The investigational etomidate analog ABP-700 causes involuntary muscle movements in humans at anesthetic doses and seizures in dogs at 10-fold higher toxicologic doses
The mechanism of seizures in dogs, and their relationship to involuntary muscle movements in humans, is unknown
Toxicologic studies in dogs using supratherapeutic ABP-700 doses caused involuntary muscle movements and seizures, but these were temporally and electroencephalographically distinct, suggesting different underlying mechanisms
Events occurred at ABP-700 and metabolite concentrations one and two orders of magnitude higher, respectively, than those found in humans
Electrophysiologic studies of the principal metabolite of ABP-700 in oocyte-expressed γ-aminobutyric acid type A receptors showed inhibition at the high supratherapeutic concentrations achieved in the dogs, and such inhibition may explain seizure activity
Proepileptiform effects of ABP-700 in dogs may not be relevant to humans at therapeutic doses
The etomidate analog ABP-700 produces involuntary muscle movements that could be manifestations of seizures. To define the relationship (if any) between involuntary muscle movements and seizures, electroencephalographic studies were performed in Beagle dogs receiving supra-therapeutic (~10× clinical) ABP-700 doses. γ-aminobutyric acid type A (GABAA) and glycine receptor studies were undertaken to test receptor inhibition as the potential mechanism for ABP-700 seizures.
ABP-700 was administered to 14 dogs (6 mg/kg bolus followed by a 2-h infusion at 1 mg · kg-1 · min-1, 1.5 mg · kg-1 · min-1, or 2.3 mg · kg-1 · min-1). Involuntary muscle movements were documented, electroencephalograph was recorded, and plasma ABP-700 and CPM-acid concentrations were measured during and after ABP-700 administration. The concentration-dependent modulatory actions of ABP-700 and CPM-acid were defined in oocyte-expressed α1β3γ2L GABAA and α1β glycine receptors (n = 5 oocytes/concentration) using electrophysiologic techniques.
ABP-700 produced both involuntary muscle movements (14 of 14 dogs) and seizures (5 of 14 dogs). However, these phenomena were temporally and electroencephalographically distinct. Mean peak plasma concentrations were (from lowest to highest dosed groups) 35 μM, 45 μM, and 102 μM (ABP-700) and 282 μM, 478 μM, and 1,110 μM (CPM-acid). ABP-700 and CPM-acid concentration–GABAA receptor response curves defined using 6 μM γ-aminobutyric acid exhibited potentiation at low and/or intermediate concentrations and inhibition at high ones. The half-maximal inhibitory concentrations of ABP-700 and CPM-acid defined using 1 mM γ-aminobutyric acid were 770 μM (95% CI, 590 to 1,010 μM) and 1,450 μM (95% CI, 1,340 to 1,560 μM), respectively. CPM-acid similarly inhibited glycine receptors activated by 1 mM glycine with a half-maximal inhibitory concentration of 1,290 μM (95% CI, 1,240 to 1,330 μM).
High dose ABP-700 infusions produce involuntary muscle movements and seizures in Beagle dogs via distinct mechanisms. CPM-acid inhibits both GABAA and glycine receptors at the high (~100× clinical) plasma concentrations achieved during the dog studies, providing a plausible mechanism for the seizures.
ABP-700 is a “soft” analog of etomidate (fig. 1).1–4 Similar to other soft drugs such as remifentanil and esmolol, it contains a metabolically-labile ester moiety that is rapidly hydrolyzed by esterases to a carboxylic acid metabolite (CPM-acid). In developing this and other soft etomidate analogs, our primary goal was to retain the potent sedative-hypnotic activity and benign cardiorespiratory actions of etomidate, while eliminating the prolonged suppression of adrenocortical steroid biosynthesis that plagues etomidate administration. A secondary goal was to produce an agent that would allow rapid anesthetic recovery. Subsequent human studies have confirmed that at clinically relevant doses, ABP-700 minimally affects respiratory and cardiovascular function.4,5 However, unlike etomidate, it does not suppress steroid biosynthesis and has a fast recovery profile even after prolonged infusion.4,6,7
With ABP-700 administration, involuntary muscle movements are commonly observed.4–6 Although such movements are also commonly observed with etomidate administration and not associated with seizures, the corporate licensee of ABP-700 (The Medicines Company, USA) undertook studies in Beagle dogs to determine whether there was any electroencephalographic seizure activity during these movements. During those toxicologic studies—which were performed at the direction of the United States Food and Drug Administration with the goal of achieving ABP-700 exposure levels that are 10× higher than the maximum associated with clinical doses—electroencephalographically-confirmed convulsive seizures were sometimes observed. Although no seizures or postictal behaviors were reported in clinical studies, establishing the likely mechanism(s) for these seizures in Beagle dogs and assessing the potential of ABP-700 for causing seizures in humans is critical for the future development of ABP-700 as a therapeutic drug.4,5 More broadly, defining the cause of these seizures may be important for guiding the design and development of other novel sedative-hypnotic agents that act by similar receptor mechanisms.
Many general anesthetics are widely believed to produce their behavioral effects—at least in part—by acting on γ-aminobutyric acid type A (GABAA) receptors in the central nervous system (CNS).8 Within the clinically relevant concentration range, these agents enhance (“potentiate”) GABAA receptor function in a concentration-dependent manner leading to a reduction in CNS excitability that manifests as sedation or hypnosis. This receptor action also accounts for the anti-convulsant activity of many general anesthetic agents.9,10 Conversely, at lethally high concentrations and similar to classic convulsants such as pentylenetetrazol and picrotoxin, general anesthetics commonly inhibit GABAA receptor function.11–14 For propofol and etomidate, this inhibitory action is typically evident at concentrations greater than or equal to 30 μM, which is more than 10-fold higher than their respective sedative-hypnotic concentrations.11,14,15 This inhibitory action presumably results from anesthetic binding to a low affinity site(s) on the GABAA receptor that is distinct from those responsible for potentiation. Consequently, concentration-response curves that define anesthetic actions on GABAA receptors are typically bell-shaped, exhibiting potentiation at clinically-relevant anesthetic concentrations and inhibition at lethally high ones.11,16
In the current article, we report the results of the dog toxicology studies along with in vitro studies aimed at characterizing the GABAA and glycine receptor modulatory actions of ABP-700 and CPM-acid. Our goal in undertaking the receptor studies was to test the hypothesis that ABP-700 or its metabolite CPM-acid produced seizures in these dogs because it had reached concentrations sufficient to inhibit the function of these inhibitory receptors, which is a well-established mechanism for producing seizures.17,18
Materials and Methods
Sources of Drugs and Chemicals
ABP-700 for Beagle dog studies was obtained from KABS Laboratories (Canada) as a sterile formulation at 20 mg/ml in 10% sulfobutylether-β-cyclodextrin (pH 2.5) and stored frozen (−20°C) before being brought to room temperature within 6 h of administration. ABP-700 and CPM-acid for GABAA receptor studies were synthesized by Aberjona Laboratories (USA) using our previously described methods.1,19 Etomidate-acid was synthesized as previously described.20 Other chemicals were purchased from Sigma-Aldrich (USA).
Toxicologic Studies of ABP-700 in Beagle Dogs
All Beagle dog studies were performed by CiToxLAB North America (Canada) and conducted in accordance with their institutional animal care and use committee and the principles outlined in the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health. Dogs (n = 14 dogs; 7 males and 7 females; age: 8 to 9 months at the time of ABP-700 dosing) were purchased from Marshall BioResources (USA). Dogs were individually housed in stainless steel cages equipped for telemetry data acquisition and day/night vision cameras for continuous video monitoring. Room temperature was maintained at 21° ± 3°C with 12-h light-dark cycles. At least 26 days before ABP-700 dosing, dogs were anesthetized with propofol and isoflurane and surgically instrumented for telemetric monitoring of electroencephalography and electromyography, intraarterial blood pressure, heart rate, body temperature, and physical activity generally as previously described.21 Sterile transmitters (D70-EEE and M10; Data Science International, USA) were inserted between the internal abdominal oblique muscle and the aponeurosis of the transverses abdominis and secured with nonabsorbable suture material. The electroencephalographic and electromyographic leads were tunneled to a dorso-cervical neck incision. Femoral venous and intraarterial catheters were also inserted at this time. Electroencephalographic electrodes were attached to the cranium to monitor two standard bipolar derivations (CZ-OZ and C4-O2) and the electromyographic electrodes were placed in the temporalis muscle. Both electroencephalographic and electromyographic signals were continuously recorded and analyzed beginning at least 30 min before ABP-700 administration and ending at least 24 h after administration was complete using Dataquest ART software and NeuroScore software with the seizure detection module (Data Sciences International, USA). Electroencephalographic interpretation was completed by a board certified veterinary neurologist who included a description of the involuntary movements and possible electroencephalographic correlates in a report to the corporate sponsor. That interpretation was peer-reviewed by an independent neuroscientist. A seizure was defined as an electroencephalographic pattern lasting greater than or equal to 10 s and satisfying either of the following criteria: (1) repetitive generalized or focal spikes, sharp-waves, spike-and-wave, or sharp-and-slow wave complexes at greater than or equal to 3 Hz; or (2) sequential rhythmic, periodic, or quasiperiodic waves at greater than or equal to 1 Hz and unequivocal evolution in frequency (gradual increases/decreases by greater than or equal to 1 Hz), morphology, or location.22 Heart rate, intraarterial blood pressure, and body temperature were acquired and analyzed using DSI Ponemah 6.30 software (Data Sciences International).
Before study, dogs were fasted for 12 to 24 h. Each dog was premedicated with glycopyrrolate (0.02 mg/kg intramuscularly) and then received an ABP-700 bolus (6 mg/kg) immediately followed by a 2-h ABP-700 infusion at 1 mg · kg-1 · min-1 (total dose:126 mg/kg; n = 4 dogs), 1.5 mg · kg-1 · min-1 (total dose: 186 mg/kg; n = 6 dogs), or 2.3 mg · kg-1 · min-1 (total dose: 282 mg/kg; n = 4 dogs). When sufficiently anesthetized, dogs were placed on a heating pad to maintain core body temperature and endotracheally intubated. Ventilation was mechanically assisted with oxygen (ventilation rate: 7 to 20 breaths/min, inspiratory pressure: 18 to 25 cm H2O). Hemoglobin oxygen saturation, end-tidal carbon dioxide (target: 25 to 55 mmHg), heart rate, and body temperature (target: 37.2° to 38.8°C) were monitored continuously and maintained at target values. Involuntary muscle movements were identified from video recordings and their time, duration, severity (mild, moderate, or severe), and body locations were noted. The severity of each discrete episode of involuntary muscle movement was judged using two criteria: (1) Were there full muscle contractions affecting at least two different body parts? and (2) Did the movements last more than 3 min? Each individual episode was graded as slight, moderate, or severe if it met neither, one, or both criteria, respectively. When necessary to control severe involuntary muscle movements that risked interfering with electroencephalographic signal acquisition and interpretation (and thus seizure detection) during ABP-700 infusion, rocuronium (0.1 to 2.0 mg/kg, IV) was administered. At the end of ABP-700 infusions, dogs were extubated as they emerged from anesthesia and exhibited adequate respiratory function. Midazolam (0.2 mg/kg, IV) and phenytoin (0.2 mg/kg, IV) were administered after the ABP-700 infusion when needed to control repeated episodes of clonic and/or tonic convulsions.
Blood samples (0.5 ml each) were drawn 5, 15, 30, 45, 60, 90, and 115 min after the start of the ABP-700 infusion and then 2, 10, 20, 60, 120, 240, and 480 min and 24 h after the 2-h ABP-700 infusion was completed. These samples were immediately transferred into pre-chilled tubes containing K2EDTA and 16 μl NaF, and then centrifuged for 10 min under refrigeration (4°C; 1,500 g) within 30 min of collection. ABP-700 and CPM-acid concentrations were analyzed by Envigo CRS (USA) using a validated liquid chromatography/tandem mass spectroscopic protocol that had lower limits of quantitation of 1.00 and 20.0 ng/ml for ABP-700 and CPM-acid, respectively. A noncompartmental toxicokinetic analysis of the plasma concentration versus time data was performed by CiToxLAB North America using Phoenix WinNonlin 6.3 software (Certara, USA) to define toxicokinetic parameters.
GABAA Receptor Electrophysiology
Oocytes were harvested from adult female Xenopus laevis frogs using procedures that were approved by the Massachusetts General Hospital Institutional Animal Care and Use Committee (Boston, Massachusetts) and are in accordance with the principles outlined in the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health (Bethesda, Maryland).20 For GABAA receptor studies, oocytes were injected with 1 ng of messenger RNA (mRNA) encoding the α1, β3, and γ2L subunits of the human GABAA receptor at a subunit ratio of 1:1:3. For glycine receptor studies, oocytes were injected with 1 ng of mRNA encoding the α1 and β subunits of the human glycine receptor at a ratio of 1:1. After injection, oocytes were incubated for 18 to 48 h in ND96 buffer (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5 mM 4-[2-hydroxyethyl]-1-piperazineethanesulfonic acid, pH=7.4) containing 0.05 mg/mL of gentamicin and then studied using the whole cell two-electrode voltage-clamp technique.23
Potentiation of 6 μM GABA-evoked GABAA Receptor–mediated Currents by ABP-700 and CPM-Acid
Each oocyte was perfused with ND96 buffer containing 6 μM γ-aminobutyric acid (GABA) alone for 20 s followed immediately by 6 μM GABA plus test compound (ABP-700 or CPM-acid) at the desired concentration for 60 s and the current response was recorded. This GABA concentration was chosen because it produces a current whose peak amplitude is ~5% of that produced by 1 mM GABA (commonly termed an EC5 GABA concentration). To account for variable receptor expression among oocytes, all peak current responses were normalized to the peak current response evoked by 1 mM GABA measured in the same oocyte.
Inhibition of Maximally-activated GABAA and Glycine Receptor–mediated Currents by ABP-700, CPM-Acid, and Etomidate-acid
To maximally activate receptors, each oocyte was perfused for 70 s with ND96 buffer containing 1 mM GABA (for GABAA receptor experiments) or 1 mM glycine (for glycine receptor experiments). Ten seconds into this activation period, the test compound (ABP-700, CPM-acid, or etomidate-acid) was added for 30 s. The inhibitory effect of the test compound on currents was then quantified from the recorded electrophysiologic traces as the reduction in current amplitude at the end of test compound application. To account for desensitization during test compound application, an interpolated straight line was fit between the pre- and posttest compound phases of the current recording period. That line was then used as the baseline against which the effect of the test compound was quantified.
Voltage-dependence of GABAA Receptor Inhibition by ABP-700 and CPM-Acid
Oocytes expressing GABAA receptors were voltage-clamped at the desired transmembrane potential ranging from −110 mV to +10 mV. Receptors were then maximally activated by perfusing each oocyte with 1 mM GABA for 70 s. Ten seconds into this activation period, ABP-700 or CPM-acid was added for 30 s. The inhibitory effect of the test compound on currents was then quantified from the recorded electrophysiologic traces as the reduction in current amplitude at the end of test compound application as described in the previous section.
Impact of CPM-acid on GABAA and Glycine Receptor Agonist Concentration-response Curves
Each oocyte expressing either GABAA receptors or glycine receptors was perfused with ND96 buffer containing CPM-acid for 30 s, followed immediately by CPM-acid plus the desired concentration of receptor agonist (GABA or glycine, respectively) for 15 s and the current response was recorded. Control studies were performed in the absence of CPM-acid. The peak current response obtained in the absence or presence of CPM-acid was normalized to the peak current response evoked by 1 mM agonist (without CPM-acid) measured in the same oocyte.
Concentration-response curves for potentiation were fit using Prism 6.0h software (GraphPad, USA) using its built-in four-parameter equation for stimulation (equation 1):
where minimum is the normalized peak current amplitude in the absence of test compound (ABP-700, CPM-acid, or GABA), maximum is the normalized peak current amplitude at infinitely high test compound concentrations, [test compound] is the test compound concentration, EC50 is the test compound concentration that evokes a peak current amplitude that is half way between the maximum and minimum values, and n is the slope of the relationship.
Concentration-response curves for inhibition of 1 mM GABA-evoked and 1 mM glycine-evoked currents were similarly fit using Prism 6.0h software using its built-in four-parameter equation for inhibition (equation 2):
where maximum is the normalized peak current amplitude in the absence of test compound (ABP-700, CPM-acid, or etomidate-acid), minimum is the normalized peak current amplitude at infinitely high test compound concentrations, [test compound] is the test compound concentration, half-maximal inhibitory concentration is the test compound concentration that evokes a peak current amplitude that is half way between the maximum and minimum values, and n is the slope of the relationship.
In Beagle dog studies, each individual data point represents a single measurement in a dog unless otherwise indicated. In receptor studies, individual data points are reported as the mean ± SD of five separate oocyte experiments whose results were normally distributed as assessed using the Kolmogorov–Smirnov test with the Dallal-Wilkinson-Lilliefor corrected P value. The extra sum-of-squares F test was used to test whether the agonist EC50 or the at maximum peak current at infinitely high agonist concentrations differed significantly in the presence versus the absence of CPM-acid. The uncertainties in fitted parameters are reported as 95% CI. There was no lost or missing data. To avoid output saturation, oocytes producing 1 mM GABA-evoked peak currents greater than 10 μA were discarded. Fitting and statistical tests were performed with GraphPad Prism 6.0h (USA). No a priori statistical power calculations were conducted and the sample size was based on previous experience.24,25 Statistical significance was assumed for P < 0.05.
Toxicologic Studies of ABP-700 in Beagle Dogs
ABP-700 induced and maintained anesthesia in all 14 Beagle dogs at a depth sufficient to allow endotracheal intubation without a paralytic agent within 5 to 15 min of initiating ABP-700 administration. Involuntary muscle movements were observed in all 14 dogs during 2-h ABP-700 infusions and in 9 of the dogs after the infusions were complete. These movements ranged from slight tongue movements to severe whole body shaking. The total number of recorded episodes of involuntary muscle movements in the 14 dogs was 321 (157 in males and 164 in females), with 278 episodes occurring during ABP-700 infusions and 43 episodes occurring afterward. Only one of the episodes of involuntary muscle movements (0.36%) recorded during ABP-700 infusions coincided with electroencephalographic evidence of seizure activity. This occurred in a dog from the 186 mg/kg group that had a seizure 107.5 min after the start of an ABP-700 infusion (i.e., 12.5 min before the end of the 2-h infusion and after receiving 167 mg/kg of ABP-700). The percentages of all episodes of involuntary muscle movement characterized as mild, moderate, and severe were 80%, 14%, and 6%, respectively. There was no clear correlation between the ABP-700 dose and the number or severity of these episodes (data not shown). The total dose of rocuronium administered to a dog during ABP-700 infusion to control severe movements ranged from 1.6 to 4.8 mg/kg (126 mg/kg group), 1.6 to 5.6 mg/kg (186 mg/kg group), and 3.2 to 5.6 mg/kg (282 mg/kg group), respectively. Only one dog did not receive any rocuronium because all of its involuntary muscle movements were slight, and no dog received rocuronium during the final 20 min of ABP-700 infusion. Dogs began to right themselves 17 to 58 min after the conclusion of the 2-h ABP-700 infusions.
Heart rate remained within normal values for this species throughout ABP-700 administration (data not shown). However, figure 2 shows that mean blood pressures progressively decreased in all dogs during ABP-700 infusions from 115 ± 10 mmHg (mean value in all three dosage groups during the one hour prior to beginning the ABP-700 infusions ± SD) to mean minimum values of 88 ± 13 mmHg (126 mg/kg group), 76 ± 9 mmHg (186 mg/kg group), and 65 ± 6 mmHg (282 mg/kg group) by the end of the 2-h ABP-700 infusions. Mean blood pressures returned to baseline values within an hour of completing ABP-700 infusions, paralleling recovery from anesthesia.
Spikes, isolated sharp waves, and repeated sharp waves were identified in the electroencephalographic recordings of all dogs during ABP-700 infusions. Tonic and/or clonic convulsions associated with electroencephalographically-confirmed seizures were observed in 5 of the 14 dogs. A representative electroencephalographic recording of one such seizure is shown in figure 3A to 3C. By group, there were two dogs (one male and one female) in both the 126 mg/kg and 186 mg/kg groups, and one dog (female) in the 282 mg/kg group that had seizures. In one dog from each of the three groups, the first seizure was preceded by premonitory signs including salivation, chewing, excessive vocalization, tremors, head shaking, and/or muscle twitching or contractions. Seizures first occurred 6.5 and 152 min (126 mg/kg group), 81 min (186 mg/kg group), and 743 min (282 mg/kg group) after completion of ABP-700 infusion (fig. 3D). As noted in the first paragraph of the Results section, one dog (186 mg/kg group) had its first (and only) seizure 12.5 min before the end of the 2-h ABP-700 infusion. Of the dogs that had seizures, two dogs had just one seizure and two dogs had three seizures. The fifth dog, which was in the 186 mg/kg group, had a total of 64 seizures over the 24 h that followed ABP-700 infusion. Thus, 71 out of 72 of the electroencephalographic-confirmed seizures occurred during the post-infusion period. These seizures were self-limiting, lasting 10 to 226 s. However, two dogs received midazolam and phenytoin to control convulsive behavior in the post ABP-700 infusion period that the post hoc electroencephalographic review determined were not due to seizures.
For each of the three dosage groups, figure 4A and 4B plot the time-dependent changes in plasma ABP-700 and CPM-acid concentrations, respectively. They reveal that the plasma concentrations of both ABP-700 and CPM-acid generally increased during ABP-700 infusions. Plasma concentrations of ABP-700 and CPM-acid were dose-dependent and achieved peak values near the end of the infusions. After completion of the infusions, plasma ABP-700 and CPM-acid concentrations decreased by 90% on approximate timescales of 10 min and 2 h, respectively. For each of the five dogs that had seizures, figure 4A and 4B also indicate the time of its first seizure. Two dogs (one in each of the 126 mg/kg and 186 mg/kg groups) had their first seizures within minutes of the end of the ABP-700 infusion (6.5 min after and 12.5 min before, respectively) when the plasma concentrations of ABP-700 and CPM-acid were both near their maximum values. The other three dogs had their first seizure 1.4, 2.5, and 12.4 h after both ABP-700 and CPM-acid had reached their peak plasma concentrations.
Table 1 shows the results of noncompartmental analysis of the toxicokinetic data from the three groups for both ABP-700 (table 1A) and CPM-acid (table 1B). The toxicokinetic parameters suggest no large differences between male versus female dogs. The mean (combined male and female data) maximum measured plasma concentrations of ABP-700 were 11,025 ng/ml (35 μM), 14,000 ng/ml (45 μM), and 32,180 ng/ml (102 μM) in the 126 mg/kg, 186 mg/kg, and 282 mg/kg groups, respectively. The mean maximum measured plasma concentrations of CPM-acid were an order of magnitude higher than those of ABP-700 with values of 84,180 ng/ml (282 μM), 142,840 ng/ml (478 μM), and 331,750 ng/ml (1,110 μM) in the 126 mg/kg, 186 mg/kg, and 282 mg/kg groups, respectively.
Potentiation of 6 μM GABA-evoked GABAA Receptor-mediated Currents by ABP-700 and CPM-Acid
We characterized the effects of ABP-700 and CPM-acid over a range of concentrations on currents mediated by α1β3γ2L GABAA receptors expressed in Xenopus oocytes and evoked by 6 μM GABA. We chose to use this low, submaximally-activating GABA concentration for our initial receptor studies because such concentrations commonly allow one to detect the GABAA receptor potentiating and inhibiting actions of test compounds in a single electrophysiologic trace.13 Figure 5A shows representative electrophysiologic traces obtained during ABP-700 studies and demonstrates that the sedative-hypnotic modified 6 μM GABA-evoked currents in a concentration-dependent manner. Within the clinically-relevant ABP-700 concentration range (less than or equal to10 μM), ABP-700 potentiated GABA-evoked currents. The magnitude of this potentiation progressively increased with ABP-700 concentration before reaching a plateau. However, at high ABP-700 concentrations (greater than or equal to 100 μM), electrophysiologic traces also exhibited “surge” currents upon ABP-700 and GABA washout and the peak current amplitude of the GABA response began to decrease with ABP-700 concentration. Together, these phenomena are indicative of GABAA receptor inhibition.13,26 Figure 5B shows analogous electrophysiologic traces obtained during studies of CPM-acid. At equivalent concentrations, CPM-acid produced consistently less potentiation of 6 μM GABA-evoked currents than ABP-700, although surge currents were still observable in electrophysiologic traces at the very high CPM-acid concentrations where potentiation occurred. Figure 5C plots the relationship between the mean peak current amplitude of the GABA response and the concentration of ABP-700 or CPM-acid. For both ABP-700 and CPM-acid, this relationship was bell-shaped with the peak current amplitude first increasing before decreasing with concentration. A fit of the rising portions of the ABP-700 and CPM-acid concentration-response relationships to equation 1 yielded respective EC50s for current potentiation of 2.3 μM (95% CI, 1.6 to 3.3 μM) and 350 μM (95% CI, 120 to 970 μM) and respective maximal peak current amplitudes that were 96% (95% CI, 87 to 105%) and 28% (95% CI, 15 to 41%) of that produced by 1 mM GABA.
Inhibition of 1 mM GABA-Evoked GABAA Receptor-mediated Currents by ABP-700 and CPM-Acid
To better quantify the GABAA receptor inhibitory potencies of ABP-700 and CPM-acid, we defined their actions on GABAA receptor–mediated currents evoked by 1 mM GABA. Under these high agonist conditions, the confounding effect of simultaneous potentiation on the inhibitory action is greatly reduced because most (~85%) GABAA receptors will already have been opened by the agonist prior to the addition of ABP-700 or CPM-acid.27–29 This receptor-saturating GABA concentration is also physiologically relevant because it approximates that achieved at synapses.30 Figure 6A and 6B show representative electrophysiologic traces obtained during such studies. They show that at an intermediate concentration (30 μM), ABP-700 modestly potentiated 1 mM GABA-evoked currents whereas CPM-acid had no measurable effect. At high concentrations, both ABP-700 and CPM-acid inhibited currents in a concentration-dependent manner. Figure 6C plots the relationship between the mean amplitude of the current trace recorded during ABP-700 or CPM-acid application and the ABP-700 or CPM-acid concentration. It shows that ABP-700 modestly potentiated 1 mM GABA-evoked currents when applied at low to intermediate concentrations (3 to 30 μM) but inhibited them at high concentrations (greater than or equal to 300 μM). In contrast, the only measurable action of CPM-acid was to inhibit 1 mM GABA-evoked currents. A fit of the two datasets to equation 2 yielded half-maximal inhibitory concentrations for current inhibition of 770 μM (95% CI, 590 to 1010 μM) for ABP-700 and 1,450 μM (95% CI, 1,340 to 1,560 μM) for CPM-acid.
Voltage-dependence of GABAA Receptor Inhibition by ABP-700 and CPM-Acid
To determine whether this receptor inhibition was voltage-dependent, we quantified the magnitude of inhibition produced by ABP-700 and CPM-acid at different membrane potentials in receptors activated with 1 mM GABA. Figure 7A and 7B respectively plot the relationship between the fractional inhibition produced by ABP-700 and CPM-acid (at their approximate respective half-maximal inhibitory concentrations of 770 μM and 1,500 μM) as a function of membrane potential. These figures demonstrate that GABAA receptor inhibition by ABP-700 was voltage-independent as the slope of this relationship (−0.056; 95% CI, −0.15 to 0.04) was not statistically significantly different from zero (P = 0.247). In contrast, the magnitude of receptor inhibition produced by CPM-acid was voltage-dependent, increasing as the membrane potential was varied from −110 mV to +10 mV. The slope of this relationship was −0.24 (95% CI, −0.38 to −0.09) which was statistically significantly non-zero (P = 0.002).
Impact of CPM-Acid on the GABA Concentration—Response Curve
We sought to further define the mechanism of CPM-acid inhibition of GABAA receptors by assessing its impact on the GABA concentration-response relationship. We chose to focus on CPM-acid rather than ABP-700 because only the former reached plasma concentrations in the dog toxicology studies that were sufficient to measurably inhibit GABAA receptor function. In addition, almost all of the seizures in dogs occurred in the minutes to hours after terminating the 2-h high-dose ABP-700 infusions suggesting that the rapidly formed (and relatively slowly eliminated) metabolite, rather than the parent compound itself, was the causative agent. Finally, animal studies indicate that as is typical for sedative-hypnotic agents, ABP-700 is an anti-convulsant rather than a convulsant.31 Figure 8 plots the relationship between the GABA concentration and the amplitude of the peak current response in the absence and presence of 1,500 μM CPM-acid. A fit of the two datasets to equation 1 yielded essentially identical EC50s of 44 μM (95% CI, 35 to 54 μM) in the absence of CPM-acid and 45 μM (95% CI, 32 to 62 μM) in the presence of CPM-acid (P = 0.9148). However, the maximal peak current amplitudes were statistically significantly different with respective values of 104% (95% CI, 99 to 108%) and 64% (95% CI, 60 to 69%) of that produced by 1 mM GABA (P < 0.0001).
CPM-Acid Also Inhibits the Glycine Receptor
To further evaluate the ways in which CPM-acid might produce seizures in dogs, we assessed whether it could also inhibit the glycine receptor. This is another important inhibitory neurotransmitter receptor in the CNS that is functionally and structurally similar to the GABAA receptor, may colocalize at the same neuronal synapses in the brain and spinal cord as the GABAA receptor thus providing potential redundancy when GABAA receptors are inhibited, and whose inhibition can also produce seizures.32–35 Figure 9A shows representative electrophysiologic current traces demonstrating the inhibitory effect of CPM-acid on electrophysiologic currents evoked by 1 mM glycine and mediated by α1β glycine receptors. Similar to its effects on GABAA receptors, CPM-acid inhibited glycine receptors in a concentration-dependent manner. Figure 9B plots the relationship between the mean amplitude of the current trace during CPM-acid application and the CPM-acid concentration. A fit of the dataset to equation 2 yielded a half-maximal inhibitory concentration for current inhibition of 1,290 μM (95% CI, 1,240 to 1,330 μM).
We then studied the effects of CPM-acid on the glycine concentration–response curve for current activation. Figure 9C plots the relationship between the glycine concentration and the amplitude of the peak current response in the absence and presence of 1,300 μM CPM-acid, an approximate half-maximal inhibitory concentration in the glycine receptor. A fit of the two datasets to equation 1 yielded EC50s of 115 μM (95% CI, 100 to 133 μM) in the absence of CPM-acid and 260 μM (95% CI, 200 to 340 μM) in the presence of CPM-acid (P < 0.0001) and respective maximal peak current amplitudes of 98% (95% CI, 94 to 102%) and 75% (95% CI, 69 to 81%) of that produced by 1 mM glycine (P < 0.0001).
GABAA and Glycine Receptor Function Is Also Inhibited by Etomidate-acid
The structural similarity between CPM-acid and etomidate-acid (fig. 1) suggested to us that the latter may also inhibit receptor function at high concentrations. To test this, we defined the ability of etomidate-acid to inhibit GABAA and glycine receptor function using agonist concentrations of 1 mM GABA and 1 mM glycine, respectively. Figure 10 plots the relationship between the mean amplitude of the current trace during etomidate-acid application and the etomidate-acid concentration. A fit of the two datasets to equation 2 yielded etomidate-acid half-maximal inhibitory concentrations for current inhibition of 7,050 μM (95% CI, 6,300 to 7,890 μM) in the GABAA receptor and 2,930 μM (95% CI, 2,710 to 3,170 μM) in the glycine receptor.
The purpose of the Beagle dog toxicologic studies was to determine whether the involuntary muscle movements that commonly occur during ABP-700 administration are manifestations of seizures. They utilized supra-therapeutic ABP-700 doses and achieved plasma concentrations of ABP-700 and CPM-acid that were on the order of 10× and 100× higher, respectively, than those typically reached during Phase 1 studies in humans.4 Involuntary muscle movements were observed in all 14 dogs and seizures in 5 dogs. However, these two phenomena were temporally and electroencephalographically distinct. The large majority of involuntary muscle movements (278 episodes, 87% of all recorded episodes) occurred during ABP-700 infusion and essentially all of these (99.6%) occurred in the absence of electroencephalographic evidence of seizure activity. Because rocuronium was given intermittently during (but not after) ABP-700 infusion to eliminate involuntary muscle movements when they risked interfering with electroencephalographic signal acquisition and interpretation, this incidence of involuntary muscle movements almost certainly underestimates that which would have occurred during ABP-700 infusion had the paralytic not been used. In contrast, most seizures (i.e., 80% of first seizures and 99% of all seizures) occurred in the minutes to hours after the 2-h ABP-700 infusions were complete. Only one seizure occurred during ABP-700 infusion, and this occurred during the final 12.5 min of the 2-h infusion. These results indicate that the involuntary muscle movements caused by ABP-700 administration are not of epileptic origin and strongly suggest that the underlying mechanisms responsible for ABP-700-induced involuntary muscle movements and seizures are different.
Electroencephalographic spikes and sharp waves were also recorded in the electroencephalographs of dogs during ABP-700 infusion. These are phenomena that are also frequently found in the electroencephalographs of nonepileptic patients receiving etomidate, sevoflurane, or high doses of opiates.36–40 In epileptic patients, numerous sedative-hypnotic agents that act via a variety of different mechanisms produce (or increase the frequency of) spikes and/or sharp waves, including dexmedetomidine, opiates (alfentanil, remifentanil, and fentanyl), methohexital, clonidine, enflurane, and diphenhydramine.41–48 The clinical significance of this pharmacologic action is not completely clear; although, it has been used as a tool to help locate epileptiform foci in the brain during neurosurgical resection.42,49
Involuntary muscle movements are also seen upon administration of other anesthetic induction agents, particularly etomidate where the incidence has been reported to be as high as 70 to 87% in unpremedicated patients.50–52 Although the etiology of etomidate-induced involuntary muscle movements is unknown, it has been hypothesized to result from either a disequilibrium of the drug and/or differential sensitivity to the drug among its various CNS effect sites, and is not associated with seizures.37,53,54 A recent study by our group showed that when the structure of etomidate is modified to abolish its potent and efficacious GABAA receptor positive modulatory activity, the ability of the drug to produce myoclonic muscle movements (in rats) is also eliminated.55 This supports an important role for GABAA receptor positive modulation in mediating this side effect.
What, then, caused the seizures in Beagle dogs during the high dose toxicology studies, and why did they sometimes occur hours after the 2-h ABP-700 infusions were complete? We hypothesized that the seizures were caused by inhibition of GABAA receptors—perhaps together with glycine receptors—by either ABP-700 itself or its principal metabolite, CPM-acid. This hypothesis was based upon the results of previous studies demonstrating that anesthetics and anesthetic-like compounds commonly inhibit GABAA receptors at very high (typically lethal) concentrations along with the data generated by the industry sponsor and reported in this article revealing that ABP-700 and, in particular, CPM-acid reached very high concentrations in these dogs by the end of these high-dose infusions.11–13,56 It was also informed by numerous previous studies showing that inhibition of GABAergic neurotransmission (either chemically or resulting from genetic mutations) causes seizures.57,58
The results of the current studies support this hypothesis for CPM-acid (but not ABP-700) as they reveal that within the plasma concentration range achieved in these dogs, the metabolite inhibits GABAA receptors. Studies utilizing a range of GABA concentrations showed that CPM-acid reduces the maximal peak current response obtained at high agonist concentrations without affecting the agonist EC50. Such insurmountable antagonism implies that CPM-acid inhibits GABAA receptors in a noncompetitive manner, a molecular mechanism it shares with the convulsant picrotoxin.59,60 The magnitude of this inhibition was voltage-dependent, suggesting direct blockade of the ion channel by the negatively charged metabolite.
In addition, our studies showed that at high concentrations, CPM-acid also inhibits glycine receptors. Similar to its effect on GABAA receptors, CPM-acid reduced the maximal peak current response obtained at high agonist concentrations. However, it also increased the agonist EC50 of the receptor, suggesting that there may be multiple CPM-acid inhibitory mechanisms (i.e., competitive and noncompetitive) with this receptor. Such multiple mechanisms may explain why inhibition of glycine receptors by CPM-acid is characterized by such a steep concentration-response relationship (slope: −5.5).
We also found that at high concentrations, the principal metabolite of etomidate (i.e., etomidate-acid) similarly inhibits both GABAA and glycine receptors. However, etomidate-acid is formed in vivo at a much slower rate than CPM-acid. Consequently, when their respective parent drugs are administered to dogs at rates that achieve similar levels of sedation/hypnosis, etomidate-acid reaches plasma concentrations that are only one-twentieth that of CPM-acid.6 Such concentrations would be far below those required to inhibit either GABAA or glycine receptors and cause seizures.
In spite of the frequency with which high-dose ABP-700 infusions produced seizures in the 14 Beagle dogs (36%), no seizures or postictal behaviors were observed and reported in Phase 1 clinical studies involving 90 humans who received ABP-700 as boluses or continuous infusions lasting as long as 30 min.4,5 In addition, an independent analysis of raw frontal electroencephalographic waveforms from bispectral index monitoring recorded from 52 of these volunteers and a subsequent review of such waveforms from an additional 29 volunteers who received fentanyl premedication prior to ABP-700 administration found no evidence of seizures during or after ABP-700 administration even during episodes of severe involuntary muscle movements (Report to The Medicines Company and Addendum, Brad J. Kolls, M.D., M.M.Ci. and David S. Warner, M.D., September 15, 2015). In these Phase 1 clinical studies, typical peak plasma CPM-acid concentrations reached 1,500 ng/ml (5 μM) and the highest concentration measured at any time in any individual was only ~3,000 ng/ml4 (Michel Struys, M.D., Ph.D., F.R.C.A., the Department of Anesthesiology,University Medical Center Groningen, University of Groningen, Groningen, The Netherlands; written communication, April 3, 2018). Such concentrations are just 1% of the peak values achieved in the Beagle dog studies and far below those necessary to inhibit either GABAA or glycine receptors and produce seizures via these receptor mechanisms (fig. 11).
An additional potential contributing factor is that Beagle dogs may be relatively susceptible to seizures. A study utilizing a colony of more than 1,200 Beagle dogs determined that 5.7% of all dogs experienced clinically observable spontaneous convulsive seizures resembling epilepsy.61 This is approximately an order of magnitude higher than the prevalence of epilepsy in humans.62,63 In that Beagle study, an analysis of the pedigree data along with the seizure rates showed a nonrandom incidence that led the authors to conclude that there was a genetic basis for this seizure-susceptibility. Such genetic susceptibility differences may explain why some of our Beagle dogs seized while others did not. Finally (and perhaps most relevantly), Beagle dogs may be relatively sensitive to GABAA receptor–inhibiting compounds as they convulse with lower doses of pentylenetetrazol than Sprague–Dawley rats or cynomolgus monkeys, two other preclinical models commonly used during drug development.21 Such high sensitivity is mirrored in in vitro studies as hippocampal slices from Beagle dogs are more sensitive to pentylenetetrazol-induced changes in population spike area and number than slices from rats, minipigs, and cynomolgus monkeys.64
The U.S. Food and Drug Administration has issued nonbinding guidance for establishing the safety of drug metabolites that provides useful context for the present studies.65 It advocates for the safety testing of drug metabolites prior to initiating large scale human trials if such metabolites reach plasma concentrations that are greater than 10% of total parent drug-related exposure at steady-state. Such concentrations are routinely achieved when administering soft drugs because they are specifically designed to be metabolized rapidly.66,67 Safety testing may include the use of animal models, provided that the chosen model produces adequate metabolite exposure levels. Such levels are defined as being at least as high as those encountered in humans. In the current dog studies, however, plasma metabolite concentrations reached levels that are 100× greater than those achieved in human studies, so high that they inhibited both GABAA and glycine receptors. Thus, it seems highly unlikely that inhibition of these receptors with resultant seizure production is relevant for ABP-700 administration to humans, particularly when such administration is brief.
A limitation to this study is that we cannot directly relate the concentrations of CPM-acid that we found inhibit GABAA and glycine receptors with those present in the CNS of the Beagle dogs because the latter were not measured. We used peak venous plasma concentrations as a surrogate, but peak CNS concentrations may have been higher and occurred later in time. In particular, if ABP-700 was metabolized within the CNS of these dogs, then CPM-acid—which is negatively charged at physiologic pH—could have become trapped within their CNS by the blood brain barrier. With prolonged ABP-700 infusions, this could have led to peak CPM-acid concentrations in the CNS that were even higher than those measured in peripheral venous blood. This is a phenomenon that we previously observed in rats that received 30-min infusions of methoxycarbonyl etomidate, the lead compound for the development of ABP-700 and other soft etomidate analogs.7 In that study, peak metabolite concentrations measured in the cerebrospinal fluid were twice those measured in peripheral blood. Such trapping would cause CPM-acid to accumulate within the CNS during ABP-700 infusion, followed by its slow release from the CNS into the peripheral blood after the ABP-700 infusion was complete. Such a release might explain the slow component of the biphasic decay evident in the CPM-acid toxicokinetic plot that has a half-life of ~5 h (fig. 4B). The resulting disequilibrium between CPM-acid concentrations in the CNS and plasma could explain why seizures sometimes occurred in dogs hours after plasma CPM-acid concentrations had reached their maximum values. A similar disequilibrium has been described for tranexamic acid, a carboxylic acid antifibrinolytic that also inhibits GABAA and glycine receptors, has poor blood–brain barrier permeability, and produces seizures approximately 5 to 10 h after administration has ceased and blood concentrations have peaked.33,68,69 Such a large disequilibrium cannot occur with ABP-700 because it is uncharged, highly hydrophobic, and crosses the blood–brain barrier readily to allow rapid anesthetic induction and emergence.
In summary, high doses of ABP-700 produced both involuntary muscle movements and seizures in Beagle dogs. However, these two phenomena were temporally and electroencephalographically distinct. These results indicate that ABP-700–induced involuntary muscle movements are not of epileptic origin and strongly suggest that the underlying mechanisms responsible for ABP-700-induced involuntary muscle movements and seizures are different. The principal metabolite of ABP-700 (i.e., CPM-acid) reached plasma concentrations during these toxicology studies that are sufficient to inhibit the function of both GABAA and glycine receptors. Such inhibition provides a plausible mechanism for the seizures that occurred in this seizure-prone animal model. It occurs at CPM-acid concentrations that are approximately 100× higher than those achieved in clinical studies of ABP-700 and during which no seizures were observed, strongly suggesting that these actions are irrelevant to the administration of therapeutic ABP-700 doses to humans.
The authors thank Clive Meanwell, Ph.D., and Scott Johnson, M.D., (The Medicines Company, Parsippany, New Jersey) for encouragement in pursuing this project, and Michel Struys, M.D., Ph.D., (University Medical Center Groningen, Groningen, The Netherlands) for his helpful comments and support.
Receptor studies were funded by grant Nos. GM087316 and GM122806 from the National Institutes of Health, Bethesda, Maryland; the Department of Anesthesia, Critical Care, and Pain Medicine, Massachusetts General Hospital, Boston, Massachusetts; and the Department of Anesthesiology, University Medical Center Groningen, Groningen, The Netherlands. Beagle dog studies were initiated and funded by The Medicines Company, Parsippany, New Jersey.
Dr. Raines is the lead inventor of ABP-700. He, his department, his laboratory, and his institution could receive royalties relating to the development of ABP-700 or related analogs. Dr. Raines was the scientific founder of Annovation BioPharma (Cambirdge, Massachusetts) and previously served on the Scientific Advisory Boards of Annovation BioPharma and The Medicines Company (Parsippany, New Jersey). The other authors declare no competing interests.