Cysteine Reversal of the Novel Neuromuscular Blocking Drug CW002 in Dogs: Pharmacodynamics, Acute Cardiovascular Effects, and Preliminary Toxicology

• ❖ The experimental neuromuscular blocking agent CW002 can be reversed by systemic administration of l-cysteine

• ❖ l-Cysteine can be neurotoxic at high doses in some species

• ❖ In dogs, the optimal dose of l-cysteine for reversal of a large dose of CW002 is 50 mg/kg

• ❖ At four times this dose, l-cysteine had no acute organotoxicity in dogs

RECENT studies of the neuromuscular blocking drug gantacurium have demonstrated an ultra-short duration of action largely due to rapid inactivation of the molecule by endogenous l-cysteine.1–3This chemical reaction occurs without an enzymatic catalyst at pH 7.4 and entails adduction of cysteine to the chlorofumarate component of the gantacurium molecule. The unique pharmacology of gantacurium has led to the development of related compounds with an intermediate duration of action that can be rapidly reversed by intravenous injection of l-cysteine, even in the absence of any spontaneous muscle recovery.4# One of these compounds is CW002, a nonhalogenated, symmetrical benzylisoquinolinium fumarate diester, which exhibits an intermediate duration and reversal by l-cysteine in rhesus monkeys and dogs.5,6 

To date, the optimal intravenous l-cysteine dose for reversal of CW002 as well as any acute cardiovascular effects and systemic organ toxicity associated with the relevant dose range remain unclear. Toxicological evaluation in rats has indicated that the l-cysteine dose required to produce death in half of the study subjects (LD⁵⁰) is 1.89, 1.62, 1.55, and 1.14 g/kg when administered orally, intraperitoneally, subcutaneously, and intravenously, respectively.7Although there are data indicating that l-cysteine can be neuroprotective,8,9when given to rats in doses close to the LD⁵⁰, l-cysteine has been deemed to be neurotoxic, both in neonatal animals with an incomplete blood–brain barrier10and in adults.11Other data have raised the prospect that sustained excess of l-cysteine within the central nervous system may be neuroexcitatory via  a number of mechanisms and contribute to neurodegenerative processes.12Manifestations of acute neuroexcitation by l-cysteine injected directly into the brain or cerebrospinal fluid of rats include an increase in blood pressure modulated by both sympathetic nervous activation and vasopressin release.13 

However, it remains unclear how studies in rodents using l-cysteine injected directly into the brain or administered in high systemic doses relate to the effects of smaller intravenous doses in large animals or human. l-Cysteine, a nonessential amino acid in adults, is conditionally essential in preterm infants because of enzymatic immaturity of the transsulfuration pathways that convert methionine to cysteine.14As such, l-cysteine has been commonly infused intravenously at doses up to 80 mg · kg⁻¹· day⁻¹to supplement parenteral nutrition in preterm infants.15In both adults and children, the acetylated derivative (N -acetyl l-cysteine) has been administered to treat a wide variety of conditions (organ ischemia/reperfusion, hepatorenal injury, cocaine and amphetamine toxicity, and psychiatric illness) in which increased glutathione production and augmented antioxidant capacity are felt to be potentially beneficial.16–21Currently, both oral and intravenous N -acetyl cysteine preparations (Acetdote) are approved by the Food and Drug Administration for use in the treatment of acute acetaminophen toxicity.22 

This study in dogs was designed with two goals. The first goal was to provide information about the optimal l-cysteine dose range for CW002 reversal and to broadly characterize any acute or sustained side effects. To accomplish this goal, the dose–response relationship for l-cysteine reversal of a fixed CW002 dose was determined from multiple l-cysteine injections at weekly intervals. These data were then interpreted in the context of both acute dose-related effects on blood pressure and heart rate (HR), and any physical or behavioral abnormalities evident in the interval between doses. The second goal was to more closely characterize the systemic toxicity of l-cysteine relative to the defined therapeutic dose range. This goal was accomplished by reversing a fixed CW002 dose with an amount of l-cysteine four times greater than the derived optimal dose and then observing the dogs for 2 or 14 days before necropsy and biochemical and histopathologic evaluation.

Six adult male beagles weighing between 9.1 and 12.5 kg were used for the study after approval of the protocol by the Institutional Animal Care and Use Committee of Weill Cornell Medical College (New York, New York). Each animal was studied four times at least 7 days apart, with a single l-cysteine dose evaluated during each procedure. For half the animals, the dose sequence was 10, 20, 50, and 100 mg/kg, and for the other half, it was 100, 50, 20, and 10 mg/kg.

After an overnight fast, each dog was sedated with a subcutaneous injection of 0.1 mg/kg acepromazine, an intravenous catheter was inserted in a foreleg vein, and anesthesia was induced with 4–5 mg/kg intravenous propofol. The trachea was then intubated, and the lungs were ventilated with a mixture of 70% nitrous oxide, 28% oxygen, and 1.5–2.0% isoflurane. Stability of anesthesia was assessed by using a processed electroencephalogram (CSM-2; Danmeter a/s, Denmark). Ventilation was adjusted to maintain end-tidal carbon dioxide at 30–32 mmHg, and systemic arterial oxygen saturation was continuously monitored. Body temperature was measured with a rectal probe and maintained at 37°–38°C with a water-circulating heating blanket. Estimated fluid deficits and maintenance needs (5 ml⁻¹· kg⁻¹· h⁻¹) were provided with intravenous lactated Ringer's solution throughout the experiment. After placement of electrocardiographic leads, a femoral artery was percutaneously cannulated for the measurement of systemic arterial pressure. The electrocardiogram, HR, and mean arterial pressure (MAP) were continuously recorded in both analog and digital format throughout the experiment. A superficial segment of the tendon of the left tibialis anterior muscle was attached to a force transducer, preloaded with 50 g tension, and single twitch responses were elicited at 0.15 Hz with supramaximal square-wave pulses of 0.2 ms duration via  stimulation of the peroneal nerve. Train-of-four stimulation, 2 Hz for 2 s, was interposed during the experiment whenever appropriate.

Powdered CW002 was synthesized by Cedarburg Laboratories (Grafton, WI). l-Cysteine was purchased from Sigma-Aldrich (St. Louis, MO). Cysteine solution was prepared by dissolving the l-cysteine base in saline to a concentration of 100 mg/ml, which provided a solution with a pH of approximately 5.5.

After surgical preparation, a 15- to 20-min stabilization period was allowed before initiating the protocol. During this time, end-tidal isoflurane concentration was allowed to equilibrate at approximately 1.3%, and consistency of twitch height was confirmed. CW002 was then administered as an intravenous bolus of 0.08 mg/kg (CW002 alone), a dose reflecting approximately nine times the ED⁹⁵for the beagle.5Fifteen minutes after complete recovery of twitch height and verification of normal train-of-four (defined as the fourth twitch being equal in amplitude to the first), the same dose of CW002 was repeated followed 1 min later by intravenous injection of 10, 20, 50, or 100 mg/kg l-cysteine (CW002 with cysteine). Fifteen minutes after complete recovery of twitch after the second dose, 0.08 mg/kg CW002 was given a third time (CW002 after cysteine) to determine the effect of residual l-cysteine on CW002 pharmacodynamics. For each of the three CW002 injections, the duration of action (defined as the time from injection of CW002 to 95% recovery of twitch height) and recovery indices (defined as the time from 5 to 95% recovery of twitch height and the time from 5% recovery of twitch height to normal train-of-four) were determined. After return of neuromuscular function, the animals were awakened and the trachea was extubated, and activity was monitored for clinical signs of recurrent muscle weakness over the next 2 h. Between l-cysteine reversal experiments (separated by at least 7 days), dogs were observed daily for frank moribundity as well as changes in activity, behavior, and appetite by animal care personnel blinded.

To more fully assess the cardiac effects of high-dose l-cysteine, after completion of the full protocol in four animals, the chest was opened and additional instrumentation was placed for measurement of left ventricular (LV) volume and pressure (conductance/micromonometer catheter; Millar Medical Instruments, Dallas, TX) along with ascending aortic blood flow (electromagnetic flow probe; Carolina Medical, King, SC). From these data, stroke volume, systemic vascular resistance, and LV ejection fraction were recorded. In addition, LV inotropy was assessed as dP/dt normalized to LV end-diastolic volume, and lusitropy indexed from the time constant of isovolumic relaxation. For each experiment, 1 ml/kg of saline, the vehicle for l-cysteine, was first injected via  peripheral vein to evaluate the effect of volume alone. Fifteen minutes later, the same amount of saline containing 100 mg/kg l-cysteine was injected. To shed light on the relative contributions of autonomic activation and vasopressin release to the effect of high-dose l-cysteine on blood pressure and HR, in four dogs, an additional dose of 100 mg/kg was given after ganglionic blockade with 20 mg/kg hexamethonium and 1 mg/kg atropine (n = 2), or pretreatment with 10 μg/kg of the vasopressin-1 receptor antagonist (β-mercapto-β, β-cyclopentamethylenepropionyl1, O-me-Tyr2, Arg8)-vasopressin (n = 2).

Sixteen beagles weighing between 7.2 and 9.6 kg were used for this part of the study under a protocol approved by the animal care and use committee of Charles River Laboratories, Worchester, Massachusetts. As described earlier in the Cysteine Dose–Response and Acute Cardiovascular Side Effects section, animals were anesthetized with propofol for tracheal intubation and thereafter maintained with isoflurane. Body temperature was controlled with a circulating warm water blanket or forced air blanket. After sterile skin preparation, antibiotic administration (cefotaxime, 50 mg/kg, intravenously), and infiltration of 2% lidocaine, a femoral vessel was cannulated via  small cut down for blood pressure measurement and blood sampling.

Eight animals received 0.4 mg/kg CW002 followed 10 min later by 200 mg/kg l-cysteine; the CW002 dose was fivefold higher than that used for the dose ranging study, and the l-cysteine dose was 2- to 20-fold higher than the doses used in the dose–response analysis. The other eight dogs served as controls, receiving only CW002 and vehicle. For both groups, neuromuscular function was monitored, and after full recovery, the arterial catheter was removed and the wound was closed, the dogs were awakened, and the trachea was extubated. For postoperative analgesia, buprenorphine (0.01 mg/kg, subcutaneously) was administered every 8 h starting the afternoon/evening after the vascular access procedure and continued for a total of four injections.

All animals were observed twice daily for mortality/moribundity. Four animals from each group were sacrificed by pentobarbital overdose on day 2, and the remaining dogs from each group were sacrificed on day 14. Comparisons between groups were made with regard to hematology, coagulation, and serum chemistry. For this analysis, data collected at the time of initial treatment were compared with those obtained at the time of sacrificing. After sacrificing, gross necropsy was performed, which included examination of the carcass and musculoskeletal system; all external surfaces and orifices; the cranial cavity and external surface of the brain; and thoracic, abdominal, and pelvic cavities with their associated organs. Tissues were embedded in paraffin, sectioned, and stained for histopathologic examination.

The dose–response effects of l-cysteine on the duration of action and recovery indices were not normally distributed and therefore assessed by analysis of variance on ranks and the Newman–Keuls test when appropriate, with data presented as median with 25–75% confidence intervals. For all statistical analyses, a P  value less than or equal to 0.05 was considered significant.

The magnitude and timing (relative to injection) of changes in blood pressure and HR produced by l-cysteine injection at each dose for each animal were assessed from continuous data sampled at 200 Hz and averaged over 5-s intervals. Pooled data were normally distributed and quantitatively assessed by analysis of variance for repeated measures and the Newman–Keuls test when appropriate and are presented as mean ± SD. Statistical analysis of cardiac performance variables was precluded by a low power because of the small sample size. Accordingly, dose-related qualitative variations in the cardiac response to l-cysteine were assessed by creating a composite representation of data from all animals. With this approach, drug injection was defined as time 0, and the mean and SD of data averaged over 5-s intervals for each dog were determined. These data were then plotted as a relative value reflecting a ratio of the steady-state data obtained just before injection of l-cysteine.

A single-factor analysis of variance was applied, with animal grouping as the factor, and Dunnett's test applied when appropriate. For all analyses, a P  value less than or equal to 0.05 was considered significant.

Representative recordings of the change in muscle twitch produced by 0.08 mg/kg CW002 alone or in combination with l-cysteine are presented in figure 1. Pooled reversal data (table 1) show that, on average, 0.08 mg/kg CW002 alone produced neuromuscular blockade with a median duration of 75 min, a median time for 5–95% recovery of 23 min, and a median time for 5% recovery to normalized train-of-four of 40 min. l-Cysteine injected 1 min after CW002 markedly reduced the duration of muscle relaxation at all doses (P < 0.001), with both 50 and 100 mg/kg facilitating twitch recovery within 5 min. Similar dose-dependent effects of l-cysteine on recovery indices are also evident. As also shown in table 1, 15 min after complete recovery of twitch after reversal with l-cysteine, there was sufficient residual l-cysteine to reduce the duration of action for additional CW002 in a dose-dependent manner.

As shown in figure 2, increasing doses of l-cysteine tended to increase MAP and reduce HR. These changes were, on average, less than 10% even for 100 mg/kg. Table 2depicts comparison of baseline values for MAP and HR between experiments (there were no differences), and the maximal change in each variable produced by l-cysteine. Of particular note is the fact that the changes in MAP were not clearly dose related: the pressor response to 20 mg/kg represented a significant increase, whereas 50 mg/kg did not change MAP. Furthermore, as depicted in figure 2, the pattern of change for both MAP and HR varied as a function of dose, with 100 mg/kg cysteine eliciting a multiphasic response with the peak increase in MAP occurring later than for other doses.

The temporal relationship between changes in cardiac performance variables in response to 100 mg/kg cysteine is shown in figure 3. Injection of saline vehicle alone had no effect on any variable (data not shown). Consistent with MAP and HR data, changes in LV pressure, stroke volume, contractility, and lusitropy after administration of 100 mg/kg cysteine were less than 10%. Nonetheless, the data do suggest a multiphasic hemodynamic response to 100 mg/kg cysteine with an initial vasodilation followed by a more sustained vasoconstriction. Although the mechanism of this response remains unclear, as shown in figure 4, ganglionic blockade altered the initial fluctuation in MAP produced by 100 mg/kg cysteine but had no effect on the subsequent, more sustained pressor response. In contrast, antagonism of the vasopressin-1a receptor did not prevent the initial fluctuation in MAP evident under control conditions, but abolished the late pressor response.

In the animals that received repeated l-cysteine doses during a 4-week period for definition of the dose–CW002 reversal response, there were no fatalities or morbid events. All dogs maintained normal activity and behavior without change in appetite or weight loss. At the end of the study, two of the dogs were examined by veterinary staff and deemed suitable for adoption. Animals that received a single 200-mg/kg dose of l-cysteine and were sacrificed 2 or 14 days later also exhibited no mortality, moribundity, or weight loss during the observation period. Relative to control animals that did not receive l-cysteine, there were also no differences in hematology, coagulation, and serum chemistry values either at 2 or 14 days (data given in  appendix 1). Similarly, neither gross nor histologic examination revealed differences among groups (data not shown, see  appendix 2for tissue list).

CW002 is a fast-onset, intermediate duration, nondepolarizing neuromuscular blocking agent that is inactivated by nonenzymatic adduction of cysteine to the fumarate component of the molecule.5This study demonstrates that injection of exogenous l-cysteine facilitates the adduction process and can rapidly reverse the neuromuscular blockade produced by CW002, even in the absence of any spontaneous recovery. The optimal l-cysteine dose for reversal of 0.08 mg/kg CW002 (9 × ED⁹⁵) in the dog is approximately 50 mg/kg, a concentration that elicits no significant cardiovascular effect.

The concept of reversing the effect of neuromuscular blocking drugs by molecular inactivation has recently been highlighted by the introduction of sugammadex, a cyclodextrin designed to bind the steroid compound rocuronium.6This idea of displacing a molecule from an active site by providing alternative, high-affinity binding sites has fostered a paradigm shift away from the conventional approach of reversing neuromuscular blockade by giving a cholinesterase inhibitor to increase the concentration of acetylcholine at the motor end plate once there is evidence of spontaneous muscle recovery. It is now possible to reverse the effects of a nondepolarizing muscle relaxant even in the absence of any spontaneous recovery, thus rendering intermediate duration drugs “ultrashort” if needed.

In contrast to the sugammadex–rocuronium interaction, data indicate that CW002 is modified by l-cysteine adduction in such a way that affinity of the molecule for skeletal muscle nicotinic receptors is reduced 60-fold.5Recent in vitro  studies by our research group (unpublished) have provided a putative pathway for CW002 metabolism. Within this construct, CW002 undergoes a slow background hydrolysis (reaction half time of 495 min). Addition of l-cysteine produces an adduction product with a reaction half time of 11.4 min, and subsequent hydrolysis of the adduct occurring with a reaction half time of 56 min. Importantly, these reactions proceed only in a forward direction, and as such, there is no potential for dissociation of cysteine and regeneration of CW002. These in vitro  data support the concept that duration of action for CW002 in vivo  can be shortened by dose-related increases in the amount of l-cysteine available for adduction with the molecule. Our data confirm that injection of exogenous l-cysteine rapidly facilitates inactivation of the CW002 molecule, with 50 mg/kg l-cysteine administered 1 min after CW002 markedly decreasing the median duration of neuromuscular blockade from 75 to 4.6 min. Our data also show that 15 min after complete reversal of CW002, there is sufficient residual l-cysteine to substantially reduce the duration of a subsequent CW002 dose.

Results of the current study indicate that there are no differences between 50 and 100 mg/kg l-cysteine in terms of how each dose shortens the duration of action, 5–95% recovery time, or time to normalization of the train-of-four response. From these data, we conclude that a dose in the 50-mg/kg range is optimal for reversal of CW002 under the experimental conditions. However, this conclusion needs to be interpreted in light of the fact that the only dose of CW002 given was 0.08 mg/kg, or approximately nine times the ED⁹⁵in beagles.4Use of this dose in the current study was based on previous observations in rhesus monkeys indicating that the potency of CW002 in primates is roughly one fifth that in dogs.5To provide a molar mass of CW002 that would approach a clinically relevant concentration in primates without producing an inordinately long duration of neuromuscular blockade in dogs, the 0.08 mg/kg dose was chosen. Importantly, even at this relatively high CW002 dose for dogs, 50 mg/kg l-cysteine injected 1 min after CW002 was remarkably effective in reversing neuromuscular blockade. Preliminary data indicate that reversal of 3 × ED⁹⁵(0.15 mg/kg) CW002 by 50 mg/kg cysteine in rhesus monkeys is similar, if not more rapid, to that observed in dogs.23 

Previous work in our laboratory has demonstrated that even at 100 × ED⁹⁵, CW002 produces only modest and transient hemodynamic changes.4Consistent with these data, 0.08 mg/kg CW002 alone caused no hemodynamic changes in the current study. When injected as an intravenous bolus 1 min after 0.08 mg/kg CW002, l-cysteine also had relatively little hemodynamic effect reflecting on average, changes in MAP and HR less than 10%, even at the highest dose of 100 mg/kg. However, blood pressure effects in particular were not clearly dose related because 20 and 100 mg/kg l-cysteine produced a significant pressor response, whereas 50 mg/kg did not. The reason for this effect remains unclear, but may reflect, at least in part, the emergence of counterbalancing effects on blood pressure, that is, simultaneous stimulation of both vasopressor and vasodilator responses. This process is most apparent after 100 mg/kg l-cysteine in which changes in both MAP and HR were multiphasic (direction of change shifting over time) with the peak increase in MAP occurring later after 100 mg/kg cysteine than for the other doses. More extensive evaluation of cardiac performance in a subset of dogs indicates that 100 mg/kg also produces multiphasic fluctuations in LV pressure, stroke volume, afterload, contractility, and relaxation. When considered in the context of simultaneous changes in MAP and HR, these data suggest an initial vasodilation and sympathetic nervous system activation followed by a more sustained vasoconstriction with a reflex parasympathetic response.

Although clearly demonstrable, the overall hemodynamic change produced by l-cysteine was relatively modest. Nonetheless, previous reports raise points of possible mechanistic and toxicological interest. In a study examining the cardiovascular effects of amino acid infusions in dogs, Hallberg and Soda24found that l-cysteine infusion increased hepatic and portal vein blood flow but did not change systemic arterial blood pressure. For this study, 100 ml of a solution containing 6.058 mg/ml cysteine was infused over 6.25 min for a total dose of approximately 40 mg/kg. In contrast, for the current study, l-cysteine in doses up to 100 mg/kg was given as an intravenous bolus. Potential implications of this include the possibility of entry into the central nervous system.25Previous reports indicate that injection of l-cysteine into the cerebrospinal fluid of conscious rats elicits a dose-related increase in arterial pressure because of autonomic nervous system activation combined with systemic vasopressin release.13Because the concentrations of l-cysteine in the brain and blood were not measured in our study, we cannot directly compare these results with our data obtained after intravenous dosing. Nonetheless, our preliminary observations indicating that the late pressor response to 100 mg/kg l-cysteine can be blocked by a vasopressin antagonist suggest that sufficient amounts of l-cysteine were achieved in the central nervous system to mimic effects noted for central injection of the compound. Given previous reports that high intraperitoneal doses of l-cysteine can produce acute neurotoxicity in both neonatal and adult rodents,10,11entry of l-cysteine into the brain may have potential consequences. However, it is important to note that the doses used in these studies—1,200 mg/kg in infants and 1,500 mg/kg in adults—are close to the reported LD⁵⁰of 1,620 mg/kg for intraperitoneal l-cysteine in the rat7and 24–30 times higher than the optimal dose for CW002 reversal found in the current study.

To more specifically characterize the toxicology of l-cysteine relative to its potency for reversal of CW002, a separate group of dogs received CW002 followed by 200 mg/kg l-cysteine, a dose fourfold higher than that deemed optimal in dose–reversal response studies. Animals were then sacrificed 2 or 14 days later. The study dogs exhibited no clinical, biochemical, or anatomic evidence of drug-related toxicity compared with control animals that did not receive l-cysteine. However, these results need to be interpreted in the context of the limitations inherent to preclinical research. In particular, although the data indicate that l-cysteine in the doses administered is without adverse side effects to the dogs, they do not independently establish that l-cysteine is safe for clinical use. Currently, l-cysteine is added to the parenteral nutrition provided to preterm infants at a dose of approximately 80 mg · kg⁻¹· day⁻¹,14,15and intravenous N -acetyl l-cysteine is approved by the Food and Drug Administration for the treatment of acute acetaminophen toxicity at a dose of 150 mg/kg over 60 min, followed by 50 mg/kg over 4 h, and then 100 mg/kg over 16 h.22Formulated in a preparation containing EDTA and sodium hydroxide as stabilizing agents, Acetadote® (Cumberland Pharmaceuticals Inc., Nashville, TN) has been associated most prominently with nausea and cutaneous anaphylactoid reactions related, at least in part, to histamine release.22The relative roles played by the acetyl and l-cysteine moieties along with the EDTA and sodium hydroxide excipients in reactions to intravenous administration remain uncertain.22However, in the current study of l-cysteine given as bolus injections, we did not observe the cutaneous flushing, hypotension, and reflex tachycardia commonly seen with histamine release in dogs, suggesting that l-cysteine alone may not be the cause of cutaneous anaphylactoid reactions. Nonetheless, clinical trials involving l-cysteine will be required to verify safety in human.

In conclusion, intravenous l-cysteine can rapidly reverse the neuromuscular blockade produced by a CW002 dose nine times greater than the ED⁹⁵in the beagle. The optimal dose for reversal seems to be approximately 50 mg/kg, with no added benefit to 100 mg/kg. After l-cysteine administration for reversal of CW002, the duration of a subsequent CW002 dose is also decreased in a dose-related manner. Over the dose range evaluated, l-cysteine has minimal hemodynamic effects inducing a small (≤ 10%), but consistent, increase in MAP at 20 and 100 mg/kg and modest, but significant, decrease in HR at 100 mg/kg. Preliminary toxicology studies of reversing CW002-induced neuromuscular blockade with 200 mg/kg l-cysteine (fourfold higher than that deemed optimal) did not demonstrate clinical, biochemical, or anatomic evidence of drug-related toxicity.