Comparative Effectiveness of Calabadion and Sugammadex to Reverse Non-depolarizing Neuromuscular-blocking Agents

REVERSAL of neuromuscular-blocking agents (NMBAs) is an important strategy for accelerating recovery from neuromuscular blockade (NMB).^1–4  Acetylcholinesterase inhibitors, typically neostigmine, are the only drugs to reverse NMB approved by the US Food and Drug Administration. Acetylcholinesterase inhibitors cannot reverse deep NMB^5–7  and may even induce muscle weakness when administered in large doses after recovery from NMBAs.⁸ 

Sugammadex, an anionic γ-cyclodextrin derivative, reverses NMB by binding and thereby inactivating steroidal NMBAs.⁹  Hypersensitivity to sugammadex and dose-dependent sugammadex-induced anticoagulation (increases activated partial thromboplastin time and prothrombin time) were reported.^10–12  Although sugammadex is used outside the United States, the Food and Drug Administration has declined it twice based on the observed side effects.¹³ 

Over the past decade, the preparation and molecular recognition properties of a new class known as cucurbiturils have been thoroughly explored.^14,15  Calabadion 1—a cucurbituril derivative—reverses the steroidal nondepolarizing NMBAs rocuronium and vecuronium as well the benzylisoquinoline cisatracurium by encapsulation.¹⁶  However, the in vitro binding affinity of calabadion 1 toward rocuronium is slightly less than the reported binding affinity of sugammadex.¹⁷  Calabadion 2 (fig. 1) is a major leap forward in selectivity and binding affinity toward NMBAs. The affinity of calabadion 2 for rocuronium is 18,900 times that for acetylcholine, whereas the affinity of calabadion 1 for rocuronium is only 350 times higher than for acetylcholine.

A major goal of this study was to conduct a head-to-head comparison of the binding affinity with steroidal NMBAs between calabadion 2 and sugammadex in vitro, ex vivo (in the absence of drug metabolism), and in vivo. In addition, we evaluated the effectiveness of calabadion 2 to reverse cisatracurium-induced NMB.

Calabadion 1 and 2 were synthesized in the Isaacs laboratory (University of Maryland, USA) according to the published procedures.¹⁸ 

To determine the relative binding affinity of calabadion 2 and sugammadex toward rocuronium, we allowed calabadion 2 and sugammadex to compete for a limited amount of rocuronium in one solution (pH = 7.4; 25°C). We determined the concentrations of calabadion 2•rocuronium and sugammadex•rocuronium by ¹H NMR spectroscopy by integration of the diagnostic resonances for the CH₃ groups of rocuronium in the calabadion 2•rocuronium complex that appear between 0.3 and −0.3 parts per million (ppm) relative to an internal standard of known concentration. Substitution of the concentrations into the usual equilibrium expressions along with mass balance considerations allowed us to calculate the difference in binding affinity between calabadion 2 and sugammadex toward rocuronium.

To determine the concentrations of calabadion 2 and rocuronium in the urine samples from the rats, we used a different 1H NMR assay. For this assay, we evaporate the urine samples and then redissolve them in buffered D₂O to allow for 1H NMR analysis. We determine the concentrations of calabadion 2 and rocuronium by integration of the diagnostic resonances for the methyl groups of rocuronium in the calabadion 2•rocuronium complex that appear between 0.3 and −0.3 ppm relative to an internal standard of known concentration.

To monitor the transformation of cisatracurium over time in the Krebs–Henseleit buffer to laudanosin and pentamethylene diacrylate (H₂C=CHCO₂(CH₂)₅O₂CCH=CH₂) by Hoffmann degradation, we performed a third different 1H NMR assay. In this assay, we monitored the formation of pentamethylene diacrylate by the appearance of the diagnostic resonances for the terminal HC=CH₂ groups. We integrate these diagnostic resonances normalized with respect to the initial concentration of cisatracurium reflected by the integral for the resonances of the central (CH₂)₃ group of cisatracurium and pentamethylene diacrylate to determine the concentration of pentamethylene diacrylate as a function of time.

To determine the binding constant for the calabadion 2•succinylcholine complex, we performed a UV/Vis competition assay as published previously.¹⁷  In brief, we prepare a solution of calabadion 2•rhodamine 6G that has a known K_a (4.8 ± 0.1 × 10⁵ M⁻¹) and which displays a different UV/Vis spectrum than rhodamine 6G alone. On addition of increasing concentrations of succinylcholine, we observe that the UV/Vis spectrum reverts toward that of free rhodamine 6G that reflects the competitive binding. We monitor the change in UV/Vis absorbance and fit it to a standard competition binding model as published previously to determine the K_a value for calabadion 2•succinylcholine.¹⁷ 

A total of 34 male Sprague–Dawley rats whose weights ranged from 340 to 406 g were used. Institutional guidelines for animal care and usage for research purposes were strictly followed. All procedures involving animals were approved by the Regierung von Oberbayern (AZ: 55.2-1-54-2532.3-88-13), and experiments were conducted at the research laboratories of Klinikum Rechts der Isar, Munich, Germany.

Rats were anesthetized with carbon dioxide for decapitation, to dissect two hemidiaphragm preparations per rat with intact phrenic nerve. Each hemidiaphragm was secured to a phrenic nerve diaphragm tissue holder and mounted in a 65 ml tissue bath (ISO-07-TSZ2D, Experimetria Ltd., Hungary) with a Krebs–Henseleit buffer solution (1.0 mM KH₂PO₄, 0.6 mM MgSO₄·7H₂O, 5.0 mM KCl, 30.0 mM NaHCO₃, 20.0 mM α-d-glucose, 118 mM NaCl, 2.5 mM CaCl₂) maintained at 37°C (AMP-09 Temperature controller, Experimetria Ltd.) and 95% O₂ and 5% CO₂ (Vol%) insufflations.

The hemidiaphragms were attached to an isometric force-displacement transducer (FSG-01/200 Force Transducer, Experimetria Ltd.) at the centrum tendineum using a crocodile clip (AGF1 Crocodileclip, SKS Hirschmann, Germany). Measurements were amplified by AMP-01-SG Classic bridge amplifier and recorded with the S.P.E.L. Advanced Isosys software (Experimetria Ltd.). A resting force of 10 mN was applied for 20 min followed by 50 mN to assess a resting and peak forces during the experiment. The phrenic nerve was stimulated with a 2-Hz train-of-four (TOF) stimulus every 15 s (rectangular pulses of 0.2 ms duration with a supramaximal voltage) using a square wave stimulator (ST-03-O4, Experimetria Ltd.). To ensure a steady state, stimulation was continued for 15 min before further measurements were taken.

The NMB induced by rocuronium and cisatracurium was quantified as the depression of the force amplitude of the first twitch response (T1) to the TOF stimulus. To decrease the interindividual variability when making comparison of calabadion 2’s effects to reverse benzylisoquinolines and steroids, one hemidiaphragm preparation of each rat was treated with rocuronium and the other was treated with cisatracurium.

Rocuronium and cisatracurium were added in steps of 150 and 100 µg, respectively, until the T1 value was decreased by more than 10% compared with the control value. After each NMBA increment, a stable T1 value, i.e., three consecutive T1 with less than 3% deviation was achieved after 5 to 8 min. Therefore, the NMBA boluses were added every consecutive 10 min. On the basis of these measurements, we then subsequently applied a 1.5 times the EC₉₉ of rocuronium (62 µM) and cisatracurium (16 µM) to evaluate effectiveness of the reversal agents.

Neuromuscular transmission recovery was assessed by the recovery of T1 and TOF ratio (ratio of the twitch response of the fourth to the first stimulus of each TOF stimulation). To evaluate their effectiveness, the reversal agents were administered every 10 min until recovery of the TOF ratio to 0.9. To evaluate rocuronium reversal, we administered an initial bolus of sugammadex or calabadion 2, each 5 mg, followed by repetitive doses of 0.4 mg. For cisatracurium reversal, an initial dose of 8 mg calabadion 2 was added to the tissue bath, followed by repetitive doses of 1.5 mg. To address the impact of the Hoffmann degradation on the concentration of cisatracurium when cisatracurium was used as NMBA, 10 additional hemidiaphragm preparations were incubated in 6 µM cisatracurium bath (EC₉₀) to evaluate spontaneous recovery. In these preparations, the twitch response was measured every 10 min, and the expected spontaneous sigmoidal recovery course was observed. Linearization by logit transformation revealed a mean increase of twitch responses by 0.26 each 10 min. The twitch response values of the reversal studies were corrected accordingly before the individual concentration response relations were calculated. To account for the time-dependent degradation of cisatracurium into laudanosin and pentamethylene diacrylate (Hoffmann degradation), we calculated the increase in the concentration of pentamethylene diacrylate. The concentration of cisatracurium in Krebs–Henseleit buffer (initial concentration: 3.5 mM, pH = 7.4, 37.5°C, ingredients as described earlier in Experimental Procedures) was repeatedly measured by ¹H NMR every 10 min for 1 h and then hourly for 5 h. The concentration of pentametyldiacrylate was measured. The 1H NMR resonances for the characteristic terminally located alkene protons increased over time and were quantified. Under these conditions, the magnitude of decrease in cisatracurium concentration amounted to 18% ln(t[h]) + 41%, i.e., 8.3% every 10 min. The cisatracurium concentrations used to calculate the affinity to calabadion 2 were corrected based on the measurements of Hoffman degradation by ¹H NMR obtained over 1 h.

A total of 108 male Sprague–Dawley rats (weight, 227 to 356 g) were used. Institutional guidelines for animal care and usage for research purposes were strictly followed. All procedures involving animals were approved by the Institutional Animal Care and Use Committees at Harvard Medical School, and experiments were conducted at the research laboratories of the Massachusetts General Hospital, Boston, Massachusetts.

Rats were anesthetized (5% isoflurane, maintained to 1.5%) before cannulation of their left femoral vein and artery for drug administration, blood pressure analyses, and blood gas analyses. The animals were tracheotomized to maintain normal breathing throughout the surgery. The body temperature was monitored with a rectal temperature probe (regulation to 37° ± 1°C). Blood pressure was recorded continuously throughout the experiment, and blood gas analyses were performed 1 min before and 10 and 20 min after administration of the relaxant. Total urine volume was collected in eight rats at the end of 60 min after the injection of calabadion 2 by a single puncture in the bladder, after which the animal was killed.

For evaluation of neuromuscular transmission, the femoral nerve was stimulated with two subcutaneously inserted needle electrodes on a supramaximal level. The evoked response of the quadriceps femoris muscle was detected using accelerometry with a TOF-Watch SX Monitor (Organon Ireland Ltd., a part of Schering-Plough Corporation, Ireland) as has been published previously.^16,19  The transducer was inserted subcutaneously ventromedially at the proximal end of the thigh, next to the tuberosity of the tibia. After determination of the supramaximal stimulation current, the femoral nerve was continuously stimulated at 1 Hz until twitch height reached a stable plateau, which was defined as baseline. Stimulation was applied for at least 10 min before any drug injection. We prepared envelopes for a block randomization in groups of 12 rats at a time. After surgical preparation of the rat had been finished, a card was blindly drawn to define the reversal drug and dose to be administered. The person who conducted the experiment was blinded throughout the experiment. After twitch recovery to 50% of baseline, the single twitch stimulation mode was changed to the TOF stimulation mode with a 12-s interval to capture shallow levels of NMB. TOF monitoring was continued until 20 min after injection of the reversal agent.

To compare the dose–response relationship of NMB reversibility among calabadion 2, calabadion 1, neostigmine, and sugammadex, we measured time to recovery of neuromuscular transmission after intravenous injection of the twofold ED₉₀ of rocuronium (3.5 mg/kg), vecuronium (0.7 mg/kg), or cisatracurium (0.6 mg/kg; table 1).^16,20  Previous experiments indicated a group size of four to be sufficient.¹⁶  At the onset of apnea, mechanical ventilation was started, and the test drug (reversal agent or saline) was injected 30 s later. Mechanical ventilation was terminated when animals started sufficient breathing. The rats were observed for at least 30 min, and rats with bladder puncture were observed for at least 60 min after administration of the relaxant.

To evaluate the effects of calabadions on the subsequent administration of succinylcholine (calabadion 2•succinylcholine: K_a = (2.8 ± 0.1) × 10⁶ M⁻¹), the depolarizing NMBA was administered 60 s after calabadion 1/2–assisted recovery from vecuronium-induced relaxation. We compared the onset of NMB after succinylcholine-induced relaxation between the conditions pretreatment (with calabadion 1 or calabadion 2) or no pretreatment (placebo).

Rocuronium 3.5 mg/kg (two times ED₉₀), 0.7 mg/kg vecuronium (two times ED₉₀), 0.6 mg/kg cisatracurium (two times ED₉₀), 0.9 mg/kg succinylcholine (two times ED₉₀), 0.06/0.012 mg/kg neostigmine/glycopyrrolate, and sugammadex were diluted in 0.5 ml water. Calabadion 1 and 2 were diluted in 0.5 ml H₂O. All drugs were administered over 5 s. Isoflurane (Flurane; Baxter Healthcare Corporation, USA), cisatracurium (Abbott Laboratories, USA), rocuronium (Zemuron, Merck, USA), vecuronium (Novaplus, USA), succinylcholine (Hospira, Inc., USA), and neostigmine/glycopyrrolate (West-Ward, USA) were obtained from clinical supplies. Sugammadex (Merck) was imported.

From the urine samples of rats injected with rocuronium or cisatracurium, we took 0.1 ml aliquots of the urine and evaporated them under high vacuum. The residue was dissolved in a solution containing a known concentration of 2,2,3,3-d(4)-3-(trimethylsilyl) propanoic acid sodium salt (TMSP) as internal standard, and the sample was diluted to a total volume of 0.6 ml with 20 mM phosphate-buffered D₂O (pH = 7.4). The concentrations of the calabadion 2•rocuronium complexes were determined by ¹H NMR spectroscopy (fig. 2) by comparing the integral for diagnostic peaks of the complex (CH₃ groups at 0.23 and −0.32 ppm, 3H each) relative to the internal standard (−0.02 ppm, 9H). The concentrations of free rocuronium were measured by adding a twofold molar excess of calabadion 2 to the aforementioned solution. Addition of excess calabadion 2 bound any free rocuronium as the calabadion 2•rocuronium complex. Conversely, addition of excess rocuronium allowed the measurement of calabadion 2 not originally bound to rocuronium. We measured only total concentrations of calabadion 2 for the cisatracurium samples.

We conducted a preclinical hypothesis-driven drug development study and tested calabadion 2 for one indication, reversal of nondepolarizing NMB. We tested the research hypotheses that calabadion 2 increases speed of recovery from NMB compared with available comparators. The primary aim of the in vivo experiments was to compare the effect of calabadion 2 in reversing benzylisoquinolines and steroidal NMBAs relative to neostigmine, placebo, and sugammadex (the latter comparison was made for steroids only). The primary outcome variable was time to recovery from NMB expressed as time to onset of breathing and time to recovery of TOF ratio to 90%.

To address the primary aim, we tested the hypotheses that calabadion 2 increases speed of recovery from benzylisoquinoline and steroidal NMBA–induced NMB compared with neostigmine, calabadion 1, and placebo in a dose-dependent fashion. We included all measurements of time to recovery across muscle function types, using a mixed linear model (compound symmetry repeated covariance type).

To evaluate effects on time to recovery from NMB, we included relaxant (vecuronium, rocuronium, or cisatracurium), reversal agent (calabadion 2, calabadion 1, neostigmine, or placebo), and reversal agent dose in milligrams per kilogram and observed muscle function type (recovery of spontaneous breathing vs. extremity TOF ratio) as repeated independent variables and time to recovery as the dependent variable. We tested for an interaction between the administered reversal agent and the slope of the reversal agent dose.

To evaluate the secondary aim, we hypothesized that calabadion 2 reverses steroidal NMBA–induced NMB with a higher potency than sugammadex. As for the primary aim, we included all measurements of time to recovery across muscle function types using a mixed linear model (compound symmetry repeated covariance type) and evaluated the effects of both reversal agents on the dependent variable time to recovery from NMB including relaxant (vecuronium or rocuronium), reversal agent (calabadion 2 or sugammadex), and reversal agent dose (adjusted for molecular weight [1,641 g/mol for calabadion 2 vs. 2,178 g/mol for sugammadex]) and observed muscle function type as repeated independent variables. We tested for an interaction between the administered reversal agent and the slope of the reversal agent dose.

In the ex vivo experiments, twitch response was defined as the difference of resting force and peak force. Assuming that the relationship between NMBA and reversal agent concentrations with the TOF ratio is governed by the Hill equation, linear regressions of the TOF ratio in a logit scale and the concentrations in natural log scale were created for each preparation.²¹  The transformed variables of the individual linear regressions, i.e., the individual slopes and intercepts, were statistically evaluated by an analysis of covariance for repeated measurements using either NMBA or reversal agent (between groups) or regression variable (within group) as independent factors. The concentration effect relation is described by the slope and the intercept and by the effective concentration for a 50% effect (EC₅₀).

Data are presented as mean ± SD unless otherwise specified. P < 0.05 was considered to be the minimum criterion for statistical significance, and no attempts were made to adjust for multiple comparisons. Statistical analysis was performed using SPSS 20.0 (SPSS Inc., USA).

Figure 3 shows the ¹H NMR spectra recorded for rocuronium, calabadion 2•rocuronium, and sugammadex•rocuronium, and a mixture of rocuronium (199.8 μM), calabadion 2 (186.8 μM), and sugammadex (2,399.7 μM) in 20 mM phosphate-buffered D₂O at pH 7.4. As shown in figure 3, B and C, the resonances for the axial steroidal CH₃ groups appear at different chemical shifts in the calabadion 2•rocuronium (≈0.3 and −0.3 ppm) and sugammadex•rocuronium (≈1.0 ppm) complexes. Figure 3D shows that these axial steroidal CH₃ groups appear as separate resonances even when calabadion 2 and sugammadex are competing to bind to rocuronium. This advantageously allows us to integrate these resonances to determine the concentrations of calabadion 2•rocuronium and sugammadex•rocuronium in the solution at equilibrium. By using the usual equilibrium and mass balance expressions as detailed latter in this section, we are able to calculate that calabadion 2 binds 89 times more tightly to rocuronium than sugammadex does. When combined with the known K_a value for the calabadion 2•rocuronium (K_a = 3.4 × 10⁹ M⁻¹),¹⁷  we are able to determine the K_a value for sugammadex•rocuronium (K_a = 3.8 × 10⁷ M⁻¹) in the 20 mM phosphate buffer used herein. When comparing the K_a value for calabadion 2•rocuronium with the binding affinity of sugammadex•rocuronium, by using the same method that is a competition ¹H NMR assay, calabadion 2 has a substantially higher binding affinity to rocuronium (K_a value: calabadion2•rocuronium = 3.4 × 10⁹ M⁻¹vs. sugammadex•rocuronium: 3.8 × 10⁷ M⁻¹, respectively).

The mass balance expression ([rocuronium]_total = 199.8 μM = [calabadion 2•rocuronium] + [sugammadex•rocuronium]) allowed us to calculate the concentration of the complex calabadion 2•rocuronium (136.1 μM) and sugammadex•rocuronium (63.7 μM). Assuming a closed system, we calculated the concentration of free calabadion 2 (50.7 μM) and sugammadex (2336 μM) based on the spectral analysis (fig. 3) of their complexes with rocuronium. The relative binding constant (K_rel = ([calabadion 2•rocuronium] [sugammadex]_free)/([sugammadex•rocuronium] [calabadion 2]_free)) was obtained as K_rel = 98.4. In a solution containing 186.8 μM calabadion 2, 199.8 μM rocuronium, and 4.7993 mM sugammadex, the relative binding constant was comparable (K_rel = 81.45). The value of K_a for calabadion 2•rocuronium of 3.4 × 10⁹ M⁻¹ and the K_a of sugammadex•rocuronium of 3.8 × 10⁷ M⁻¹ allows us to conclude that the binding affinity of calabadion 2•rocuronium compared with sugammadex•rocuronium is 89 times higher. We determined the K_a value for calabadion 2•succinylcholine as 2.8 × 10⁶ M⁻¹ by monitoring the change in UV/Vis absorbance and fit it to a standard competition binding model as published previously.¹⁷ 

Rocuronium and cisatracurium induced a concentration-dependent decrease in twitch height of the rat hemidiaphragm with an EC₅₀ of 12 and 3 μM, respectively (table 2). The EC₅₀ of calabadion 2 and sugammadex in reversing rocuronium (1.5 × EC99) was within the same range (assessed by TOF, 61 vs. 49 μM and assessed by T1, 53 vs. 43 μM, respectively, fig. 4), suggesting that calabadion 2 binds rocuronium with a 1:1 binding ratio similar to that of sugammadex. For the reversal of cisatracurium, we identified an EC₅₀ of calabadion 2 of 80 μM when assessed by T1 and 100 μM when assessed by TOF. We did not identify any reversal of cisatracurium with sugammadex at concentrations up to 1,000 μM.

Calabadion 2 reversed vecuronium-induced (P < 0.01 for main effect of reversal agent dose, fig. 5), rocuronium-induced (P < 0.001, fig. 6), and cisatracurium-induced (P < 0.01, fig. 7) NMB dose dependently. At the highest doses, the time to recovery of NMB amounted to 16.6 ± 7.9 s, 13.6 ± 9.2 s, and 15.8 ± 5.6 s after reversal administration, respectively, with significantly faster recovery of spontaneous breathing compared with TOF recovery (P < 0.001 for interaction effect of reversal agent and muscle group). Time to recovery from NMB was dose dependently significantly faster across all three relaxants when reversed with calabadion 2 compared with neostigmine and placebo (P < 0.001 for interaction effect of reversal agent and reversal agent dose on recovery time).

Calabadion 2 reversed vecuronium-induced (P < 0.01 for main effect of reversal agent dose) and rocuronium-induced (P < 0.001) NMB dose-dependently. Time to recovery from vecuronium-induced NMB to onset of breathing was significantly faster with calabadion 2 compared with sugammadex, calabadion 1, neostigmine, and placebo (figs. 5 and 6). The molar potency (interaction between number of molecules of reversal agent administered and time to recovery of spontaneous breathing after administration of steroidal NMBA) of calabadion 2 to reverse the effects of vecuronium was greater compared with sugammadex (P < 0.05; fig. 7).

Calabadion 2 reversed cisatracurium-induced NMB dose-dependently (P < 0.01 for main effect of reversal agent dose). Time to recovery of muscle function (expressed as TOF 0.9 and spontaneous breathing recovery) after cisatracurium-induced NMB was significantly shorter with calabadion 2 compared with placebo and neostigmine/glycopyrrolate (fig. 8).

Onset of action and peak NMB of succinylcholine did not differ in animals pretreated with calabadion 1 and 2 to reverse vecuronium-induced NMB compared with those received placebo (fig. 9).

We did not observe any adverse effects of calabadion 2 on heart rate, blood pressure, or arterial blood gas parameters. No signs of residual blockade or recurarization were detected.

One hour after intravenous administration of calabadion 2 (40 to 80 mg/kg), 49 ± 31% (mean ± SD) of the drug was detected in the urine samples. When administered in low dosage (5 to 10 mg/kg), 62 ± 17% of calabadion 2 was detected in urine.

Calabadion 2 is a broad-spectrum agent used to rapidly reverse deep vecuronium-, rocuronium-, and cisatracurium-induced NMB in a dose-dependent manner. Calabadion 2, like sugammadex, reverses steroidal NMBAs with a 1:1 binding ratio, has a higher binding affinity in vitro, and provides a higher molar potency to reverse steroidal-induced NMB in vivo. Calabadion 2 was well tolerated in the rat, and a substantial amount of calabadion 2 was eliminated by the kidney within 1 h.

The 1:1 binding between calabadion 2 and NMBAs was established previously by ¹H NMR integration of resonances for calabadion 2 relative to NMBA and also by Job plots.¹⁷  In this study, the 1H NMR experiments used for head-to-head binding assay and the urine analysis reconfirmed the 1:1 binding.

In combination, our current head-to-head competition experiment between calabadion 2 and sugammadex for rocuronium (calabadion 2 binds 89 times stronger than sugammadex) and our previous determination that calabadion 2 binds rocuronium stronger than calabadion 1 establish that calabadion 2 has superior binding affinity.¹⁷  We attribute the superior binding affinity of calabadion 2 toward rocuronium to several factors. First, the naphthalene walls of calabadion 2 define a hydrophobic box that is complementary to the hydrophobic steroidal skeleton of rocuronium. Second, and more importantly, the ureidyl C=O portals of calabadion 2 exhibit strong ion-dipole interactions toward the cationic N atoms of rocuronium that are not possible with sugammadex.

The higher in vitro binding affinity to steroidal NMBAs of calabadion 2 compared with that of sugammadex has the potential to translate to a high in vivo molar potency in the rat. The high binding affinity is an encouraging pharmacologic observation. Reversing drugs by encapsulating carries the risk of nonspecific binding to biologically relevant plasma molecules or medications the patients may also have taken. The experience with sugammadex shows that its side effects (inhibiting effects on coagulation) are dose-dependent.²²  Although we have so far not observed inhibiting effects of calabadion 2 on coagulation, it will be important to use the lowest effective dose of a drug that inactivates NMBAs by binding. Our data showing the higher in vitro and in vivo (fig. 5 to 8) binding affinity of calabadion 2 compared with sugammadex suggest that calabadion 2 may have a slightly wider therapeutic range. Currently, further experiments to evaluate the potential of calabadion 2 to bind to alternate targets are being performed.

Sugammadex, calabadion 1, and calabadion 2, as shown in this study, are renally eliminated within a short time.¹⁶  This pharmacokinetic profile helps to prevent long-term effects of the containers on medications administered during the postoperative course.

In our ex vivo experiments, we observed similar values of the EC₅₀ of calabadion 2 and sugammadex (TOF 0.5 reversal after calabadion 2 injection: 61 µM [95 CI, 41 to 88] and after sugammadex: 54 µM [32 to 73]) as well as slope of the recovery curves (fig. 4; TOF 0.5 reversal after calabadion 2 injection: 22 to 27 and after sugammadex: 20 to 24) in reversing rocuronium (table 2), indicating that calabadion 2, like sugammadex, binds to steroidal NMBAs in a 1:1 ratio. Of note, under the conditions of the steady-state ex vivo preparation, differences in binding affinity between sugammadex and calabadion 2 did not and are not expected to translate into effects on EC₅₀ of these compounds.

Within the rat hemidiaphragm preparation, we created an environment of steady-state NMB that does not contain plasma proteins. Therefore, high binding affinity/selectivity is not expected to translate into increased speed of recovery from NMB under these conditions. Because the ex vivo diffusion distance is relatively long compared with a perfused muscle or an in vitro solution, the ability of rocuronium to pass through the tissue is limited, even with a maximum concentration gradient created by calabadion 2 or sugammadex encapsulation.

Accordingly, because of the initial excess of rocuronium molecules at postjunctional receptors, their removal does not result in an instant recovery of twitch height. However, as soon as we achieve a reduction to 70% receptor occupation, we will observe a complete recovery reversal of the twitch height.²³ 

Sugammadex reverses rocuronium- and vecuronium-induced NMB by encapsulation into its lipophilic cavity but does not bind to benzylisoquinolines, e.g., cisatracurium, which account for about one third of the market volume of NMBAs.^24–26  Not all mammalian species react in the same way to neuromuscular-blocking substances.²⁷  Rats are more resistant to nondepolarizing NMBAs and more sensitive to depolarizing NMBAs than humans. The dose–response curve of calabadion and sugammadex to reverse the effects of nondepolarizing NMBAs should be shifted to the left in humans compared with this study in rats. Accordingly, lower doses of an encapsulating agent should be required in humans compared with those required to reverse rocuronium in the rat, which is also supported by published preclinical sugammadex data.^28,29  Ultimately, dose–response studies in humans are required to define the risk–benefit ratio of calabadion 2 when given for reversal of nondepolarizing NMBAs.¹⁸ 

Calabadion 1 was the first reversal agent that reversed deep cisatracurium-induced NMB by binding and encapsulation, an unmet need because no alternative medication is available to reverse deep cisatracurium-induced NMB. Calabadion 2 has an almost five times higher binding affinity toward cisatracurium compared with calabadion 1, and substantially lower doses are required to reverse cisatracurium in vivo.¹⁷  The affinity of calabadions to cisatracurium is lower than their affinity toward steroidal NMBAs. In contrast, sugammadex does not bind to cisatracurium, and the binding affinity of calabadion 2•cisatracurium Ka = 4.8 × 10⁶ M⁻¹ is one eighth that of lower than sugammadex•rocuronium (K_a = 3.8 × 10⁷ M⁻¹). We demonstrated a stable and comparable clinical effect, when the doses had been adapted appropriately. Importantly, reappearance of breathing after reversal of cisatracurium with calabadion 2 was sufficiently fast (17.5 ± 6.5 s for calabadion 2 to 60 mg/kg compared with 465 ± 196 s for placebo and 291 ± 99 s for neostigmine and compared with 47 ± 13 s for the previously published calabadion 1 to 150 mg/kg).¹⁶ 

We attribute the enhanced binding affinity of calabadion 2 (figs. 5 to 8) for its targets to its larger hydrophobic cavity that is shaped by two naphthalene walls as opposed to calabadion 1, which features two benzene walls.³⁰  Currently, calabadion 2 is developed as a new broad-spectrum NMBA reversal agent. Unlike sugammadex, calabadion 2 also reverses cisatracurium-induced NMB. We were able to show this effect at each level of our experiments, i.e., in vitro, ex vivo, and in vivo.

In clinical practice, surgical complications may occur that require a second (emergency) surgical procedure with anesthesia, shortly after the reversal of the NMBA with an agent that inactivates NMBAs, such as sugammadex or calabadion. It is possible to administer a higher dose of the nondepolarizing NMBA injected before the reversal. Cammu et al.³¹  found an inverse relationship between the onset time and the time interval between sugammadex and the repeat administration of rocuronium, and a direct relationship between the duration of NMB and the time interval between sugammadex and the repeat administration of rocuronium.

However, currently, it is recommended that after initial reversal of NMB with sugammadex, 24 h should be allowed before rocuronium or vecuronium can be readministered. Based on the European Medicines Agency recommendations, succinylcholine can be given.³²  Given the substantial in vitro affinity of calabadion 2 to succinylcholine, we conducted an additional set of experiments to evaluate whether succinylcholine can be used after calabadion reversal of nondepolarizing NMBA. Our data (fig. 9) show that succinylcholine can be used safely and effectively directly after administration of calabadion 2 to reverse a nondepolarizing NMBA.

The concept behind our study brings important and new input in the field of drug development. The model of drug inactivation by encapsulation will become even more important in the future because the clinically meaningful problems such as of adverse and lingering effects of anesthetics and NMBAs or cocaine intoxication have not yet been solved.³³  By using the principle of encapsulation, the relationship between binding affinity to the target drug to be reversed and the binding affinity to other chemically similar compounds that are not to be inactivated (that is binding selectivity) has to be considered.

New encapsulating drugs that affect neuromuscular transmission have to be tested in vitro, ex vivo, and in vivo for measurements of binding affinity and selectivity. The clinician scientist should consider that differences in binding affinity and selectivity may translate to effects on speed of reversal in vivo, as has been demonstrated in our study. In contrast, in a steady-state ex vivo model (such that the phrenic nerve hemidiaphragm model) that does not contain plasma proteins, high binding affinity/selectivity does not translate to increased drug potency.

In summary, we show that calabadion 2 is a broad-spectrum reversal agent to rapidly reverse deep vecuronium-, rocuronium-, and cisatracurium-induced NMB in a dose-dependent fashion. Calabadion 2, like sugammadex, reverses steroidal NMBAs with a 1:1 binding ratio, has a higher binding affinity in vitro, and provides a higher molar potency to reverse steroidal-induced NMB in vivo. Calabadion 2 was well tolerated in the rat, and substantial amount of calabadion 2 was eliminated unchanged by the kidney within 1 h.

This study was supported by the Department of Anesthesia and Critical Care of the Massachusetts General Hospital (Boston, Massachusetss) and in part by the National Institutes of Health (Bethesda, Maryland; CA-168365).

Drs. Isaacs and Eikermann hold equity stakes in Calabash Bioscience, Inc. (College Park, Maryland), which is developing calabadion 2 for biomedical application. The other authors declare no competing interests.