Clinical Pharmacology of GW280430A in Humans

The Institutional Review Board of Weill Cornell Medical Center (New York, New York) approved this open-label ascending-dose study of GW280430A administered by bolus injection into healthy male volunteers aged 18–59 yr, within 30% of ideal body weight, during nitrous oxide–oxygen–opioid–propofol anesthesia. A total of 31 subjects participated in this study after giving informed consent. The first part (groups A and B) was performed to determine an approximate ED⁹⁵dose in 11 subjects. Subject 1 was to be given an initial dose of 0.02 mg/kg followed by an estimated dose that would produce 50% block (ED⁵⁰). A lack of response from the initial 0.02 mg/kg dose would lead to a doubling of each successive dose until some twitch suppression was observed. The ED⁵⁰dose was based on a log-probit analysis with a historic slope estimate of 7.0 (based on a log-probit transformation, where the dose was log transformed and the response was transformed with the probit function). Subject 2 was to be given three doses of estimated ED²⁵, ED⁵⁰, and ED⁷⁵. Subject 3 was to be given doses of estimated ED⁵⁰, ED⁷⁵, and ED⁹⁰. Group B, subjects 4–11, received doses of estimated ED²⁵, ED⁵⁰, ED⁷⁵, and ED⁹⁰. Data were sequentially reviewed and summarized at the completion of each group. An estimated ED⁹⁵was generated on site with SAS 6.08 software (SAS Institute Inc., Cary, NC) after the completion of the group A subjects, 1–3. A more accurate estimate of ED⁹⁵was generated after the completion of group B subjects, 4–11, with WinNonLin software (SAS Institute Inc.). The second part (groups C–G) was performed to evaluate the safety and pharmacodynamics of ascending multiples of ED⁹⁵doses in 20 volunteers, including reversal in one group (G) (table 1). The pharmacodynamic profile of GW280430A was determined at the adductor pollicis during the study.

Anesthesia was induced intravenously with midazolam (0.01–0.05 mg/kg), fentanyl (2–8 μg/kg), and propofol (2.0–2.5 mg/kg) and was maintained with nitrous oxide (70%), oxygen (30%), and propofol (50–200 μg · kg⁻¹· min⁻¹) in all subjects. Electrocardiogram, pulse oximetry, and temperature were monitored and maintained within normal limits. Volunteers’ tracheae were intubated without use of a muscle relaxant at least 20 min before a dose of GW280430A was administered. The study drug was injected over 5 s into a rapidly running intravenous line. Blood pressure (BP; systolic, diastolic, and mean via  radial arterial line) and heart rate (HR) were recorded continuously throughout the study to assess hemodynamic safety.

The onset, duration, and depth of neuromuscular block were assessed at the adductor pollicis using standard mechanomyographic neuromuscular monitoring techniques in all parts of the study. Neuromuscular transmission at the adductor pollicis was monitored using the evoked response to electrical stimulation of the ulnar nerve at the wrist. Supramaximal twitch stimulation (single 0.2-ms square wave) was delivered via  surface electrodes to the ulnar nerve every 6.7 s (0.15 Hz). The evoked responses were recorded continuously on a polygraph via  mechanomyography using a force displacement transducer attached to the thumb. A baseline of at least 20 min duration was established before a dose of the test drug was given. On evidence of beginning twitch recovery, the stimulation mode was changed from single twitch to train-of-four (TOF) every 10 s. Subsequent doses were given after at least 30 min after full neuromuscular recovery (fourth twitch to first twitch ratio ≥ 0.9) from the previous dose. Eight volunteers were also stimulated using TOF stimulation from baseline through recovery for two doses as a comparison of the effect of mode of stimulation on onset times at the adductor pollicis. Four subjects who received two doses of 0.40 mg/kg were pharmacologically reversed with 0.5 mg/kg edrophonium with 0.10 mg/kg atropine after the second dose for comparison of recovery times.

The nonlinear model examined was a two-staged, nonlinear, repeated-measures model where the dependent variable was the T1 suppression response and the independent variable was the dose. An estimate of the ED⁹⁵dose was computed for each subject in groups A and B, from each of the individual dose–response curves, and a composite ED⁹⁵estimate was taken by averaging these ED⁹⁵estimates. The software WinNonLin was used to fit the nonlinear models.

Arterial blood samples (5 ml each) for the determination of histamine concentration in plasma were collected immediately before and at 2 and 5 min after each dose in all groups. Clinically significant histamine release was defined as a doubling from baseline and greater than 1,000 pg/ml.

Subject 1 received three doses, 0.02, 0.04, and 0.08 mg/kg, because the first dose produced no block. Subject 2 received 0.05, 0.065, and 0.10 mg/kg. Subject 3 received 0.08, 0.10, and 0.13 mg/kg. Subjects–11 received 0.06, 0.09, 0.12, and 0.19 mg/kg. The ED⁹⁵initial estimate from the nonlinear regression model using the data from the first 11 subjects was 0.18 ± 0.015 mg/kg, and this value was used to compute the doses for subjects in groups C–G. The ED⁹⁵s from the final data were actually 0.19 ± 0.014 mg/kg (table 2). Figure 1displays the average dose–response relation.

The time to onset of maximum block at the adductor pollicis ranged from 2.6 ± 0.3 min to 1.5 ± 0.3 min for doses ranging from 0.18 mg/kg (ED⁹⁵) to 0.72 mg/kg (4 × ED⁹⁵). Time to 90% block for these doses ranged from 2.1 ± 0.7 min to 1.3 ± 0.2 min (table 3). Neuromuscular block onset data using TOF stimulation was collected at the 0.30- and 0.40-mg/kg doses. This method of stimulation yielded the shorter onsets of 1.2 and 1.0 min for these doses, respectively, compared with single-twitch onset data (table 4).

The duration of neuromuscular block increased with increasing dose. Clinical and total durations were 4.7–10.1 min and 9.9–16.1 min, respectively, for doses of 0.18–0.72 mg/kg (table 5). Times to return of TOF of 70% or greater and 90%, with means and ranges for each dose, are also shown in table 5.

Fire to 95% and 25–75% recovery rates were approximately 7 and 3 min, respectively, for all doses of GW280430A studied and did not increase with increasing dose (table 6).

Four subjects who received 0.40 mg/kg GW280430A were given 0.5 mg/kg edrophonium plus 10 μg/kg atropine at 10% T1 recovery. The total duration, the time from injection to 95% T1 recovery, and the recovery time to TOF 70% were shortened to 5.7 ± 1.1 and 2.1 ± 0.6 min, respectively. These values are compared with the times from spontaneous recovery in table 7.

There was very little change from baseline for mean HR, mean BP, systolic BP, and diastolic BP in the first 5 min after the administration of the study drug up to and including a dose of 0.45 mg/kg (2.4 × ED⁹⁵). However, at a dose of 0.54 mg/kg and above, there was some evidence of a mean increase in HR. A mean decrease in mean arterial pressure of 17% was evident at a dose of 0.72 mg/kg. The duration of the BP decrease was less than 5 min and was self-limited (figs. 2 and 3). Figure 4graphically plots simultaneous HR and mean arterial pressure maximal changes for each patient within the 5 min after the dose. Most of the large changes are in the upper left-hand quadrant, suggestive of typical histamine-related effects observed at high doses.

Plasma histamine concentrations were determined at baseline and 2 and 5 min as described. One subject who received 0.54 mg/kg and three subjects of four who received 0.72 mg/kg exhibited 2-min postinjection plasma histamine concentrations that were more than a doubling of baseline value and more than 1,000 pg/ml, indicative of clinically relevant histamine release. In all of these individuals, HR and BP changes and facial flushing were associated with these increases in plasma histamine (table 8).

The need for a nondepolarizing replacement for succinylcholine prompted the clinical investigation of GW280430A after animal studies identified it as a promising candidate. This study showed that the ED⁹⁵of GW280430A in man is 0.19 mg/kg and that it has a rapid onset and ultrashort duration of action. The spontaneous recovery rate of GW280430A is rapid, predictable, and independent of dose. Edrophonium facilitated reversal, shortened duration by approximately two thirds, and was used because of its faster onset than neostigmine. An ideal ultrashort-acting nondepolarizing NMB must have an onset time of 60–90 s to allow comparably acceptable intubating conditions to succinylcholine. A clinical duration of approximately 10 min is also desirable. The doses of GW280430A that fit this onset profile are 0.45–0.54 mg/kg, with clinical durations of action less than 10 min. The absence of significant side effects at doses allowing reliable, rapid intubation with conditions similar to those of succinylcholine is considered an important criterion for its replacement. In this small sample, it seems that doses up to and including 0.45 mg/kg are free of side effects. Doses at or above 0.72 mg/kg caused transient side effects in most volunteers. Further studies must be performed to determine a reliable dose that has the potential of meeting all the necessary criteria for rapid tracheal intubation and that is free of significant side effects. Pharmacodynamically, it is likely that this dose will be in the range of 0.50 mg/kg. Because the dose of 0.54 mg/kg given in this study caused histamine release in one of four patients, the side effect profile must be elucidated further.

The doses of rapacuronium that had been used for rapid tracheal intubation, 1.5–2.0 mg/kg, yielded clinical durations of 15–20 min, similar to mivacurium. 6,7We looked at the onset and recovery of 1.5 mg/kg rapacuronium in 35 patients under identical experimental conditions and found the onset at the adductor pollicis at 0.15 Hz to be 1.6 ± 0.5 min and recovery to 90% TOF to be within 44.2 ± 11.4 min without reversal and 31.3 ± 13.9 min with reversal. 8Rapacuronium can produce decreased BP and bronchospasm, with or without concomitant histamine release. Rapid administration of rapacuronium can produce a decrease in mean arterial pressure when administered in doses from 1 to 3 mg/kg, and patients in whom bronchospasm develops over this dose range had no correlation with histamine concentrations. 9It was removed from the market because of these side effects in 2001.

Another potential advantage of GW280430A exists in its lack of accumulation as implied by the identical recovery rates independent of dose in the range of 1–4 × ED⁹⁵seen in this study. This is in contrast to rapacuronium, which, either because of its metabolite or its own pharmacokinetics, changes its time course characteristics from that of a short-acting agent to that of an intermediate-duration NMB after an infusion as short as 1 h. 10The 25–75% recovery index for GW280430A reported here, 3 min, is the shortest reported for any nondepolarizer to date. This compares favorably with the 9 min that had been reported for rapacuronium, which placed its recovery profile between the short-acting mivacurium, 7 min, and the intermediate-acting cisatracurium, 14 min. 7,10–12The time to full recovery to 90% TOF is also the shortest reported for any nondepolarizer, and this may be critically important in the ambulatory setting.

In conclusion, this study showed that this new bisquaternary compound, GW280430A, has the most promising pharmacodynamic profile to date of a nondepolarizing NMB to be considered as a replacement for succinylcholine. Doses in the 2.5–3 × ED⁹⁵range showed maximum block onsets within 90 s, clinical durations of less than 10 min, and complete recovery to TOF of 90% or greater within 15 min. Doses in this range may be able to be used when a rapid-sequence induction is necessary and securing of the airway must be attempted within 60 s. The side effects in this small study were limited in this dose range. An ideal dose, balancing the need of rapid onset with the safety of minimal to absent side effects, should be more carefully studied. Intubation studies in patients must be performed with real induction sequences as well as studies investigating continuous infusions of this drug that should allow for continuous profound neuromuscular block with a rapid recovery.

The authors thank Benjamin Duncan, Ph.D. (Statistician, GlaxoWellcome Company, Research Triangle Park, North Carolina), for his statistical analyses, and JoAnn Rush (Administrative Assistant, Department of Anesthesiology, Weill Medical College of Cornell University, New York, New York) for her technical assistance with this manuscript.