Preclinical Pharmacology of CW002: A Nondepolarizing Neuromuscular Blocking Drug of Intermediate Duration, Degraded and Antagonized by l-cysteine—Additional Studies of Safety and Efficacy in the Anesthetized Rhesus Monkey and Cat

CW002 (fig. 1), a novel neuromuscular blocking agent (NMBA), is degraded in vitro by l-cysteine adduction (t½ = 11.4 min) followed by alkaline hydrolysis.¹  Neuromuscular blockade (NMB) can be readily antagonized by exogenous l-cysteine in the Rhesus monkey¹  and dog.^2,3 

We classify the duration of CW002 (approximately 30 min in the monkey at approximately 3.75 × ED₉₅ for NMB) as intermediate because its duration is about two thirds that of cisatracurium at comparable dosage yet about 3× that of the ultrashort-acting NMBA gantacurium in that species.¹ 

CW002 was selected for additional studies of efficacy and safety. The desirable properties of CW002^1–7  are intermediate duration and rapid antagonism of NMB by l-cysteine,¹  and minimal cardiopulmonary effects for both CW002 and l-cysteine.^2,4 

Additional studies included in the current investigations have addressed the following:

• a.

  A comparison of spontaneous recovery from CW002-induced NMB versus antagonism by l-cysteine (50 mg/kg) at +1 or +5 min or at 1 to 2% twitch height during early recovery from a dose of 0.15 mg/kg, 3.75 × ED₉₅ (estimated dose producing 95% twitch inhibition), or after discontinuation of infusions 30 to 180 min long, and at +1 min after doses of 10 and 20 × ED₉₅.

• b.

  Development of dose ratios comparing ED causing 20% decrease in mean arterial pressure (MAP) and/or 20% increase in heart rate (HR) versus ED₉₅ for NMB in the monkey.^8,9 

• c.

  A comparison of changes in MAP and HR after CW002 given to monkeys as a single bolus, as the “first dose of the day,” versus administration in increments to achieve the same total doses. This comparison is pertinent since in the clinic, a large dose of NMBA is commonly given first.

• d.

  Development of dose ratios in the cat for ED₅₀ for vagal block (VB) or sympathetic block (SB) versus ED₉₅ (NMB).^8,9 

All protocols were approved by the Institutional Animal Care and Use Committees (IACUC) of Weill Cornell Medical College (New York, New York) and of Albany Medical College (Albany, New York), where the studies were conducted.

CW002 was synthesized at Cedarburg Hauser Pharmaceuticals, a division of Albany Molecular Research, Inc., Grafton, Wisconsin. CW002 was given intravenously at concentrations of 1 to 10 mg/ml in 0.9% NaCl, with pH adjusted to 3.0 with 1 N HCl to improve stability. l-cysteine hydrochloride was purchased from Ajinomoto Pharmaceuticals (USA). In our new formulation for l-cysteine, to maximize stability, vials contained 20 ml (4.0 g at 200 mg/ml) in 0.9% NaCl). The pH was adjusted and buffered to 4.5 to 5.5 using 1 N NaOH and concentrated (10×) phosphate buffer (pH 7.4). The vials were stored at −20°C and thawed before injection. Stability for 8 h, once thawed, has been shown in pilot studies (Laboratory of John J. Savarese, M.D., Weill Cornell Medical College, 2008 to 2015). Solutions of CW002, also stored at −20°C, were freshly thawed on the day of the experiment and kept in an ice bath to further ensure stability.

Eight adult male Rhesus monkeys weighing 8 to 18 kg were studied at intervals of 4 to 6 weeks. Animals were housed in accordance with the Guide for Care and Use of Laboratory Animals (National Research Council, Washington, DC). Care and maintenance of the animals have been described.¹ 

Experimental setup was performed as previously described.^1,8  Monkeys received ketamine (7 to 10 mg/kg) intramuscularly, followed by tracheal intubation under topical anesthesia with 2% lidocaine. Ventilation was controlled at V_T 12 to 15 ml/kg at 18 to 20 breaths/min. Isoflurane anesthesia (1.0 to 2.0%) in N₂O/O₂ (2 l/1 l mixture) was maintained during each experiment, and lactated Ringer’s solution was given at a rate of approximately 6 ml kg⁻¹ h⁻¹. Arterial pressure was monitored directly. HR was measured by tachograph from the arterial waveform. Temperature was kept at 36 to 38°C. Oxygen saturation was kept at more than 96%.

Needle electrodes (25 ga) were placed at the peroneal nerve. Square-wave pulses of 0.2-ms duration at supramaximal voltage were applied to the nerve at 0.15 Hz to elicit twitch responses of the extensor digitorum of the foot; 20% of the tendon was tied to a Grass FT 10 force transducer at a baseline tension of 50 g. Train-of-four (TOF) stimulation (2 Hz for 2 s) was interposed at appropriate time points, especially at 1 to 2 min before injection of CW002 and after recovery of twitch to 95% of baseline.

CW002 dose 1, the “first dose of the day,” was given after a more than 15-min stable baseline period as a rapid (5-s) bolus. Dose 2 was given at least 30 min after recovery of TOF ratio to 100% or more after dose 1. Doses 1 and 2 of CW002, given at the beginning of each experiment, were included in data describing neuromuscular blocking potency and duration and recovery. Onset was measured after dose 1 only. At the end of each experiment, analgesics were given per veterinary practice, and the animals were awakened and attended until standing.

The dose–response relationship was generated by nonlinear regression of log dose versus percentage inhibition of twitch. Doses 1 and 2 from each experiment were used to calculate the dose–response curve. ED₅₀ and ED₉₅ for twitch inhibition were derived. Onset time from injection to maximum twitch suppression (dose 1 only), total duration from injection to recovery of twitch to 95% of baseline, and the recovery interval from 5 to 95% twitch height (doses 1 and 2) were measured. All recoveries were followed until TOF ratio was at least 95%. (Most Rhesus monkeys have usually shown TOF ratio of 100 to 120% at baseline.^1,8 )

Continuous infusions of CW002 were given for durations varying from 30 to 180 min. Two groups of infusions were done: group 1: spontaneous recovery; group 2: antagonism by l-cysteine (50 mg/kg) at the end of infusion. Infusions were initiated by a bolus dose of 0.08 to 0.15 mg/kg; infusion was then begun at recovery to 25% twitch height after the initial bolus, and the infusion rate was thereafter adjusted to maintain 98 to 99% twitch suppression.

At the end of infusions in group 1, spontaneous recovery was measured to 95% twitch height and TOF more than 95%. In group 2, at discontinuation of infusion, the line was flushed with 5 ml lactated Ringer’s solution, and l-cysteine (50 mg/kg) was given 1 min later, at 0 to 2% twitch height.

Spontaneous recovery times and l-cysteine–accelerated recovery times, i.e., reversal times, were compared by assessment of the 5 to 95% twitch recovery interval and its relationship to the duration of infusion using linear regression.

Antagonism of CW002-induced NMB by l-cysteine has been shown to peak in rapidity and completeness at doses of 30 to 50 mg/kg in the Rhesus monkey and dog.^1,2  In the current studies, antagonism by l-cysteine given as a 5-s bolus was evaluated under various circumstances, always administering the same dose (50 mg/kg). Antagonism was evaluated at the following time points: at +1 min (100% block) after CW002 doses of 0.15, 0.40, and 0.80 mg/kg (approximately 3.75 × to 20 × ED₉₅); at +5 min (100% block) after 0.15 mg/kg; at the beginning of twitch recovery (1 to 2% twitch height) after 0.15 mg/kg; and at the termination of infusions in group 2, 1 min after discontinuation of infusion, at 0 to 2% twitch height.

All doses of CW002 were given as rapid (5-s) boluses. Two dose–response relationships were developed for HR and MAP.

• A.

  Single-bolus method (group 1): Data from the first dose of the day only were applied to construct the curves. Doses of 0.20 to 2.40 mg/kg (5 to 60 × ED₉₅) were given on separate occasions, spaced 4 to 6 weeks apart. Method A was employed particularly to closely mimic clinical practice where a large dose of NMBA is usually given first.

• B.

  Cumulative method (group 2): Doses from 5 × ED₉₅ to 20 × ED₉₅ (0.20 to 0.80 mg/kg) were given successively; the cumulative dose was doubled every 10 min to a total of approximately 40 × ED₉₅ or 1.60 mg/kg: 5 ×, 5 ×, 10 ×, and 20 × ED₉₅. Maximum changes from the original baseline were noted for each dose increment to yield cumulative dose–response curves.

Curvilinear dose–response curves were calculated for method A, and the EDs for 20% decrease of MAP and 20% increase of HR after single-bolus doses were derived.

Anesthesia, setup, ventilation, and monitoring were as described in previous studies in Rhesus monkeys and cats.^1,8  Six male cats weighing 4 to 6 kg were studied three times each at intervals of 2 to 4 weeks. About 20% of the tibialis anterior tendon was freed in sterile fashion and tied to an FT 10 transducer at baseline tension of 50 g. Monitoring of neuromuscular function was as described for monkeys. Data obtained were as in monkeys: duration to recovery to 95% of baseline twitch height and 5 to 95% twitch recovery interval.

Onset of NMB was not studied in the cat, nor was continuous infusion. At the end of each experiment, animals were given analgesics per veterinary practice and attended until standing. Cats were offered for adoption after completion of studies.

Studies were not as extensive as in the Rhesus monkey. Antagonism of approximately 1 × ED₉₅ (doses of approximately 0.03 mg/kg) by l-cysteine (50 mg/kg), given at +6 min after CW002 injection, was documented twice.

Inhibition of vagal (parasympathetic, VB) and sympathetic (ganglionic, SB) responses was studied in the third of the three experiments in four of the six cats. The protocol has been described.⁸ 

The right cervical vagus nerve and sympathetic trunk were exposed under sterile technique and tied off centrally. The nerves were placed on separate electrodes and were stimulated with 10-s trains of square waves (0.5 m at 20 Hz) every 3 to 5 min. Responses recorded were as follows: VB—inhibition of a baseline response of about 50% reduction of HR; SB—block of the contraction of the right nictitating membrane as transduced by a Grass FT-03 instrument.

Dose–response curves were constructed in cumulative fashion: 0.10, 0.10, 0.20, and 0.40 mg/kg boluses of CW002 were given sequentially every 10 min. The total dose studied was therefore 0.80 mg/kg (approximately 23 × ED₉₅ in the cat). Nonlinear regression of dose versus percentage blockade of the autonomic responses was performed to estimate ED₅₀ for VB and SB.

From studies in the cat, the following were calculated as indicators of the relative likelihood of VB or SB versus NMB potency:

Similarly, from studies in monkeys (see Circulatory Observations), the following were calculated as indicators of the relative likelihood of pertinent deleterious circulatory changes versus NMB potency.

The program Microsoft Excel 2007 (Microsoft Corporation, USA) generated statistical comparisons via Student’s t test (two tailed) or the Tukey–Kramer test (for multiple comparisons).

Noninear regression of the percentage twitch inhibition or block of autonomic responses versus log dose was performed using the program GraphPad Prism version 6.0 (GraphPad Software, USA) to generate dose–response curves for neuromuscular or vagal blockade (NMB or VB).

Linear regression was used to evaluate the relation of speed of recovery, spontaneous or accelerated by exogenous l-cysteine, as indicated by the 5 to 95% twitch recovery interval, versus the duration of infusion.

SigmaPlot^® from Systat Software, Inc. (USA) was used to construct curvilinear regressions for HR and MAP changes in monkeys after single-bolus doses given as the “first dose of the day.” P < 0.05 was considered significant in all statistical comparisons. Group sample sizes of n = 4 or more were ensured before any statistical comparison. Exact P values are given whenever possible (for two-tailed Student’s t test).

All data are original and have not been previously published. Data are reported as mean (± SD) or as mean with 95% CI.

The calculated ED₅₀ and ED₉₅ of CW002 for NMB (fig. 2; n = 33) were 0.020 ± 0.008 and 0.040 ± 0.015 mg/kg, respectively.

After “first doses of the day” (n = 5) of 0.03 to 0.06 mg/kg, which produced 95 to 99% block, onset time was 195.0 (18.7) s, total duration to recovery of twitch to 95% of baseline was 17.7 (5.3) min, and the 5 to 95% recovery interval was 12.4 (3.6) min (table 1). After 0.15 mg/kg (approximately 3.75 × ED₉₅), onset was 81.4 (25.3) s, duration was 25.9 (7.0) min, and the recovery interval was 11.5 (4.6) min (n = 7). Larger doses shortened onset time at up to approximately 10 × ED₉₅ (0.40 mg/kg), where onset was 37.5 (5.9) s; doses more than 0.40 mg/kg did not shorten onset time significantly (table 1).

Spontaneous recovery intervals (5 to 95% twitch) were not different (P > 0.05 by Tukey–Kramer test) over the range 0.05 to 1.60 mg/kg (approximately 1.25 × to approximately 40 × ED₉₅); this interval did not differ significantly during spontaneous recovery after infusions in group 1. Intervals ranged from 11.5 to 16.1 min (tables 1 to 3 and fig. 3).

Data are summarized in tables 2 and 3 and figure 3.

In the two groups of experiments (group 1: spontaneous recovery, n = 17, and group 2: l-cysteine antagonism, n = 8), the rates of CW002 administration required to maintain blockade at 98 to 99% twitch inhibition did not differ: 3.60 (0.9) μg kg⁻¹ min⁻¹ (group 1) and 3.20 (0.7) μg kg⁻¹ min⁻¹ (group 2), P = 0.32. Infusion duration did not differ (group 1, 103.7 [44.2] min vs. group 2, 90.4 [21.6] min), P = 0.43.

Antagonism of NMB at end infusion by l-cysteine, 50 mg/kg (group 2), markedly shortened 5 to 95% recovery time versus spontaneous recovery: 2.5 (0.4) min versus 13.5 (4.0) min (P < 0.0001; table 2).

Infusion duration had no influence on the rate of spontaneous recovery (5 to 95% recovery interval) or on the accelerated (shortened) recovery induced by l-cysteine (fig. 3); P = 0.43 and 0.28, respectively.

The data are summarized in tables 2 and 3. l-cysteine caused rapid recovery of twitch from 5 to 95% of baseline within 1.8 to 3.6 min when it was administered 1 min after doses of CW002 ranging from approximately 3.75 × to 10 × ED₉₅ (0.15 to 0.40 mg/kg), at which time there was always 100% block of twitch. At two other time points after injection of CW002 (0.15 mg/kg), namely at +5 min and at 1 to 2% twitch height at the beginning of recovery, antagonism by l-cysteine was equally rapid (table 3). Similarly, l-cysteine given at 0 to 2% twitch height, at +1 min after discontinuation of infusions, accelerated the recovery of twitch from 5 to 95% of baseline strength from 13.5 (4.0) min to 2.5 (0.4) min (P < 0.0001), during spontaneous versusl-cysteine–accelerated recovery (table 2 and fig. 3).

CW002 was slightly more potent in the cat than in the Rhesus monkey and was longer lasting (table 1). ED₉₅ was 0.035 ± 0.01 mg/kg.

After doses inducing 95 to 99% block (approximately 0.03 mg/kg), the duration was 39.7 (8.8) min (n = 7); after 0.10 mg/kg (approximately 3.0 × ED₉₅), the duration was 57.2 (18.1) min (n = 5). The 5 to 95% recovery intervals after doses of approximately 0.03 and 0.10 mg/kg were 27.5 (10.8) versus 21.0 (6.4) min, n = 7 versus n = 5. They did not differ significantly (table 1), P = 0.259.

When given at 6 min after CW002 (approximately 0.03 mg/kg), which induced mean 99% block, l-cysteine (50 mg/kg) accelerated the 5 to 95% recovery interval from 27.5 (10.8) min (n = 7) to a mean of 5.8 min (table 3, n = 2).

A comparison was made of the circulatory effects of cumulative administration versus single-bolus doses, the latter given as the “first dose of the day” (table 4 and figs. 4 and 5). Circulatory data shown in tables and figures are maximum changes from baseline measurements. Any changes noted nearly always occurred within 5 min or less.

Curvilinear regressions done for changes of MAP and HR after the “first dose of the day” show that after the doses of 0.80, 1.60, and 2.40 mg/kg (20, 40, and 60 × ED₉₅), increases in HR were dose related, while a decreasing trend was seen for MAP (figs. 4 and 5). On the other hand, a decreasing trend for both HR and MAP was seen during cumulative administration (fig. 4). At 0.80 and 1.60 mg/kg, differences in responses of MAP versus baseline were significant after bolus (P = 0.049 and 0.001) but not after cumulative doses (P = 0.125 and 0.096). A table of P values is attached to figure 4 and table 4. The ED₂₀ with 95% CI for HR increase after the “first dose of day” boluses was 2.16 (1.61 to more than 2.40 mg/kg); the ED₂₀ for MAP decrease was 1.06 (0.94 to 1.21) mg/kg (fig. 5). The dose ratios for ED (20% change in MAP or HR) versus ED₉₅ for NMB were 54 × ED₉₅ for increase in HR and 27 × ED₉₅ for decrease in MAP.

Data are summarized in table 5. ED₅₀ (VB) was 0.59 ± 0.07 mg/kg; ED₅₀ (SB) was >>0.80 mg/kg. Dose ratios [ED₅₀ (VB)/ED₉₅ (NMB)] and [ED₅₀ (SB)/ED₉₅ (NMB)] were 17 × ED₉₅ and >> 23 × ED₉₅ for the vagal and sympathetic systems.

The studies reported herein were undertaken to assemble a pharmacologic profile of CW002; they provide additional data on efficacy: on antagonism of 99 to 100% twitch inhibition by the same dose (50 mg/kg) of l-cysteine under a variety of circumstances and on the pharmacology of continuous infusion of CW002. The minimal circulatory changes after CW002 in large doses (5 to 10 × ED₉₅ for NMB) provide information assuring an adequate safety margin in the NMB dose range. An accidental overdose of CW002 (as large as 20 to 60 × ED₉₅ for NMB), as suggested in table 1, would likely lengthen its NMB effect less than proportionately due to increased reaction kinetics (the mass action effect), a consideration of both efficacy and safety.

Documentation herein of the circulatory and autonomic side effects of very large doses of CW002 (20 to 60 × ED₉₅) not only provides increased assurance of safety, but also suggests upper dose ranges for future evaluations of potential histologic, biochemical, or hematologic pathologic effects.

Definition of dose ratios relating the ED₅₀ for autonomic blockade in the cat divided by the neuromuscular blocking potency (ED₉₅) requires the determination of the ED₉₅ in that species. Similarly, dose ratios for circulatory changes of 20% decrease in MAP and increase in HR versus ED₉₅ in the monkey were generated as indicators of safety (table 4 and figs. 4 and 5). Obviously, higher dose ratios suggest greater safety, i.e., less likelihood of occurrence of side effects.

The numbers of individual animals studied seem appropriate since the numbers are normative to our previously published work in monkeys and cats.^1,8  The dose range chosen for study of CW002 represents the range of potency of published¹  and unpublished compounds (Laboratory of John J. Savarese, M.D., Weill Cornell Medical College, 2008 to 2015). That CW002 has a duration of NMB that is typically “intermediate” in the Rhesus can be inferred from a crossover study showing a ratio of 2:3 of comparative duration of CW002 versus cisatracurium.¹ 

We are reluctant to compare speed of onset of NMB after CW002 versus other NMBAs in the monkey model without our own data. In our unpublished observations (Laboratory of John J. Savarese, M.D., Weill Cornell Medical College, 2008 to 2015), rocuronium showed approximately 40% the potency of CW002 (ED₉₅ approximately 0.10 vs. 0.040 mg/kg) and a longer duration of action at approximately 4 × ED₉₅; the onset of effect of CW002 in our monkeys seems longer than the onset of effect of rocuronium and shorter than that of cisatracurium. Onsets of CW002 given in table 1 are intended to show at which dose (0.40 mg/kg) a limit is reached, presumably due to circulation time, and do not imply comparisons with other NMBAs.

Comparisons of onset of effect of NMBAs in humans or other species must require well-controlled anesthetic conditions, measuring onsets obtained only after “first doses of the day.”

l-cysteine adduction (fig. 1) initiates a degradation cascade ending in molecular fragments that show only about 0.01 to 0.001 × the potency of CW002 to produce NMB.¹  The key feature of the profile of CW002, antagonism of its NMB effect by exogenous l-cysteine, most likely occurs because the chemistry of degradation is accelerated markedly by a mass-action effect.¹ 

The peak effect of l-cysteine, in terms of speed and completeness of reversal, is reached at doses of 50 mg/kg^1,2 ; therefore, that dose was given under all conditions reported in this article: l-cysteine was found equally effective at +1 min after CW002 doses of 0.15 to 0.40 mg/kg (approximately 3.75 to approximately 10 × ED₉₅), at +5 min after 0.15 mg/kg, and at 1 to 2% twitch height at the beginning of recovery from a dose of 0.15 mg/kg or from infusions. The 5 to 95% recovery interval was shortened to 1.8 to 3.6 min under all circumstances and did not differ significantly (tables 2 and 3).

Because of its mechanism of reversal, i.e., inactivation and destruction of the NMBA molecule, l-cysteine antagonism of deep CW002-induced NMB is rapidly effective, as shown herein (tables 2 and 3). The competitive mechanism of neostigmine is well understood, on the other hand, to be the reason for its lack of effectiveness in antagonism of deep block caused by other NMBAs⁹  as well as CW002.¹ 

l-cysteine concentrations in plasma remain constant throughout adult life¹⁰ ; sources are the diet, cleavage of the –S–S– bond of the dimer (l-cystine), yielding two molecules of l-cysteine, and enzymatic conversion of glutathione. The metabolic pathways sourcing l-cysteine are bidirectional, depending on the local redox state of tissue.¹¹ 

A toxicokinetic study in dogs of l-cysteine (data from Laboratory of John J. Savarese, M.D., Weill Cornell Medical College, 2008 to 2015) reports a plasma half-life of 30 to 40 min, relatively short, possibly because of uptake into various metabolic pools.¹¹ 

The plasma half-life of l-cysteine in humans has been estimated at 30 to 60 min.¹¹  The plasma half-life of sugammadex is 2 to 4 h in normal subjects.¹²  These t½ are pertinent to clinical practice because such agents, which effectively acutely lower the plasma concentration of the NMBA, will show a residual effect, after reversal, of inhibiting the activity of subsequent doses of the NMBA. Such doses given after the reversal agent, to reestablish paralysis and clinical relaxation, would be inhibited for a length of time proportional to the t½ in plasma of the reversal agent.

Decreased availability of endogenousl-cysteine in practice, such as in malnutrition or cachexia, might slow spontaneous recovery from CW002-induced NMB; the remedy would be administration of exogenous l-cysteine.

Both l-cysteine adduction and Hofmann degradation are pH dependent.^1,13,14  In Hofmann degradation, hydroxyl ion concentration in living tissue is limited to the narrow range compatible with life and cannot be altered acutely to accelerate recovery from atracurium/cisatracurium; exogenous l-cysteine, however, presumably induces reversal of CW002 by an acute increase in ambient l-cysteine concentration in plasma, markedly accelerating the conversion of CW002 to its inactive adduct (fig. 1).

Neostigmine is ineffective in shortening the timeframe from CW002 injection to complete recovery from CW002-induced blockade in monkeys when given 1 min after 0.15 mg/kg (approximately 3.75 × ED₉₅).¹  An analogous failure was reported when early administration of neostigmine to humans after intubating doses of vecuronium¹⁵  did not shorten the time from injection of vecuronium to complete recovery. This is basis for the commonly advised clinical practice of waiting (ideally) for at least two twitches to be perceptible on TOF stimulation during recovery from nondepolarizing block before neostigmine administration.⁹ 

Antagonism of rocuronium-induced blockade may be achieved early, and complete recovery shortened, with sugammadex, by increasing the dose from 2 mg/kg when TOF count is 1 to 4^16,17  to 8 to 16 mg/kg when TOF count is 0.^18–20  Cost may be a consideration in choosing this option.

The new cucurbituril derivative calabadion, like sugammadex, probably complexes with the NMBA (rocuronium or cisatracurium) in antagonizing NMB,²¹  and so may also require higher dosage for early reversal.

The same dose of l-cysteine (50 mg/kg) is equally effective in antagonism of CW002-induced NMB at any time after CW002 administration; as shown herein, the interval from 5 to 95% recovery of twitch does not differ significantly under a variety of circumstances during reversal of doses of CW002 from 3.75 to 10 × the ED₉₅ for NMB in monkeys (tables 2 and 3). Hypothetically, issues of both convenience and economics might also be pertinent in future practice with the CW002/L–cysteine combination.

The wide separation of the vagal blocking effect from the NMB property suggested in the dose ratio (17 × ED₉₅, table 5) is desirable. By comparison, the dose ratio of pancuronium in the cat for VB versus NMB is only about 2.0 to 3.0 × ED₉₅ for NMB, and pancuronium does accelerate HR in clinical practice.^22,23  Rocuronium shows a dose ratio for VB versus NMB in the range of 5.0 to 7.0 × ED₉₅ in the cat²⁴  and may cause an increase in HR in humans after doses of 3 to 4 × ED₉₅ given for tracheal intubation.²⁵  The dose ratios for VB versus NMB of cisatracurium and vecuronium, which do not accelerate HR,⁹  are higher: 25 and 76 to 80.^23,26 

The above dose ratios, obtained from small numbers of experiments, customary in such studies,^8,23,24,26  should be considered approximate, but may suggest increases in HR in humans,^22,25  when the ratio [ED₅₀ (VB)/ED₉₅ (NMB)] is less than 5.

M2 block in the airway, combined with allosteric sensitization of M3 receptors, may cause bronchospasm. Rapacuronium was withdrawn from clinical practice likely due to this mechanism,^27,28  demonstrated in a guinea pig model.^29–33  CW002 does not block M2 receptors in the guinea pig.⁴ 

CW002 has a high dose ratio (>>23 × ED₉₅ for NMB) versus inhibition of nicotinic ganglionic responses. In fact, this mechanism is really no longer a consideration, in terms of safety, and has been “engineered” out of current NMBAs.^9,23,26 

Large doses of CW002 (approximately 20 or more × ED₉₅, or 0.80 to 2.40 mg/kg), given to monkeys as the “first dose of the day,” result in dose-related decreases in MAP and increases in HR, which are not noted after incremental (cumulative) administration (table 4 and figs. 4 and 5). After bolus injection of 2.4 mg/kg (approximately 60 × ED₉₅), flushing was noted in each of three monkeys, but was not observed during cumulative administration. This contrast has also been observed in the dog, where plasma histamine levels increased after 0.40 to 0.60 mg/kg (approximately 40 to 60 × ED₉₅) only after single-bolus doses; decreases in MAP and increases in HR were also noted (unpublished, Paul M. Heerdt, M.D., Ph.D., Weill Cornell Medical College, 2010 to 2013).

Before clinical trials, protocols evaluating the potential of NMBAs to cause circulatory and/or histaminoid changes in dogs or monkeys should measure the effects of single-bolus doses given as the “first dose of the day” to better mimic future clinical practice, where a large dose of NMBA is commonly given first; cumulative administration may underestimate the circulatory effects of NMBAs later observed in humans.

For example, in dogs, gantacurium caused histaminoid responses during cumulative administration at about 12.5 to 25 × ED₉₅³⁴ ; in monkeys, mivacurium and gantacurium elicited such responses at approximately 12 and 50 × ED₉₅, respectively, when given cumulatively.⁸  After “first doses of the day” given to humans, histaminoid changes resulted after 2.5 × ED₉₅ for mivacurium (0.20 mg/kg), and 3 to 4 × ED₉₅ for gantacurium (0.50 to 0.70 mg/kg).^35,36 

Additional properties of the novel NMBA CW002, including its rapid reversal by l-cysteine under various circumstances and its relative lack of circulatory or autonomic effects, are described herein. The data further support its safety and efficacy.

The authors thank and appreciate the chemists, Jeff D. McGilvra, Ph.D., Cedarburg Pharmaceuticals, A Division of Albany Molecular Research, Inc., Grafton, Wisconsin; and Scott G. Van Ornum, Ph.D., Department of Chemistry, Concordia University, Mequon, Wisconsin. They also thank Tammy Smith-Stewart, B.S., Department of Anesthesiology, Weill Medical College of Cornell University, New York, New York; Farrell Cooke, B.S., Department of Anesthesiology, Weill Medical College of Cornell University; and Bryce Petty, B.A., Department of Anesthesiology, Weill Medical College of Cornell University, for producing the manuscript and figures and double-checking the statistics.

Supported in part by the Nancy Paduano Investment Retirement Account, New York, New York; C.V. Starr Foundation, New York, New York; and Department of Anesthesiology, Weill Medical College of Cornell University, New York, New York.

Drs. Savarese and McGilvra declare inventorship. The other authors declare no competing interests.