Share this article on:

The Effect of Nitrous Oxide on the Dose-Response Relationship of Rocuronium

Kopman, Aaron F. MD*; Chin, Wanda A. MD; Moe, Jimmy MD; Malik, Rawshan MD

doi: 10.1213/01.ANE.0000148076.17331.51
Anesthetic Pharmacology: Research Report
Chinese Language Editions

It has been generally assumed that nitrous oxide (N2O) enhances the effects of nondepolarizing muscle relaxants only weakly if at all. More recent evidence suggests that drug potency may be more intense under N2O anesthesia compared with total IV anesthesia (TIVA). However, the magnitude of this effect has not been well defined. We measured the 50% effective dose of rocuronium in 35 patients receiving N2O-propofol-opioid anesthesia and a comparable group receiving TIVA. A single dose of rocuronium was given to each patient and drug potency was calculated for each individual from the Hill equation assuming a log-dose/logit slope of 4.5. In both groups, the relaxant was administered 15 min after induction of anesthesia. Neuromuscular function was measured using electromyography with single stimuli at 0.10 Hz. We measured a 50% effective dose of 0.209 ± 0.051 mg/kg during TIVA and of 0.166 ± 0.041 mg/kg during N2O anesthesia, a decrease of 20% (P < 0.001). The clinical importance of this effect must be considered modest; however, estimates of potency that are usually obtained during N2O anesthesia may underestimate drug requirements at the time of induction of anesthesia.

IMPLICATIONS: The potency of rocuronium is approximately 20% less after 15 min of nitrous oxide than after 15 min of total IV anesthesia. Thus, estimates of drug potency that are obtained under nitrous oxide anesthesia may underestimate drug requirements at the time of induction.

Departments of Anesthesiology, *New York Medical College, Valhalla; and †Saint Vincent’s Hospital Manhattan, New York City, New York

Support was provided solely from institutional and departmental sources.

WAC is currently a resident in anesthesiology at New York University Medical Center.

Accepted for publication October 4, 2004.

Address correspondence to Aaron F. Kopman, MD, Department of Anesthesiology, Room N.R. 408, St. Vincent’s Hospital Manhattan, 170 West 12th St., New York City, NY 10011. Address e-mail to

When performing dose-response studies of neuromuscular blocking drugs in humans, the use of volatile anesthetics are generally avoided. All such anesthetics, to one degree or another, enhance the effects of muscle relaxants. Thus, studies of potency are usually performed during nitrous oxide (N2O) anesthesia supplemented with a hypnotic and/or an opiate, the assumption being that N2O has little or no effect at the neuromuscular junction. Despite this general consensus, objective evidence for this position is meager.

There is at least some evidence, however, that N2O may not be devoid of neuromuscular effects. Fiset et al. (1) estimated that 70% N2O was associated with an apparent increase in the potency of vecuronium compared with thiopental, fentanyl, oxygen anesthesia. Unfortunately, the confidence limits for their estimation were quite wide, and the 50% effective dose (ED50) values they calculated (N2O versus O2) differed by <15%. In addition, the duration of N2O administration before the administration of vecuronium was only 5 min. Thus, their results were suggestive but not totally convincing. A somewhat more compelling case for the proposition that N2O enhances the nondepolarizing neuromuscular block was put forth by Plaud et al. (2). They compared the effects of 0.10 mg/kg mivacurium in 10 patients receiving propofol-O2 anesthesia to 10 patients who had received a similar dose of propofol plus 70% N2O for 15 min. Maximal block in the 2 groups was 84% ± 19% versus 99% ± 2%, respectively (P < 0.01). However, because of small sample size, and the fact that a significant number of the subjects in both groups achieved 100% twitch suppression, the ED50/95 values could not be compared in the 2 groups. Thus, the magnitude of the neuromuscular enhancement produced by N2O was undetermined. We decided to revisit this issue because we believed that the answer to this question had not heretofore been adequately addressed. In our present study, we have modified Plaud et al.’s protocol so that the magnitude of the N2O effect on drug potency (if any) could be quantified.

Back to Top | Article Outline


Seventy ASA physical status I–II, adult patients (aged 18–61 yr) undergoing elective surgical procedures were included in the study. All patients were free from neuromuscular disease and had a body mass index <32.5. The protocol was approved by our hospital’s Human Subject Review Committee and informed patient consent was obtained. Patients were randomly divided into two groups.

In Group 1 (total IV anesthesia, TIVA) (n = 35), anesthesia was induced with remifentanil 3 μg/kg plus propofol 2.0 mg/kg IV, and tracheal intubation was accomplished without the use of neuromuscular blocking drugs. Anesthesia was maintained with air-oxygen, and infusions of propofol 75–100 μg · kg−1 · min−1 and remifentanil 0.15–0.20 μg · kg−1 · min−1. Ventilation was controlled, and end-tidal Pco2 was maintained between 34–40 mm Hg.

The indirectly evoked integrated compound action potential of the first dorsal interosseous muscle to supramaximal stimulation of the ulnar nerve at the wrist was measured and recorded using a Datex® NMT 100 monitor (Datex Instrumentarium, Helsinki, Finland). Single stimuli at 0.10 Hz were administered during the period of observation, and twitch depression was continuously recorded. Control twitch height was established after a 15-min period of baseline stabilization. Immediately after twitch height calibration, a single dose of rocuronium was administered. An upper body forced air warming blanket maintained hand temperature >33°C and esophageal temperature >35°C in all subjects.

The first patient received a bolus of 0.17 mg/kg rocuronium. This dose was selected to approximate what we anticipated (from previous work in our department) to be an ED50 dose. Using the Hill equation (with a postulated log-dose/logit slope of 4.50), the ED50 was calculated for this patient. The second subject received a dose that approximated the calculated ED50 for patient 1. Each successive subject was given a dose that equaled the running average of the estimated ED50 of the previously studied subjects. No attempt was made to administer larger or small multiples of the ED50 dose because we wished to avoid responses (0% or 100%) that could not be plotted on a logit scale. For each subject the estimated ED50 was computed from the Hill equation by a method previously described (3). The arithmetic mean, median, standard deviation, and standard error of the mean were then calculated from the individual ED50 values.

In Group 2 (N2O) (n = 35), the protocol was identical to that used in Group 1 except that each patient was ventilated with a mixture of 70% N2O and 30% oxygen after tracheal intubation. Initial gas flows were 10 L/min. Flows were decreased to 5 L/min when the end-tidal N2O concentration exceeded 65%. Anesthesia was supplemented with a propofol infusion of 50–75 μg · kg−1 · min−1 and additional IV opioids as required.

The mean estimated ED50 values in Groups 1 and 2 were compared using an unpaired Student’s t-test. A P value < 0.05 was considered to be statistically significant. The 95% confidence limits for the respective average values were also calculated.

Back to Top | Article Outline


There were no important differences in the demographics of the 2 groups studied (Table 1). The ED50 for rocuronium in the N2O group (0.166 ± 0.041 [sd] mg/kg) was less than in the TIVAgroup (0.209 ± 0.051 [sd] mg/kg) (P < 0.001) (Table 2; Fig. 1).

Table 1

Table 1

Table 2

Table 2

Figure 1

Figure 1

Back to Top | Article Outline


We have previously published a full description and discussion of our method for estimating ED50/95 values (3). Briefly, a line can be defined if one point and the slope of the line is known. Because the slope of the dose-response relationship for all neuromuscular blocking drugs is essentially parallel, once a single point along that line is known, the ED50 and ED95 for a patient can be calculated. Based on our previous experience with rocuronium (3), we chose to use a log-dose/logit slope of 4.5 for estimating the ED50 and ED95 of rocuronium. If this approach is correct, there is no logical reason for administering doses other than the estimated ED50 when constructing dose-response relationships. Because linear regression analysis is not used, there is no reason to give doses such as the estimated ED10 or ED90. Avoiding these latter doses markedly reduces the frequency with which either zero or 100% twitch depression may occur, responses that cannot be plotted using log-dose/logit analysis. In the present study, we did not encounter any such “unplottable” values. Our final estimate of rocuronium’s ED50 in the N2O group (0.166 mg/kg) is not different from the value we calculated in our previous electromyographic study (3) of 40 subjects using a similar protocol. In that study, we calculated an ED50 of 0.174 ± 0.039 mg/kg (at a slope of 4.50, P < 0.01).

Perhaps our calculations of potency are only approximate because the actual slope of the dose-response relationship was never measured. However, fairly wide variations in the log-dose/logit slope have very little effect on the estimated ED50. Our present electromyographic estimate for the ED50 (slope = 4.5) would increase by about 3% at a slope of 3.5, and decrease by only 1% at a slope of 5.5. Thus, this variable is quite robust. Errors in estimating the “true” slope have a greater impact on the calculated ED95 which would change by +24% to −12% at slopes of 3.5 and 5.5, respectively.

The duration and pattern of neuromuscular stimulation in both groups were kept the same because there is credible evidence that the duration of baseline stabilization before drug administration can markedly affect the onset, duration, and presumably the potency of neuromuscular block (4,5).

In his 1985 discussion of the neuromuscular effects of N2O, Miller (6) concluded that available data “…do not supply a conclusive answer to the question of enhancement of a nondepolarizing blockade by nitrous oxide, … evidence indicates that nitrous oxide enhances the effects of nondepolarizing muscle relaxants only weakly if at all.” Fiset et al. (1) subsequently concluded that N2O did in fact have a slight enhancing effect on neuromuscular blockade. They reported that administration of N2O was associated with about a 13% decrease in the ED50 of vecuronium. However, their 95% confidence limits for this value were quite wide and ranged from as little as +2% to as much as +40%. Thus, although their observations were suggestive, Miller’s cautious judgment still seemed appropriate.

Work by Plaud et al. (2) is more convincing. After 15 minutes of propofol anesthesia (n = 10), they found that mivacurium 0.10 mg/kg produced on average 84% ± 19% block at the adductor pollicis muscle (range, 35%–100%) with a median value of 90% block. In a similar group of 10 patients receiving propofol-N2O, the average peak effect was 99% (range, 93%–100%) and the median value was 100%. An increase of this magnitude (in effect from 85%–90% to 98%–99% block) represents considerably more than a 15%–20% increase in potency. In fact, assuming a slope for the log-dose/logit dose-response relationship of 4.5, an ED98 dose represents an effect site drug level that is 46% more than an ED90 dose. Unfortunately, the design of the study of Plaud et al. did not permit a quantitative assessment of the effect of N2O on neuromuscular potency. Our protocol in contrast allowed us to measure the magnitude of this effect.

Our results in large measure confirm the observations of Fiset et al. (1), although a direct comparison of the two studies is difficult. Fiset et al. used regression analysis when estimating drug potency, and only administered N2O for 5 minutes. Although we observed a larger decrease (20%) in the ED50 of rocuronium associated with administration of N2O than the 13% decrease in the ED50 of vecuronium reported by Fiset et al., it is not clear that this difference is meaningful or statistically different.

Our calculated ED50 for rocuronium was 0.166 mg/kg in the N2O group versus 0.209 mg/kg in the TIVA group (P < 0.001). Published values for the ED50 of rocuronium during N2O anesthesia vary from a low of 0.15 mg/kg (7) to as much as 0.22 mg/kg (8). Our estimate agrees most closely with the values of 0.168, 0.170, and 0.183 mg/kg proposed respectively by Bock et al. (9), Foldes et al. (10), and Lowry et al. (11). Estimates of the ED50 during TIVA range from a low of 0.167 mg/kg (12) to a high of 0.21 mg/kg (13). Because of the wide discrepancies in reports from different investigators, it is not possible to determine whether any difference in the potency of rocuronium exists during N2O anesthesia versus TIVA based on historic controls.

Although our results confirm that estimates of relaxant potency determined during TIVA are not interchangeable with those measured when N2O is added to the anesthetic regimen, they shed no light on the mechanism that is responsible for the increase in neuromuscular potency seen with N2O. Plaud et al. suggested that N2O augments the effect of vecuronium by altering delivery of drug via hemodynamic changes rather than a direct effect at the neuromuscular junction. They reasoned that saturation of muscle tissue by N2O after 15 minutes of anesthesia is only about 30% complete, so some other mechanism for the enhancement they observed must be responsible. We do not find this argument convincing. There is no reason to assume that perfusion of muscle tissue as a whole and that of the neuromuscular junction are similar. Equilibration between plasma and the neuromuscular effect compartment is very rapid. The peak effect of succinylcholine (in subparalyzing doses) is seen in <2 minutes (14), and the equilibration half-life between plasma and the effect compartment for nondepolarizing relaxants such as rocuronium approximates 5 minutes (15). Thus, 15 minutes in all likelihood gives ample time for the peak effect of N2O at the neuromuscular junction to manifest itself.

The close agreement between Fiset et al.’s1 observations (taken at 5 minutes) and our results (15 minutes) suggests that the period required before the full effect of N2O on neuromuscular transmission manifests itself is relatively short. However, because we only examined one point in time, we can only speculate on the point.

It is possible of course that Plaud et al. are correct. N2O may not alter the concentration of relaxant at the neuromuscular junction required for 50% neuromuscular block (the CE50). Miller (6) describes a study (no peer-review citation provided) in which steady-state partial neuromuscular block with d-tubocurarine was achieved in five patients during halothane-oxygen anesthesia. The addition of 70% N2O to the anesthetic mixture produced no discernable change in twitch tension. If these results can be duplicated in a larger investigation, it would support the suggestion that N2O produces its effect by increasing drug delivery to the myoneural junction. Nevertheless, there is convincing evidence (16,17) that the enhancement seen with volatile anesthetics has a pharmacodynamic rather than a pharmacokinetic basis. Ultimately, if the mechanism of N2O’s neuromuscular effect is to be settled beyond dispute, a pharmacokinetic-pharmacodynamic comparison during TIVA versus a N2O-based anesthetic will be required.

In an attempt to shorten the duration of action of rocuronium whereas still achieving good-to-excellent conditions for tracheal intubation, it has been suggested that an “intubation dose” only 1.5 times the ED95 (0.45–0.5 mg/kg) is required provided that laryngoscopy is delayed to 75–90 seconds after drug administration (18–20). A common finding after doses of this magnitude is that complete twitch depression at the adductor pollicis may not take place. This observation is no longer surprising. Assuming an ED50 of 0.21 mg/kg and a slope of 4.5, the ED98 of rocuronium after 15 minutes of TIVA would approximate 0.50 mg/kg. In addition, McCoy et al. (5) have demonstrated that increasing periods of control stimulation are associated with decreasing time to onset of neuromuscular block with atracurium, vecuronium, and mivacurium at the adductor pollicis muscle. Presumably this represents an apparent decrease in the ED50. It follows that the dose of rocuronium required to reliably abolish an indirectly evoked response at the hand shortly after induction of anesthesia may approach 2 times the ED95.

Satisfactory conditions for tracheal intubation can, of course, be achieved without the administration of neuromuscular blocking drugs (21). However, if the above line of reasoning is extended to muscles such as the masseter, the diaphragm, and the vocal cord abductors it becomes clear why doses of relaxant less than twice the “usually accepted” ED95 may result in suboptimal conditions for tracheal intubation in the absence of adequate levels of anesthesia.

In summary, N2O has a modest but measurable influence on the ED50 value of rocuronium. After 15 minutes of N2O administration, this variable is reduced by about 20%. Published reports on the potency of neuromuscular blocking drugs (which are usually obtained during N2O anesthesia) must be read with an awareness that they may underestimate drug requirements at the time of induction of anesthesia.

Back to Top | Article Outline


1. Fiset P, Balendran P, Bevan DR, Donati F. Nitrous oxide potentiates vecuronium neuromuscular blockade in humans. Can J Anaesth 1991;38:866–9.
2. Plaud B, Debaene B, Donati F. Duration of anesthesia before muscle relaxant injection influences level of paralysis. Anesthesiology 2002;97:616–21.
3. Kopman AF, Klewicka MM, Neuman GG. An alternate method for estimating the dose-response relationships of neuromuscular blocking drugs. Anesth Analg 2000;90:1191–7.
4. Symington MJ, McCoy EP, Mirakhur RK, Kumar N. Duration of stabilization of control responses affects the onset and duration of action of rocuronium but not suxamethonium. Eur J Anaesthesiol 1996;13:377–80.
5. McCoy EP, Mirakhur RK, Connolly FM, Loan PB. The influence of the duration on control stimulation on the onset and recovery of neuromuscular block. Anesth Analg 1995;80:364–7.
6. Miller RD. Neuromuscular effects of nitrous oxide. In: Eger EI, ed. Nitrous oxide/N2O. New York: Elsevier, 1985:179–84.
7. Cooper RA, Mirakhur RK, Elliot P. Estimation of the potency of ORG 9426 using two different modes of peripheral stimulation. Can J Anaesth 1992;39:139–42.
8. Booij LH, Knape HT. The neuromuscular blocking effect of ORG 9426. Anaesthesia 1991;46:341–3.
9. Bock M, Klippel K, Nitsche B, et al. Rocuronium potency and recovery characteristics during steady-state desflurane, sevoflurane, isoflurane or propofol anaesthesia. Br J Anaesth 2000;84:43–7.
10. Foldes F, Nagashima H, Nguyen HD, et al. The neuromuscular effects of ORG 9426 in patients receiving balanced anesthesia. Anesthesiology 1991;75:191–6.
11. Lowry DW, Mirakhur RK, McCarthy GJ, et al. Neuromuscular effects of rocuronium during sevoflurane, isoflurane, and intravenous anesthesia. Anesth Analg 1998;87:936–40.
12. Oris B, Crul JF, Vandermeersch E, et al. Muscle paralysis by rocuronium during halothane, enflurane, isoflurane, and total intravenous anesthesia. Anesth Analg 1993;77:570–3.
13. Meistelman C, Plaud B, Donati F. Rocuronium (ORG 9426) neuromuscular blockade at the adductor muscles of the larynx and adductor pollicis in humans. Can J Anaesth 1992;39:66–9.
14. Kopman AF, Klewicka MM, Kopman DJ, Neuman GG. Molar potency is predictive of the speed of onset of neuromuscular block for agents of intermediate-, short-, and ultra-short duration. Anesthesiology 1999;90:425–31.
15. Dragne A, Varin F, Plaud B, Donati F. Rocuronium pharmacokinetic-pharmacodynamic relationship under stable propofol or isoflurane anesthesia. Can J Anaesth 2002;49:353–60.
16. Cannon JE, Fahey MR, Castagnoli KP, et al. Continuous infusion of vecuronium: The effect of anesthetic agents. Anesthesiology 1987;67:503–6.
17. Stanski DR, Ham J, Miller RD, Sheiner LB. Pharmacokinetics and pharmacodynamics of d-tubocurarine during nitrous oxide-narcotic and halothane anesthesia in man. Anesthesiology 1979;51:235–41.
18. Kopman AF, Klewicka MM, Neuman GG. Reexamined: the recommended “intubating dose” for nondepolarizing blockers of rapid-onset. Anesth Analg 2001;93:954–9.
19. Eikermann M, Hunkemöller LP, Armbruster W, et al. Optimal rocuronium dose for intubation during inhalation induction with sevoflurane in children. Br J Anaesth 2002;89:277–81.
20. Barclay K, Eggers K, Asai T. Low-dose rocuronium improves conditions for tracheal intubation after induction of anaesthesia with propofol and alfentanil. Br J Anaesth 1997;78:92–4.
21. Scheller MS, Zornow MH, Saidman LJ. Tracheal intubation without the use of muscle relaxants: a technique using propofol and varying doses of alfentanil. Anesth Analg 1992;75:788–93.
© 2005 International Anesthesia Research Society