The Maximum Depth of an Atracurium Neuromuscular Block Antagonized by Edrophonium to Effect Adequate Recovery

Thirty adult patients, ASA physical status 1 or 2, scheduled for elective surgery gave informed consent to participate in the study, which was approved by the ethics committee of the Royal Melbourne Hospital. Patients with neuromuscular, hepatic, or renal disease, those with electrolyte abnormalities, and those taking medications that could interfere with neuromuscular function were excluded. The patients received 0.15 mg *symbol* kg sup -1 diazepam orally 2 h before anesthesia. On arrival in the operating room, the patients' electrocardiogram (ECG) and blood pressure were monitored. Anesthesia was induced with 3-4 mg *symbol* kg sup -1 thiopental and 1-2 micro gram *symbol* kg sup -1 fentanyl and maintained with 70% N²O in oxygen and continuous infusion of propofol (maintenance infusion rate 0.05-0.15 mg *symbol* kg sup -1 *symbol* min sup -1). Ventilation was controlled to maintain an end-tidal P^CO2between 30 and 35 mmHg. Heat loss was minimized by surface warming, humidification of inspired gases, and warming intravenous fluids.

The ulnar nerve was supramaximally stimulated at the wrist via surface gel electrodes by TOF stimulation (four pulses delivered at 2 Hz), which was repeated every 12 s, using a DigiStim 3 Plus peripheral nerve stimulator (Neuro Technology, Houston, TX). The force of thumb adduction was measured by a Grass FT-10 force-displacement transducer and recorded continuously on a polygraph. Preload tension on the thumb was maintained at 300 g throughout the investigation. After the twitch response was stable for not less than 3 min, the patient received 0.2-0.3 mg *symbol* kg sup -1 atracurium. Tracheal intubation was performed when the maximum neuromuscular block was achieved. The block was maintained by an infusion of atracurium according to a previously developed profile, which was designed to maintain a constant level of neuromuscular block. [12]The infusion was given via a specially built syringe pump, which was capable of reading successive memory locations containing he infusion rate data at specific times and adjusting the infusion rate according to patient size and target concentration. The target concentration was adjusted to maintain the desired level of neuromuscular block, which varied in individual patients.

After the level of neuromuscular block had been stable for a minimum of 20 min and was no longer required for surgery, incremental intravenous doses of edrophonium were administered while the infusion of atracurium was continued. After each dose, the twitch response was observed for the peak antagonistic effect, as judged by three consecutive twitches of equal height, before the next incremental dose was given. Incremental doses were given until the block was adequately antagonized, as judged by a TOF ratio greater or equal to 70% or until no further antagonism of the neuromuscular block occurred. The size of the incremental doses of edrophonium were empirically chosen so that each dose had a comparable effect, and three or four doses were required to adequately antagonize the block. In patients in whom the block was antagonized to a TOF ratio < 70%, the infusion of atracurium was stopped, and the neuromuscular block was allowed to spontaneously recover.

Data were analyzed using the statistical package Systat 5.0 (Evanston, IL). The degree of antagonism of the patient's neuromuscular block by edrophonium was categorized as being adequate if antagonized to TOF ratio greater or equal to 70% or inadequate if antagonized to a TOF ratio < 70%. The probability of adequate antagonism of a neuromuscular block by edrophonium was determined using a logistic regression model, which had the following form: Equation 1where the probability of adequate antagonism varies from 0 (inadequate antagonism) to 1 (adequate antagonism); NMBⁱ(percent depression of first twitch height) is the level of neuromuscular block maintained by the atracurium infusion before administration of edrophonium; s is a dimensionless power function that determines the steepness of the slope of the probability versus level of neuromuscular block curve; and NMB^p50(percent depression of first twitch height) is the level of neuromuscular block that results in a 50% probability of adequate antagonism. In the Equation 1model, the two parameters NMB^p50and s were simultaneously derived from the patient's NMBⁱand their associated adequate (1) or inadequate (0) antagonism by an iterative approach using a maximum likelihood estimate with the NONLIN module of Systat. For the analysis, a neuromuscular block of 1% depression of the first twitch height was defined as always being adequately antagonized, and a block of 100% depression of the first twitch height was defined as always being inadequately antagonized.

Cumulative dose-response curves were constructed in patients in whom the neuromuscular block was antagonized by edrophonium to a TOF ratio < 70%. The antagonistic effect of each incremental dose of edrophonium tin the neuromuscular block (NMB^t) was related to the level of neuromuscular block being maintained by the atracurium infusion before the administration of edrophonium (NMBⁱ) by: Equation 2where NMB^e, is the level of neuromuscular block after the incremental dose of edrophonium, and NMB^maxis the plateau of neuromuscular block that occurred after the maximum antagonistic effect of edrophonium. The neuromuscular block was quantified by the percentage depression of the first twitch height. Linear regression of the logit transformation of NMB^ton the logarithm of the cumulative dose of edrophonium was used to estimate the dose of edrophonium to produce 95% of the peak antagonistic effect. A similar method was used to construct cumulative dose-response curves in patients who were antagonized to a TOF ratio greater or equal to 70%. The level of neuromuscular block was assessed by the TOF ratio, and a logit-log dose-response curve was used to estimate the dose of edrophonium to produce a TOF ratio of 70%.

Values are reported as mean plus/minus SD. The demographic data of the patients were compared using Student's t test, with P < 0.05 considered significant.

Nine men and 21 women participated in the study, with a mean age of 61 plus/minus 16 yr and a mean weight of 72 plus/minus 14 kg. There was no significant difference in the demographic data of the patients whose neuromuscular block was adequately antagonized by edrophonium compared to those in whom antagonism was inadequate.

In 14 patients, the neuromuscular block was antagonized to a TOF ratio greater or equal to 70% by edrophonium. Figure 1illustrates the typical recording of the first and fourth twitch height from such a case. In 16 patients, the neuromuscular block was inadequately antagonized by edrophonium, indicating a ceiling effect for antagonism by edrophonium. The relationship between the level of neuromuscular block and adequate and inadequate antagonism by edrophonium is summarized in the upper part of Figure 2. Three areas can be distinguished: there is a level of neuromuscular block below which antagonism to a TOF ratio greater or equal to 70% always occurred; there is an overlapping area where the neuromuscular block in some patients was adequately antagonized, whereas it was inadequately antagonized in others; and there is a level of neuromuscular block above which a TOF ratio greater or equal to 70% could never be achieved. The probability curve estimated by the logistic regression model for adequate antagonism by edrophonium versus the level of neuromuscular block is illustrated in the lower part of Figure 2. The estimated NMB^p50was 67 plus/minus 3% depression of first twitch height (mean plus/minus SE, 95% confidence interval 59-74% depression of first twitch height), and the power function s was 8.0 plus/minus 3.3 (mean plus/minus SE). The logistic regression model estimated that a neuromuscular block of 96% depression of the first twitch height had only a probability of 0.05 for adequate antagonism by edrophonium, whereas a block of 46% depression of the first twitch height had a probability of 0.95 for adequate antagonism.

(Table 1) details the adequacy of antagonism by edrophonium when the neuromuscular block was quantified as the number of responses to TOF stimulation. When fewer than four responses to TOF stimulation were present, the block could never be antagonized to a TOF ratio greater or equal to 70%. In the seven patients in whom there were four responses to TOF stimulation and the neuromuscular block was inadequately antagonized by edrophonium, the neuromuscular block was antagonized to a TOF ratio of 52-67%, and recovery to a TOF ratio greater or equal to 70% occurred within 6 min of cessation of the atracurium infusion. By contrast, in the two patients with one response to TOF stimulation, edrophonium only antagonized the block to a TOF ratio of 26% and 46% and recovery took 25 and 23 min, respectively, after cessation of the infusion.

In patients in whom edrophonium resulted in a TOF ratio greater or equal to 70%, the total dose of edrophonium administered was 0.67 plus/minus 0.28 mg *symbol* kg sup -1 and varied from 0.16 mg *symbol* kg sup -1 in a patient with a baseline block of 26% depression of the first twitch height to 1.18 mg *symbol* kg sup -1 in a patient with a baseline block of 77% depression of the first twitch height. The mean estimated dose of edrophonium to produce antagonism of the block to a TOF ratio of 70% was 0.47 plus/minus 0.26 mg *symbol* kg sup -1 in patients with a baseline neuromuscular block greater than 50% depression of the first twitch height (range 54-77% depression of the first twitch height; n = 11).

In patients in whom a TOF ratio greater or equal to 70% could not be achieved by edrophonium, the total dose of edrophonium administered was 1.67 plus/minus 0.26 mg *symbol* kg sup -1. The dose of edrophonium to produce 95% of the peak antagonistic effect was estimated from the logit-log dose-response curve to be 0.8 plus/minus 0.33 mg *symbol* kg sup -1 (range 0.4-1.4 mg *symbol* kg sup -1). This is consistent with the observation during the study that administration of an incremental dose of edrophonium after 0.75-1.0 mg *symbol* kg sup -1 had been administered produced little further antagonism of the neuromuscular block. The time to recovery of the neuromuscular block to a TOF ratio of 0.7 after cessation of the atracurium infusion was 8.4 plus/minus 6.2 min (range 2-25 min).

This study has shown that edrophonium does not have an unlimited capacity to antagonize an atracurium-induced neuromuscular block. There is a maximum depth of neuromuscular block that can be antagonized to effect adequate recovery. With a greater block, a ceiling effect for antagonism by edrophonium is observed when a recovery plateau eventually occurs after additional doses of edrophonium. Irrespective of the dose of edrophonium administered, the neuromuscular block cannot be antagonized to effect adequate recovery. This finding is consistent with in vitro experiments, in which a ceiling effect for antagonism of a pancuronium neuromuscular block by edrophonium was observed in a rat phrenic nerve-diaphragm preparation. Supraclinical concentrations of edrophonium were unable to effect adequate recovery when the pancuronium block was at 95% depression of the first twitch response. [13]Theoretically, a ceiling effect is not unexpected, as edrophonium is not a direct antagonist of competitive neuromuscular blocking agents. It acts by increasing the concentration of acetylcholine in the end-plate region of the muscle, by inhibiting the breakdown of acetylcholine by acetylcholinesterase. However, the concentration of acetylcholine that can be achieved is limited, as only a finite amount of acetylcholine is released, and elimination can occur by alternative pathways, such as reuptake and diffusion into surrounding tissues. [13,14]At higher concentrations, anticholinesterases also may interact with the ion channel of cholinergic receptor, causing channel block and receptor desensitization. [15].

The level of neuromuscular block above which even large doses of edrophonium did not effect adequate recovery was as low as 60% depression of the first twitch height. The logistic regression model estimated that a neuromuscular block of 67% depression of the first twitch height had a probability of 0.5 for antagonism to a TOF ratio greater or equal to 70%. In practical terms, the maximum depth of block that can be antagonized approximately corresponds to the reappearance of the fourth response to TOF stimulation. [16,17]Verotta, in the only other estimate that has been made in humans, suggested a substantially greater level of vecuronium neuromuscular block could be antagonized by edrophonium. Extrapolation by a complex pharmacokinetic-dynamic model suggested that adequate recovery after edrophonium would occur in two of three patients with a block of 90% depression of the first twitch height but zero of two patients at 95% depression of the first twitch height. The model was derived from a study of the time course of effect of edrophonium on a constant level of vecuronium neuromuscular block. [18]However, clinical studies of reversal by edrophonium tend to support our finding that only a lesser neuromuscular block can be antagonized. The reversal time of a vecuronium neuromuscular block at 90% depression of the first twitch height is usually greater than 10 min, even when edrophonium doses of 1.5 mg *symbol* kg sup -1 are administered. [8]If the neuromuscular block could have been antagonized, reversal should have occurred within 2-3 min, as edrophonium has a rapid onset of action. Reversal times of this order are achieved only for reversal of neuromuscular block from vecuronium and the other competitive neuromuscular blocking agents if edrophonium is administered when four responses are present to TOF stimulation. [17-19]The reversal times of deep neuromuscular block by edrophonium are highly variable for the different neuromuscular blocking agents but are consistent with the different agents having a similar maximum depth of block, which can be antagonized by edrophonium. [5-9]The different reversal times are a reflection of the different rates of spontaneous recovery of the neuromuscular blocking agents. [11].

The answer to the question of how deep atracurium-induced neuromuscular blockade, and probably that due to other neuromuscular blocking agents, should be reversed by edrophonium can be resolved. If edrophonium is administered when there are one or two responses to TOF stimulation, antagonism will be incomplete, and adequate recovery will occur only after a decrease in the concentration of the nondepolarizing neuromuscular blocking agent. A similar decrease in the concentration of the neuromuscular blocking agent would occur if the administration of edrophonium was delayed until there are four responses to TOF stimulation. At this level of block, edrophonium will be able to antagonize the block either to adequate recovery or to a level where only a further small decrease in the concentration of neuromuscular blocking agent is required. The overall reversal times should be similar whenever edrophonium is administered. It is probably safer to wait until the block has spontaneously recovered before administering edrophonium, as the time taken for the decrease in the concentration of the neuromuscular blocking agent is highly variable. Patients administered edrophonium when the block is deep will be partially paralyzed awaiting adequate recovery, which even for an atracurium and vecuronium neuromuscular block, can take 20 min or longer. [7-9]These patients may be put at risk from tracheal extubation, because it is difficult to judge when adequate recovery has occurred even when a peripheral nerve stimulator is used. [20,21].

We found that the maximum antagonistic effect on an atracurium-induced neuromuscular block is achieved with a mean edrophonium dose of 0.8 mg *symbol* kg sup -1. This may be a slight overestimate of the "true" dose because of the cumulative dose-response technique used. [22]The dose that produces maximum antagonism should be considered the optimal dose of edrophonium if it is administered when the neuromuscular block is around the maximum depth that can be antagonized, at the reappearance of the fourth responses to TOF stimulation. This recommendation is consistent with the observations of Kopman that 0.75 mg *symbol* kg sup -1 edrophonium reliably produced adequate recovery of a vecuronium and an atracurium-induced neuromuscular block within 2-4 min if administered when there were four palpable responses to TOF stimulation, whereas after 0.5 mg *symbol* kg sup -1, recovery took up to 10 min. [17]If edrophonium is administered at lesser levels of block, when four responses to TOF stimulation are present, smaller doses may be effective. However, under these conditions, it is impossible to quantify the precise level of block by simple clinical means or to estimate the dose of edrophonium required. It is probably safest to administer the maximum antagonistic dose of edrophonium; clinical doses of edrophonium under these conditions do not induce neuromuscular block. [23]If edrophonium is administered when the neuromuscular block is deep and cannot be antagonized to adequate recovery, at one or two responses to TOF stimulation, the dose must be increased to ensure maintenance of the maximum antagonistic effect while the concentration of the neuromuscular blocking agent decreases. As previously noted, this can take 20 min or longer, so edrophonium dosage under these circumstances may have to be increased to 1.5 mg *symbol* kg sup -1 or greater or repeat doses administered. [8,19].

The level of neuromuscular blockade that can be antagonized to effect adequate recovery is critically dependent on the definition of recovery. We chose a TOF ratio of 0.7 in adductor pollicis, because this is the traditional endpoint recommended in clinical practice. At this level of residual block, there is normal ventilatory function and usually no evidence of clinical weakness in other important muscle groups. [24]Jones has suggested that a TOF ratio of 0.5 is compatible with safe recovery after an atracurium neuromuscular block, and this would correspondingly increase the maximum depth of block that can be antagonized. [25]However, at a TOF ratio of 0.5, Engback found evidence of significant muscle weakness with no patient able to sustain a 5-s head lift. [26]Alternatively, there is some evidence to suggest that the TOF ratio should be 0.8-0.9 to guarantee recovery, particularly with respect to the muscles involved in swallowing. These appear to be sensitive to nondepolarizing neuromuscular blocking agents, as occasionally, awake volunteers administered small doses that only reduce the TOF ratio to 0.9 will have difficulty in swallowing. The electromyogram of the suprahyoid muscles are decreased 44% when the TOF ratio in adductor pollicis is 0.81. [27].

In summary, this study found that there is a ceiling effect for antagonism of an atracurium-induced neuromuscular block by edrophonium in humans. The maximum depth of block that can be antagonized by edrophonium to effect adequate recovery occurs approximately when the fourth response to TOF stimulation reappears. Clinical studies of reversal by edrophonium for the other nondepolarizing neuromuscular blocking used in clinical practice are consistent with similar findings. As a consequence, reliable rapid reversal by edrophonium of an atracurium-induced neuromuscular block, and probably that from other nondepolarizing neuromuscular blocking agents, can be achieved only if edrophonium is administered when there are four responses to TOF stimulation.