronium, twitch depression was greater with desflurance compared with isoflurane (Figure 1and Table 1); time to peak effect was similar during the two anesthetics. Vecuronium's potency (ED50and ED90) during 1.25 MAC anesthesia was approximately 20% greater with desflurane compared with isoflurane (P < 0.01). During steady-state vecuronium infusion, twitch depression stabilized at 83–89%; for each subject, steady-state twitch depression differed less or equal to 1% between anesthetics. The vecuronium infusion rate to maintain this twitch depression was approximately 20% less with desflurane compared with isoflurane (P < 0.05).
Onset and Duration of Action Studies
Reversal of Potentiation Studies
Vecuronium and 3-desacetylvecuronium concentrations were less during 1.25 MAC desflurane than during 1.25 MAC isoflurane anesthesia (Table 3). For both desflurane and isoflurane, as the vapor concentration was reduced (and time elapsed), vecuronium concentration did not change but 3-desacetylvecuronium concentration increased. Consequently, with isoflurane, “neuromuscular blocking activity”(vecuronium concentration +0.8 *symbol* 3-desacetylvecuronium concentration) increased by 5% as the anesthetic dose was reduced from 1.25 to 0.25 MAC. Vecuronium plasma clearance was similar during the two anesthetics (6.1 plus/minus 1.4 and 6.0 plus/minus 1.3 ml *symbol* kg sup 1 *symbol* min1for desflurane and isoflurane, respectively).
Decreasing the end-tidal anesthetic concentration from 1.25 to 0.75 MAC was 90% complete in < 2 min in all subjects. After this reduction, the increase in twitch tension and the 90% response time for this change were similar with both anesthetics (Figure 3and Table 4). The decrease in end-tidal anesthetic concentration from 0.75 to 0.25 MAC was also 90% complete in < 2 min in all subjects. After this reduction, the increase in twitch tension and the 90% response time for this change were greater with desflurane compared with isoflurane (Figure 3and Figure 4).
Volatile anesthetic agents potentiate the effects of muscle relaxants and thereby influence their dose requirement. We found that 20% less vecuronium is required for a given effect in the presence of 1.25 MAC desflurane compared with 1.25 MAC isoflurane. Similarly, desflurane produced a greater decrease in the infusion requirement for vecuronium. Because plasma clearance of vecuronium was similar during isoflurane and desflurane anesthesia, these findings suggest that desflurane potentiates the effect of vecuronium 20% more than isoflurane both during bolus and infusion administration. However, this greater potentiation with desflurane may not be evident if vecuronium is administered immediately after administration of the volatile anesthetic, before desflurane equilibrates with the neuromuscular junction.
Previous reports have not shown that desflurane affects the potency and duration of action of muscle relaxants more than does isoflurane. A study of vecuronium's potency suggested no differences between 0.5 MAC isoflurane versus desflurane.  However, vecuronium administration followed d-tubocararine and succinylcholine, and the doses of the inhaled anesthetic were small; each of these factors tends to obscure differences between the anesthetics. To address these issues, we studied greater concentrations of desflurane and isoflurane and administered no other muscle relaxants. In two other studies, one of pancuronium,  the other of atracurium,  ED50was less with 1.25 MAC of desflurane than with 1.25 MAC isoflurane; however, both studies lacked the power (sufficient sample size) to demonstrate statistical significance. Large interindividual variation obscuring clinically relevant differences is a recurring difficulty in the comparison of neuromuscular potentiation by different anesthetics. For instance, differences similar in magnitude to those we detected were suggested between enflurane and halothane potentiation of both pipecuronium  and doxacurium  and isoflurane and halothane potentiation of vecuronium ; but despite sample sizes greater than ours, no statistically significant differences were demonstrated. Our use of a crossover design eliminated the confounding influence of interindividual variation. These differences in study design enabled us to conclude that desflurane potentiated vecuronium and increased its duration of action to a greater extent than did isoflurane.
Our findings suggest that a smaller dose of vecuronium is required during desflurane, compared with isoflurane, anesthesia. For example, during the duration of action studies, had we administered a smaller dose of vecuronium during desflurane anesthesia, the clinical duration might have been similar to that during isoflurane anesthesia. Alternatively, vecuronium doses could be administered less frequently. However, this dose reduction will be small compared to the large variation in neuromuscular response to vecuronium during either anesthetic technique (e.g., twitch depression with 10 micro gram/kg vecuronium varied from 10% to 90%). This, and the marked influence of other factors that commonly occur during surgery (e.g., local hypothermia ), suggests that a clinician is unlikely to observe the decreased vecuronium requirements during desflurane anesthesia. Nonetheless, in any given individual, if supplemental vecuronium doses are given as indicated by neuromuscular monitoring (as opposed to a predetermined regimen), less vecuronium will be given during desflurane anesthesia. In turn, plasma vecuronium concentrations at the end of surgery should be less during desflurane than during isoflurane anesthesia (as evidenced in Table 3), thereby facilitating spontaneous recovery as the inhaled anesthetic is withdrawn.
As a volatile anesthetic is withdrawn, twitch tension should increase as reversal of potentiation occurs. Our study examined neuromuscular recovery as isoflurane or desflurane concentrations were decreased in two stages from an initial concentration of 1.25 MAC. The reduction of anesthetic concentration from 1.25 MAC to 0.75 MAC produced small, similar increases in twitch tension for both anesthetics. A further decrease in anesthetic concentration (from 0.75 MAC, a dose commonly used during clinical anesthesia, to 0.25 MAC) further increased twitch tension for both anesthetics but yielded the greater increment during desflurane compared with isoflurane anesthesia. This suggests that clinicians using desflurane in concentrations of 0.75 MAC or greater can advantageously apply the greater magnitude of reversal of potentiation we observed with desflurane. Only one previous study has examined the neuromuscular response to anesthetic withdrawal as muscle relaxant concentrations were maintained constant.  In that study, a two-stage reduction of enflurane from 1.25 to 0.75 MAC and then 0.25 MAC during steady-state d-tubocurarine infusion increased twitch tension from 8% to 57% and then to 91%. The time for twitch tension to reach the new value on each occasion was approximately 40 min. These values are similar to those we observed with desflurane, although methodologic differences (e.g., different muscle relaxant and different mode of neuromuscular stimulation) limit comparisons.
A necessary condition for the infusion and reversal of potentiation studies is that vecuronium plasma concentrations remain constant; these concentrations might have varied if we did not achieve steady-state or if changes in anesthetic depth influenced plasma clearance of vecuronium. Plasma vecuronium concentrations confirm minimal deviation from steady-state. However, consistent with its longer terminal half-life.  3-desacetylvecuronium concentrations increased over time with both anesthetics. This resulted in 5% greater “neuromuscular blocking activity” with 0.25 MAC compared with 1.25 MAC anesthesia during isoflurane, but not desflurane, anesthesia. To address this limitation, we calculated that plateau twitch tension during 0.25 MAC isoflurane anesthesia would be increased by only 4% if “neuromuscular blocking activity” had remained unchanged from that at 1.25 MAC isoflurane (Appendix). This deviation is not of sufficient magnitude to invalidate our observations.
Recovery of neuromuscular function during withdrawal of a volatile anesthetic should increase the margin of safety of a muscle relaxant. However, several studies suggest that the neuromuscular potentiating effects of volatile anesthetics persist after their administration. For example, in subjects in whom neostigmine antagonized vecuronium, 17 min after isoflurane was discontinued (when end-tidal concentrations were negligible) neuromuscular function was impaired compared to a group never given isoflurane.  The investigators contended that this neuromuscular effect resulted from persistence of isoflurane at the neuromuscular junction and speculated that less soluble volatile anesthetics such as sevoflurane and desflurane would have a faster response time for reversal of potentiation. That speculation is not supported by the results of the current study: twitch tension 90% response time was similar or longer during desflurane ([nearly equal] 30 min) than during isoflurane ([nearly equal] 20–30 min) anesthesia. (We are unable to explain why twitch tension 90% response time differed between the first and second reductions in isoflurane concentrations.) These twitch tension 90% response times correspond to time constants of approximately 10–20 min. much less than the time constants associated with washout of these two anesthetics from muscle (49 and 80 min for desflurane and isoflurane, respectively).  This suggests that the site of action of potentiation is not muscle, but in a tissue with greater blood flow than muscle, perhaps the neuromuscular junction. In that desflurane's tissue-blood partition coefficient is 20–30% less than that of isoflurane in both neural tissue (brain) and muscle,  the longer twitch tension 90% response time after its withdrawal is unexpected (and should probably be reconfirmed in additional studies). Despite this slower response time, the absolute rate at which twitch tension recovered when the anesthetic was withdrawn was greater during desflurane than during isoflurane anesthesia (Figure 3). Although our study design differs from clinical practice (we administered vecuronium during the period of anesthetic washout), thereby limiting its clinical applicability, our results suggest that emergence from desflurane anesthesia might offer the possibility of more reliable antagonism of neuromuscular block at the end of anesthesia than that associate isoflurane.
In summary, desflurane potentiated the effect of vecuronium 20% more than did isoflurane and prolonged the duration of a dose often used to facilitate trachea intubation. The increment in the dose-response relationship is small compared to the large interindividual differences in the response to vecuronium. Nonetheless, this greater potentiation resulted in a greater magnitude of recovery when desflurane was withdrawn during constant vecuronium infusion than that associated with isoflurane.
The authors thank Edmond I. Eger II, M.D., for assistance with study design.
Prediction of Effect from a Change in Plasma Concentration
Our studies of reversal of potentiation require that “neuromuscular blocking activity” remain constant as anesthetic concentrations are decreased. However, we observed a 5% increase in “neuromuscular blocking activity” during the isoflurane experiments. To examine the effect of this small change in neuromuscular blocking activity, we assumed that the relationship between muscle relaxant concentration and neuromuscular blockade can be expressed using the Hill equation:Equation 1where Calcium is “neuromuscular blocking activity”(plasma vecuronium concentration 10.8 plasma 3-desacetylvecuronium concentration). Calcium50is the “neuromuscular blocking activity” depressing twitch 50%, and effect is percent twitch depression. Equation 1can be rewritten as:Equation 2.
At 0.25 MAC isoflurane, mean Calcium is 153 ng/m1 (Table 3) and effect is 43%(i.e., twitch tension 57% of control, Figure 3and Table 4). Using the mean value for gamma calculated from the dose-response estimation (4.4), Calcium50can be estimated as:Equation 3.
Had Calcium remained at 117 ng/ml (i.e., no change in Calcium during the course of the experiment), instead of increasing to 153 ng/ml, effect would be:Equation 4.
Thus, had “neuromuscular blocking activity” remained constant during the course of the experiment, mean twitch tension during 0.25 MAC isoflurane should have been 61% of control (in contrast to the observed value of 57%), still significantly different from the value during desflurane anesthesia.