Neosaxitoxin (NeoSTX) is a site-1 sodium channel blocker undergoing clinical trials as a prolonged-duration local anesthetic. Rat sciatic block and intravenous infusion models were used to assess efficacy and local and systemic toxicities for NeoSTX in saline (NeoSTX-Saline), bupivacaine (Bup), and their combination (NeoSTX-Bup). Exploratory studies evaluated the effects of addition of epinephrine to NeoSTX-Bup (NeoSTX-Bup-Epi).
Rats received percutaneous sciatic blocks with escalating doses of NeoSTX-Saline or NeoSTX-Bup. Sensory-nocifensive block was assessed using modified hotplate and Von Frey filaments. Motor-proprioceptive function was assessed by extensor postural thrust. Nerves were examined histologically after 7 days and scored on the Estebe–Myers scale. Median lethal dose was estimated for NeoSTX-Saline and in combinations. Accidental intravenous overdose was simulated in isoflurane-anesthetized, spontaneously breathing rats receiving NeoSTX-Saline (n = 6), Bup (n = 7), or NeoSTX-Bup (n = 13), with respiratory, hemodynamic, and electrocardiographic endpoints. Additional groups received blocks with NeoSTX-Bup-Epi (n = 80). Investigators were blinded for behavioral and histologic studies.
NeoSTX-Bup produced more prolonged sensory and motor block compared with NeoSTX-Saline or Bup. NeoSTX-Bup-Epi further prolonged median time to near-complete recovery for 3 μg/kg NeoSTX-Bup (hotplate: 48 vs. 6 h, P < 0.001). With sciatic injections, addition of Bup did not worsen the systemic toxicity (median lethal dose) compared with NeoSTX-Saline. Intravenous NeoSTX-Saline infusion had significantly longer times to apnea, first arrhythmia, and asystole compared with Bup (P < 0.001 for each). Histologic injury scores overall were low for all groups, with median scores of 0 (interquartile range, 0 to 0) on a 5-point scale.
NeoSTX-Bup and NeoSTX-Bup-Epi hold promise for prolonged-duration local anesthesia.
In rats, combination of Neosaxitoxin with bupivacaine for sciatic nerve block resulted in motor and sensory block, which was longer than either agent alone, and was up to 48 h when epinephrine was added. Histologic examination showed no evidence of neural toxicity, and intravenous injection of Neosaxitoxin resulted in cardiotoxicity with longer delays than bupivacaine.
Neosaxitoxin, a site-1 sodium channel blocker with long duration, neither has previously undergone preclinical safety testing, nor has its interactions with traditional local anesthetics and epinephrine been examined
In rats, combination of Neosaxitoxin with bupivacaine for sciatic nerve block resulted in motor and sensory block, which was longer than either agent alone and was up to 48 h when epinephrine was added
Histologic examination showed no evidence of neural toxicity, and intravenous injection of Neosaxitoxin resulted in cardiotoxicity with longer delays than bupivacaine
NEOSAXITOXIN (NeoSTX) is a site-1 sodium channel blocker under clinical development as a prolonged-duration local anesthetic (see accompanying article).1–4 Site-1 blockers are a family of molecules long recognized for their potent and specific blockade of specific subtypes of voltage-gated sodium channels.5,6 Combinations of the site-1 blocker tetrodotoxin with bupivacaine produced long-duration sciatic nerve blockade in rats without increased systemic toxicity compared with tetrodotoxin alone.7 Subsequent study examined rat sciatic nerve blockade with several members of the saxitoxin series, including NeoSTX.1 Phase 1 studies of subcutaneous infiltration in human volunteers showed that NeoSTX caused cutaneous hypoesthesia and that combination of NeoSTX with bupivacaine resulted in more prolonged analgesia compared with NeoSTX or bupivacaine alone.3
Conventional local anesthetics are associated with local tissue toxicities to nerve8,9 and muscle10 and profound cardiovascular toxicity11–14 in overdose or with inadvertent intravascular injection. Site-1 sodium channel blockers have shown a benign histologic profile after peripheral15 and neuraxial16 administration. In a sheep model, deliberate IV injection of NeoSTX produced minimal cardiovascular effects.17 Bupivacaine cardiovascular toxicity may involve multiple cellular targets18–21 but is likely mediated at least in part by the cardiac sodium channel Nav1.5, which is much more resistant to binding and inactivation by site-1 sodium channel blockers.22 Site-1 sodium channel blockers produce dose-dependent respiratory and skeletal muscle weakness. In anesthetized sheep receiving subcutaneous injection of NeoSTX, coinjection of bupivacaine did not worsen surrogate measures of respiratory or neuromuscular toxicity from NeoSTX.17
The current study further investigates the dose–response of NeoSTX and NeoSTX-bupivacaine combinations on neurobehavioral measures of rat sciatic nerve blockade, as well as on local and systemic toxicities of NeoSTX combinations. A separate IV infusion model further assessed the systemic toxicity using several physiologic endpoints. These experiments were performed as preclinical studies for an Investigational New Drug Application by using NeoSTX formulations manufactured for clinical use in the accompanying phase 1 clinical trial.
Our hypotheses were the following.
At fixed NeoSTX doses, addition of bupivacaine increases the intensity and duration of rat sciatic nerve blockade.
In the presence or absence of bupivacaine, intensity and duration of block increase with NeoSTX dose.
Addition of epinephrine to NeoSTX-bupivacaine combinations in rat sciatic blocks produced further prolonged block durations.
In a model of rapid accidental IV infusion, NeoSTX and bupivacaine and their combinations generate respiratory and electrocardiographic endpoints with different time courses.
The histologic effects of NeoSTX (in saline or in combination with bupivacaine) on rat sciatic nerve are benign over the intended dose range and not statistically different from those of vehicle or untreated nerves.
Materials and Methods
For the remaining discussion and the figures, the terms Bup, NeoSTX-Saline, NeoSTX-Bup, and NeoSTX-Bup-Epi refer to injections performed with bupivacaine, NeoSTX-saline, NeoSTX-bupivacaine, and NeoSTX-bupivacaine-epinephrine, respectively. In all cases where bupivacaine or epinephrine was used for sciatic nerve blocks, their final concentrations were fixed at 2 mg/ml (0.2%) and 5 μg/ml (1:200,000), respectively. NeoSTX concentrations varied as described in each section.
In the sciatic nerve injection model, injectates were prepared on the day of the experiment, and injectate volume was fixed at 0.3 ml. NeoSTX (Proteus SA, Chile) was transported and stored according to the Boston Children’s Hospital safety standards, in compliance with the Harvard Committee on Microbiological Safety.
Drug substance was produced at Proteus SA. Drug product was packaged initially by Biosano Laboratories (Santiago, Chile) (denoted NeoSTX #4) and subsequently by Saval Laboratories (Santiago, Chile) (NeoSTX#5). NeoSTX#4 was used in the phase 2 clinical trial performed in Chile,3 and NeoSTX#5 was used in the U.S. phase 1 studies reported in the accompanying articles. Methods of purification and packaging for NeoSTX#4 and #5 were the same, and bridging studies showed analytical and in vivo equivalence of these two batches. In both cases, NeoSTX drug product is packaged at a concentration of 20 μg/ml in sodium chloride solution, 0.9 mg/ml, at pH 4.5 in 1-ml sealed ampules. Drug assays for different manufacturing, toxicologic, and pharmacokinetic studies used multiple approaches, including high-performance liquid chromatography followed by fluorescence detection and high-performance liquid chromatography followed by tandem mass spectrometry. A series of studies confirmed sterility, stability, and nonpyrogenicity. Absence of cyanobacterial DNA was confirmed by a reverse transcriptase polymerase chain reaction method using positive and negative controls. Absence of cyanobacterial protein or peptides was confirmed by a combined approach using Bradford protein assays, proteomics (mass spectrometry), and amino acid analysis after acid hydrolysis. NeoSTX did not appear mutagenic or carcinogenic in Ames and chromosomal aberration tests. Preparative methods, analytical methods, and Good Laboratory Practices toxicologic studies in rats and sheep were submitted to the U.S. Food and Drug Administration as part of the Investigational New Drug application.
NeoSTX was diluted in 0.9% saline or bupivacaine hydrochloride (Sensorcaine®; APP Pharmaceuticals, USA) before injection. Depending on the intended final NeoSTX concentration, commercial vials of bupivacaine as either bupivacaine 5 mg/ml (0.5%) or bupivacaine 2.5 mg/ml (0.25%) were used to reach final bupivacaine concentrations of 2 mg/ml (0.2%) in the final injectates. In the add-on study, epinephrine was added separately from 1 mg/ml ampules on the morning of each experiment to yield a final concentration of 5 μg/ml.
In the IV overdose model, infusion concentrations were as follows: bupivacaine 2 mg/ml, NeoSTX 1.88 μg/ml, “full-concentration” combination: bupivacaine 2 mg/ml and NeoSTX 1.88 μg/ml, and “half-concentration” combination: bupivacaine 1 mg/ml and NeoSTX 0.94 μg/ml. Infusion rates were adjusted according to animal weight to ensure constant weight-scaled drug delivery rates. Thus, as single drugs, bupivacaine was administered at a rate of 3.2 mg·kg−1·min−1 and NeoSTX at 3 μg·kg−1·min−1. Full-dose combination animals received both bupivacaine 3.2 mg·kg−1·min−1 and NeoSTX as 3 μg·kg−1·min−1, whereas half-dose combination animals received bupivacaine at 1.6 mg·kg−1·min−1 and NeoSTX 1.5 μg·kg−1·min−1, and all animals received a constant weight-scaled fluid administration at rate of 1.6 ml·kg−1·min−1. Infusion concentrations were chosen based on previous estimates of median lethal dose (LD50) for each drug from extravascular (sciatic perineural) injections.
Male Sprague–Dawley rats obtained from Charles River Laboratories (USA): young adults weighing 200 to 250 g were used for the sciatic injection model and 325 to 400 g animals for the IV overdose model. Animals were cared for and sacrificed in compliance with protocols approved by the Institutional Animal Care and Use Committee at Boston Children’s Hospital (Boston, Massachusetts). Handling procedures were developed to habituate animals to the testing paradigm and minimize stress-induced analgesia.
Rats were briefly anesthetized with isoflurane 3% in oxygen delivered in induction chamber for approximately 1 min to facilitate injections. A needle was introduced posteromedial to the greater trochanter, pointing in an anteromedial direction. Once bone (ischium) was contacted, 0.3 ml of solution was injected. The left leg was always used for blocks; the right served as a control and measure of sublethal systemic toxicity (see Systemic Toxicity with Sciatic Perineural Injection section).7
We used a neurobehavioral assessment battery modified from Thalhammer et al., which uses measures of sensory-nocifensive and motor-proprioceptive impairments to assess the duration and intensity of blockade after sciatic perineural injection.23 Investigators were blinded to dose and treatment assignment. Mechanical nocifensive block was assessed using Von Frey (VF) filament testing (Touch-Test Sensory Evaluator; North Coast Medical Inc., USA). After brief habituation to a wire mesh cage, VF hairs of ascending force were applied until observation of paw withdrawal. Care was taken to apply filaments only on lateral plantar surfaces receiving reliable innervation by the sciatic nerve. Filaments were applied in an escalating series until withdrawal was observed or until the maximum of 300 g was reached. Thermal nocifensive block was assessed by time to withdrawal from hotplate set at 56°C, as described in the study by Kohane et al.,7 with a cutoff value of 12 s. Extensor postural thrust (EPT) is a mixed-strength proprioceptive test measured as grams exerted in hind paw push-off on an upright balance.23
Neurobehavioral measures were taken preinjection for a baseline reading. Postinjection values were taken at the following time points: 15 min and 1, 2, 3, 4, 6, 8, 10, and 12 h. After 12 h, measurements were made every 6 h until motor recovery was measured. At each time point, three replicate measurements of EPT force and hotplate latency were taken and averaged. Previous studies of site-1 sodium blockers using this paradigm have shown that higher doses cause transient contralateral impairments of neurobehavioral measures, reflecting systemic analgesia and/or systemic weakness.1,7
In previous studies, injections with 0.3 ml of bupivacaine 2.5 mg/ml resulted in complete block (based on cutoffs defined later in this paragraph) for greater than 98% of animals. In analyses, we therefore make the assumption that incomplete block with test formulations reflects true pharmacologic effects of that dose, rather than a technical injection failure. Baseline measures were assessed before injections. Measures recorded at 15 min were considered to demonstrate peak block intensity.
Cutoff values were 300 g for the VF test and 12 s for the hotplate test.7 Two operational measures of recovery time were derived by linear interpolations for each neurobehavioral test: time to partial recovery (similar to “50% recovery” in previous publications) and time to near-complete recovery. These parameters were selected in part based on previous work7 and in part to foster comparisons with measures used in the accompanying human phase 1 study. The time to partial recovery was defined in the VF test as the time to recovery to a threshold of 150 g, in the EPT test as the time point when values return to 50% of the individual rat’s own baseline, and in the hotplate test as the time to reach a withdrawal latency of 7 s (halfway between cutoff value of 12 s and baseline of 2 s). Near-complete recovery was defined for the VF test as time to recovery to a threshold of 60 g, for the EPT test as time for recovery to 90% of the individual rat’s own baseline, and in the hotplate test as time to reach a withdrawal latency of 3 s (mean baseline latency values 2.0 ± 0.5 s).
Systemic Toxicity with Sciatic Perineural Injection
Sublethal Systemic Toxicity.
Sublethal systemic toxicity was assessed by measurement of right hindlimb sensory-nocifensive and motor-proprioceptive impairments after left hindlimb sciatic injections as described in the Neurobehavioral Testing section. These transient impairments could in principle reflect a combination of systemic analgesic, sedative, and motor effects.
At higher doses of NeoSTX, alone or in combination with bupivacaine, increasing numbers of animals developed apnea or gasping respiration. To minimize distress in this paradigm involving awake animals, any animal developing apnea or gasping was immediately euthanized with intraperitoneal pentobarbital (100 mg/kg), and this was taken as a lethal event. LD50 calculation is described in the Statistical Procedures section.
Systemic Toxicity with IV Infusion.
To model an accidental IV injection, lightly isoflurane-anesthetized, spontaneously breathing rats received infusions via tail vein cannula of drug-containing solution until the endpoint of asystole. A light plane of anesthesia was chosen for this model to permit the study of multiple physiologic endpoints without subjecting awake animals to the distress of an awake-paralyzed situation. Twenty-six rats were randomly assigned to four groups: NeoSTX-Saline (n = 6); Bup (n = 7); full-concentration NeoSTX-Bup combination (n = 7); and half-concentration NeoSTX-Bup combination (n = 6), using the drug concentrations and infusion rates detailed in the section entitled Drugs. Anesthesia was induced by inhalation of isoflurane 3 to 5% in oxygen via induction chamber. The tail vein catheter was flushed with 2 ml of 0.9% saline and connected to a Medfusion syringe pump (Smiths Medical, USA). Vital signs were monitored and physiologic data acquired continuously using PowerLab equipment and LabChart software (AD Instruments, Australia). Baseline measurements were taken (subsequent offline analysis) once all monitoring equipment was calibrated and connected (electrocardiogram, temperature, pulse oximeter, Bain circuit pressure transducer, and tail vein plethysmograph), tail vein accessed, and the rat was maintained in a stable plane of anesthesia at 1% inspired isoflurane in oxygen for at least 5 min. Infusions (as described in the paragraph entitled Drugs) were initiated immediately after a short period of baseline recording and continued until asystole was reached. Primary endpoints for analysis were as follows: (1) apnea (undetectable pressure changes in the Bain circuit) and (2) asystole. Secondary endpoints were bradycardia (heart rate <270 beats/min), deleterious change in electrocardiographic waveform (including either heart block, wide QRS complex, ectopic atrial or ventricular beats, or prolonged QTc interval), and loss of caudal artery pulsatility by plethysmography.
Seven days after sciatic injection, rats were given an overdose of pentobarbital (150 mg/kg) and fixed by transcardiac perfusion in two stages: 100 ml of 0.9% saline was infused, followed by 200 ml of a modified Karnovsky fixative containing 2.5% glutaraldehyde and 1.25% paraformaldehyde in 0.1 M phosphate buffer. The left and right sciatic nerves were dissected and stored in dilute fixative at 4°C. Sciatic tissue was plastic embedded using standard osmium tetroxide electron microscopy protocol, cut to semithin sections, and stained with toluidine blue. Sections were analyzed by an experienced neuroscientist (G.C.), using the scoring system of Estebe–Myers24 ; this neuroscientist remained blinded to group assignments throughout. Estebe–Myers scoring system follows a 0- to 5-point ordinal Likert scale: 0 = no pathology in any portion of the field, 1 = very few myelinated axons with any mild abnormality, 2 = slightly more myelinated axons with any abnormality than 1, but less than 10% abnormal, 3 = 10 to 20% abnormal myelinated axons, 4 = signs of moderate axonal degeneration, and 5 = clear signs of axonal degeneration.24
All measurements are summarized as medians with interquartile ranges (IQRs) or mean ± SD. No data have been lost or missing. Sample sizes with a minimum of eight rats per treatment group were based on our previous experience with 19 rat sciatic block local anesthetic studies over the past 20 yr. Kruskal–Wallis tests following by Dunn correction test were used to compare block intensity at 15 min and time to partial and time to near-complete recovery across doses for NeoSTX-Saline and NeoSTX-Bup groups, whereas Mann–Whitney U tests were used to make comparisons between NeoSTX-Bup and NeoSTX-Saline groups over the range of NeoSTX doses. Two-factor ANOVAs with Bonferroni correction for multiple comparisons were used to conduct an exploratory analysis of the effect of NeoSTX dose and treatment combinations on the time to partial and near-complete recovery for each one of the neurobehavioral test. Spearman rank-order correlation was used to examine the association between NeoSTX dose and recovery time for NeoSTX-Saline and NeoSTX-Bup treatments.
In the IV overdose model, time-to-event data were summarized using Kaplan–Meier curves. Overall log-rank tests, followed by multiple pair-wise comparisons of survival curves to Bupivacaine alone, were conducted for various endpoints, and P values less than 0.017 (Bonferroni corrected for multiple comparisons) were considered statistically significant. Baseline vital signs were compared with ANOVAs. Probit analysis using maximum likelihood was applied to calculate the LD50 for each drug treatment with likelihood ratio 95% CIs obtained by the profile log-likelihood method.25,26 Nerve histology was analyzed with a Kruskal–Wallis model. Statistical analyses were performed using the SPSS statistical package (version 19.0; SPSS Inc./IBM, USA) and SAS (version 9.3; SAS Institute, USA).
Sciatic Block Neurobehavioral Measures
Figure 1 shows the dependence of block intensity at 15 min on the dose of NeoSTX administered in the presence and absence of bupivacaine. All NeoSTX-Bup formulations show complete blockade by all three behavioral measures at that time point. NeoSTX-Saline formulations using NeoSTX doses less than 3 μg/kg produced incomplete block for a majority of animals.
Duration of Block.
Time to partial recovery in hours for all tests and doses is presented in table 1. NeoSTX-Bup formulations produced significantly longer times to partial recovery from thermal and mechanical sensory-nocifensive and motor-proprioceptive blockade compared with NeoSTX-Saline at all doses greater than 1 μg/kg. Time to near-complete recovery in hours for all tests and doses is presented in figure 2. NeoSTX-Bup formulations produced substantially longer times to near-complete recovery from thermal and mechanical sensory-nocifensive and motor-proprioceptive blockade compared with NeoSTX-Saline formulations at all doses 1 μg/kg or greater. Median time to near-complete recovery after injection of Bup was 2.2 h (IQR = 1.9 to 2.9 h) for modified hotplate, 2.7 h (IQR = 2.1 to 2.9 h) for EPT, and 2.0 h (IQR = 1.8 to 2.7 h) for VF testing. Compared with Bup, median time to near-complete recovery after NeoSTX-Bup injection with a NeoSTX dose of 2 μg/kg was substantially longer for modified hotplate (10.8 h; IQR = 9.1 to 17.8 h; P < 0.001), EPT response (22.0 h; IQR = 15.0 to 28.9 h; P < 0.001), and VF response (4.7 h; IQR = 3.0 to 11.0 h; P < 0.001).
An exploratory analysis of the effect of NeoSTX dose and treatment combinations on time to partial and near-complete recovery was conducted using two-factor ANOVAs. All parameters evaluated showed an overall significant independent effect of dose (P < 0.001 for all) and treatment (P < 0.001 for all). The interaction of dose and treatment combination also had a significant effect on the time to partial recovery and near-complete for modified hotplate (partial recovery P = 0.008, near-complete recovery P = 0.050) and EPT only (partial recovery P = 0.002, near-complete recovery P = 0.001). Spearman rank-order correlation demonstrated a significant, moderately strong, positive association between NeoSTX dose and time to partial (EPT: R = 0.60, P < 0.001; hotplate: R = 0.47, P < 0.001; and VF: R = 0.45, P < 0.001) and near-complete recovery for the NeoSTX-Bup group across all neurobehavioral tests (EPT: R = 0.48, P < 0.001; hotplate: R = 0.44, P < 0.001; and VF: R = 0.49, P < 0.001). The NeoSTX-Saline group only showed a significant association between NeoSTX dose and time to partial and near-complete recovery for the hotplate test only (partial recovery: R = 0.35, P < 0.021; near-complete recovery: R = 0.48, P = 0.001). This correlation was generally stronger for the NeoSTX-Bup compared with NeoSTX-Saline.
At a constant NeoSTX dose of 3 μg/kg, addition of epinephrine (NeoSTX-Bup-Epi) showed a significantly longer time to partial and near-complete recovery than NeoSTX-Saline, NeoSTX-Bup, and Bup groups for all parameters evaluated (P < 0.001 for all parameters) (figs. 3 and 4). For NeoSTX-Bup-Epi, the median time to partial (36 h; IQR = 30 to 48 h) and near-complete recovery (48 h; IQR = 30 to 48 h) for VF was at least six-fold longer than that observed for the NeoSTX-Bup, NeoSTX-Saline, and Bup groups, at least three-fold longer for hotplate and at least 1.5-fold longer for EPT (fig. 4).
Transient Contralateral Impairments.
As a marker for systemic drug distribution after sciatic injection, neurobehavioral measures were obtained from the uninjected right limb. NeoSTX-Bup and NeoSTX-Saline combinations only at doses greater than 3 μg/kg produced transient right limb impairments, at 15-min postinjection (fig. 1, D–F). Median duration of right limb impairments lasted less than 2 h across all NeoSTX-Bup and NeoSTX-Saline doses (fig. 2, D–F).
LD50 Testing with Sciatic Perineural Injection.
As NeoSTX-Saline and NeoSTX-Bup doses increased, more animals developed apnea or gasping respiration and thus were subsequently euthanized. LD50 for NeoSTX-Saline was 4.9 μg/kg (95% CI, 4.2 to 6.2) and for NeoSTX-Bup was 5.7 μg/kg (95% CI, 4.9 to 7.9) (fig. 5). NeoSTX-Saline and NeoSTX-Bup LD50 values were not significantly different.
IV Overdose Model.
There were no significant differences in rat vital signs before testing (table 2). Compared with bupivacaine infusion, NeoSTX-Saline infusion had significantly longer times to apnea, first arrhythmia, and asystole endpoints. High-dose NeoSTX-Bup caused asystole and first arrhythmia more rapidly than Bup; however, low-dose NeoSTX-Bup was not significantly different than Bup for any endpoint (fig. 6 and table 3).
Estebe–Myers scoring of nerve injury revealed a very benign histologic profile after sciatic injection. For all treatments, the median Estebe–Myers nerve injury score was 0 (IQR = 0 to 0), with no individual scores greater than 2. Numbers of nerves receiving each treatment were as follows: vehicle: 19, bupivacaine 0.25% plain: 19, NeoSTX-Saline 1 μg/kg: 4, NeoSTX-Bup 1 μg/kg: 8, NeoSTX-Saline 2 μg/kg: 4, NeoSTX-Bup 2 μg/kg: 7, NeoSTX-Saline 3 μg/kg: 19, NeoSTX-Bup 3 μg/kg: 27, NeoSTX-Saline 3.5 μg/kg: 12, NeoSTX-Bup 3.5 μg/kg: 13, NeoSTX-Saline 4 μg/kg: 1, NeoSTX-Bup 4 μg/kg: 6, and uninjected (right side control): 16. There were no statistical differences between any treatment group and noninjected control (right) sciatic nerves. As a validation of the blinded histologist’s readings, slides were obtained from sections of positive control nerves taken from animals who had received deliberate nerve injury (loose ligation model), processed under the same protocol. These positive control nerves all received high injury ratings, with Estebe–Myers scores of 3 or 4. Representative micrographs are shown in figure 7.
In this study, rat sciatic blockade was used to assess the intensity and duration of local anesthetic effect with NeoSTX alone and in combination with bupivacaine, with and without epinephrine. NeoSTX-Saline gave inconsistent and short-duration block at lower doses, and when doses were escalated to greater than 3 μg/kg, there was increasing evidence of systemic as well as local action, as evidenced by transient impairments in contralateral (right) hindlimb neurobehavioral assessments (fig. 1). NeoSTX-Bup produced dense block at all NeoSTX doses, as assessed by three behavioral tests. Increasing NeoSTX doses in NeoSTX-Bup combinations produced longer block durations (fig. 2). In an add-on study using constant NeoSTX dosing of 3 μg/kg, addition of epinephrine (NeoSTX-Bup-Epi) gave further prolongation of blocks compared with NeoSTX-Bup (fig. 4).
Two models for study of systemic toxicity were used. An intramuscular (sciatic perineural) injection model was used to simulate systemic toxicity in an extravascular site, as intended for nerve blockade or infiltration in clinical use. In this model, coinjection of bupivacaine along with NeoSTX does not increase the systemic toxicity compared with animals receiving the same dose of NeoSTX in saline (figs. 1, D–F, and 5). In this model, the bupivacaine dose was constant and always below the range that produces cardiotoxicity. The transient contralateral effects at the 3 and 3.5 μg/kg doses did not significantly differ between the NeoSTX-Saline and NeoSTX-Bup groups, except for a borderline significant difference in the EPT test at 3.5 μg/kg (P = 0.047). In this model, the presumed mode of death is from respiratory muscle weakness. Because NeoSTX-Bup produces more reliable and longer-lasting block compared with NeoSTX-Saline, whereas not increasing the systemic toxicity of NeoSTX, NeoSTX-Bup provides a significant improvement in therapeutic index compared with NeoSTX-Saline.
The second systemic toxicity model used deliberate IV rapid infusion to simulate the effects of accidental intravascular injection (fig. 6). NeoSTX and bupivacaine were each infused either alone, in full-concentration combination, or in combinations with half-concentrations of each of the two components, and time-to-event endpoints (apnea, electrocardiogram deterioration, asystole, and loss of peripheral perfusion) were used as measures of toxicity. In animals receiving IV NeoSTX alone, apnea occurred long before deterioration of the cardiac rhythm, whereas these two events occurred much closer in time with IV bupivacaine alone (table 3). With full concentrations of both drugs, the toxicity was slightly greater (time to event was shorter) than either drug alone. However, with half-concentrations of each drug, the toxicity was not greater than with bupivacaine alone. Interpretations regarding the clinical risks of intravascular injection of the combinations should be cautious. Clinical toxicity of NeoSTX-bupivacaine combinations will probably depend on the injected doses, rates of injection, baseline physiologic status of the patient, and on the speed of resuscitative interventions, including respiratory support, vasoactive medications, medications to terminate convulsions, and treatment of rhythm disturbances.11–14
Previous research has demonstrated that amino-amide local anesthetics produce local myotoxicity10 and neurotoxicity8 that increases with local anesthetic concentration and duration of exposure. The local toxicities of traditional local anesthetics appear not to arise from sodium channel blockade per se, but rather from actions on several other cellular targets.9,16 Several approaches to prolonged local anesthesia involving controlled release of bupivacaine from microparticles or liposomes found local inflammation27,28 as well as myotoxicity and neurotoxicity. Although the clinical significance of these tissue effects for different formulations is open to debate, in our view avoidance of inflammation, neurotoxicity, and myotoxicity is a desirable feature of any proposed approach to prolonged local anesthesia. In contrast to these observations with prolonged delivery of amino amides, previous histologic studies of several site-1 sodium channel blockers have shown minimal injury or inflammatory effects on muscle or nerve.9,16 Even when the site-1 blocker is delivered via liposomes over periods of weeks, the tissue response is extremely mild.15 In the current study, sciatic nerves were assessed by a blinded neuroscientist using an established nerve preparation protocol and scoring system. Overall, the results from these histologic studies gave very low injury scores, providing preliminary support for anticipating benign local tissue effects from NeoSTX with bupivacaine combinations. Additional studies are planned using different staining procedures and different time courses.
A formulation of NeoSTX developed for clinical trials provides prolonged longer-duration sciatic nerve blockade when administered in combination with bupivacaine, with or without epinephrine. Coinjection with bupivacaine improves the reliability and duration of blockade compared with NeoSTX alone, and addition of epinephrine gives further prolongation of block durations. Sciatic nerve histology is reassuring. These data provide further support for proceeding with clinical trials of NeoSTX-bupivacaine combinations, with or without epinephrine, for prolonged local anesthesia.
Support was received from the Boston Children’s Hospital’s Technology and Innovation Development Office and the Sara Page Mayo Endowment for Pain Research, Boston, Massachusetts (to Dr. Berde).
Dr. Berde and his collaborators (Drs. Kohane, Strichartz, and Langer) hold issued patents on site-1 blockers, including Neosaxitoxin, for prolonged-duration local anesthesia. Dr. Berde is Investigational New Drug holder for Neosaxitoxin. Boston Children’s Hospital has a collaboration agreement with Proteus SA (Santiago, Chile) for commercial development of Neosaxitoxin. In the event of future commercial development, Dr. Berde, his coinventors, and Boston Children’s Hospital could potentially receive royalties. Dr. Berde has received no research support, equity, consulting fees, or other income from Proteus SA or any other commercial partner related to this study. The other authors declare no competing interests.