| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
In preclinical pharmacological studies of levobupivacaine (S-bupivacaine), we determined its tolerability, cardiovascular actions, and pharmacokinetics, and we estimated its margin of safety compared with bupivacaine in conscious sheep. Levobupivacaine HCl.H2O was infused IV for 3 min into 10 previously instrumented ewes (approximately 50 kg). On subsequent days, the doses were increased by 50 mg from 200 or 250 mg until fatality occurred. All doses produced convulsions, QRS widening, and cardiac arrhythmias. With incremental doses, 4 of 4 animals survived 200 mg, 7 of 10 survived 250 mg, 3 of 7 survived 300 mg, but 0 of 3 survived 350 mg. Death resulted from sudden onset ventricular fibrillation (n = 3, within 23 min), electromechanical dissociation-pump failure (n = 5, within 45 min), or ventricular tachycardia-induced cardiac insufficiency (n = 2, >10 min). The estimated fatal dose (mean ± SD) was 277 ± 51 mg for levobupivacaine (compared with 156 ± 31 mg found previously for bupivacaine). Pharmacokinetic analysis indicated initial and total distribution volumes = 4.5 (±1.6) and 97 (±22) L, total clearance = 1.7 (±0.4) L/min, and slow half life = 70 (±29) min; these values did not differ from those found previously with smaller doses. Heart and brain tissue levobupivacaine concentrations were approximately 3 times those in arterial blood. The doses of levobupivacaine survived were larger than found previously for bupivacaine, indicating its greater margin of safety.
Implications: Levobupivacaine produced fatal cardiac toxicity at doses significantly greater than those found in previous studies with bupivacaine. As the two drugs have similar potency for producing clinical nerve blocks, the data imply that levobupivacaine should provide a safer alternative to bupivacaine in practice.
The systemic toxicity of local anesthetics predominantly affects the central nervous system (CNS), where they may cause excitation and convulsions, and the cardiovascular system (CVS), where they may cause myocardial depression and conduction abnormalities (1,2). Nancarrow et al. (3) suggest that the risk of serious side effects is greater with bupivacaine than with other contemporary local anesthetics. Bupivacaine is a racemate (an equimolar mixture of the enantiomers R-bupivacaine and S-bupivacaine; S-bupivacaine now has the nonproprietary name levobupivacaine). Levobupivacaine gives a longer duration of intradermal nerve block than R-bupivacaine (4), but similar duration of epidural (5) and brachial plexus blockade (6) when compared with bupivacaine. Various laboratory studies have shown levobupivacaine to be less toxic to both the CNS and CVS than either bupivacaine or R-bupivacaine (711). Our previous study showed that bupivacaine was more toxic to both the CVS and the CNS in sheep when infused IV in doses up to 200 mg (11). Three animals died when infused with 150 or 200 mg bupivacaine, but none died when infused with levobupivacaine at the same doses. However, not having observed death from levobupivacaine administration, we do not know its relative lethal dose or mechanism. This has important implications for clinical practice and for treatment of any toxicity arising from the use or misuse of levobupivacaine. For example, until the cause of death is described, it is not possible to prescribe treatment strategies. Hence, this study was designed to study an extended range of levobupivacaine doses in sheep to ascertain the CNS and CVS effects, to determine the whole body pharmacokinetics, and to estimate the improvement in safety of levobupivacaine compared with bupivacaine.
The study procedures were approved by the local Animal Care and Ethics Committee. Nonpregnant Merino ewes (4358 kg) were chronically instrumented under general anesthesia as described in previous publications (3,1113), with some modifications. Under general anesthesia, a left thoracotomy was performed and the hemiazygos vein, which partially drains the chest wall into the coronary sinus of the sheep, was ligated under direct vision. A transit-time flow probe (6-mm ART2; Triton Technology, San Diego, CA) was placed on the left main coronary artery for measuring coronary arterial blood flow (CABF). Two sonomicrometer crystals (segment length transducer, P/N JP52; Triton Technology) were implanted approximately 1 cm apart in the myocardium of the left ventricular free wall for measurement of shortening of the left ventricular myocardial segment length (SS). A transit-time flow probe (21-mm ART2; Triton) was placed on the pulmonary artery for measuring cardiac output (CO) and a pair of electrocardiogram (ECG) wires (20-pound nylon-coated stainless steel fishing wire) was secured to the dorsal and ventral extremities of the fifth rib. The sheep were allowed to recover 710 days before the cannulation was performed with the aid of an image intensifier. During the cannulation, two catheters were placed in the left carotid artery and advanced into the aortic arch for measurement of mean arterial blood pressure (MABP) and arterial blood sampling. Catheters were placed in the left jugular vein with the tips near the right atrium and in the coronary sinus for drug administration and blood sampling, respectively. After surgery, the animals were accommodated in metabolic crates and were allowed to recover for 710 days before drug studies. The intravascular catheters were kept patent by continuous flushing of heparinized (5 IU/mL) saline (0.9%, 3 mL/h for each catheter) with a high-pressure infusion system. Ten animals were studied. Levobupivacaine HCl.H2O, diluted in 30 mL 0.9% saline, was infused via the right atrial catheter for 3 min. As our previous study found that four of four animals survived 200 mg of levobupivacaine (11), in the present study, only four animals were started from 200 mg of levobupivacaine, and six animals were started from 250 mg. The doses were increased by 50 mg until a fatal outcome occurred. During each study, the sheep were placed in a sling to prevent recumbency during recording. The baseline data were recorded for 5 min before the drug infusions. The recording was continued for 60 min after the start of the infusion, unless the animal died. The procedures were recorded on video tape for subsequent analysis of behavioral effects. Blood samples were collected from the aorta and sagittal sinus at 0, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 5, 7.5, 10, 15, 20, 30, 45, 60, 120, and 180 min after the commencement of the drug infusion; these were submitted for high-performance liquid chromatography analysis of drug concentration (14). Tissue samples were analyzed by using the following modification of the same methodology. After homogenization of tissue in phosphate buffer (2.5 vols, pH 4) containing lidocaine (2 µg) as an internal standard, homogenate (0.4 mL) was alkalinized with NaOH (4 M, 20 µL) to pH 12, then extracted with hexane (1.2 mL). The hexane was extracted with phosphate buffer (100 µL, pH 3), and aliquots (10 µL) of the aqueous extracts were subjected to high-performance liquid chromatography analysis (column: Novapak C18 [Waters Australia Pty Ltd, Sydney, Australia] 150 x 3.9 mm; detection: 210 nm; mobile phase: 29% acetonitrile in pH 6.0 phosphate buffer containing 0.02% triethylamine). Blood gases were analyzed in arterial blood samples drawn at 0, 1, 3, 5, 10, 30, and 60 min.
The analog signals were acquired to disk in digital form by a personal computer interfaced with a physiological monitoring system (System 6; Triton Technology) under Acknowledge 3.03 software (Biopac Inc, Santa Barbara, CA). The acquired and derived data consisted of ECG and ECG variables including P-R interval, QRS complex width, and corrected Q-T interval (Q-Tc = QT/ The entire ECG trace was analyzed for the occurrence of supraventricular tachycardia, ventricular premature contractions (VPCs), bigeminy, trigeminy, ventricular tachycardia (VTac), and multiform ventricular tachycardia (MF-VTac). These data were expressed as the ratio of the number of affected animals to all treated animals for any dose. The mode of death was determined where relevant. The arterial blood drug concentration at the onset of convulsive behavior was determined by inspection. Whole body pharmacokinetic variables (initial dilution volume, total apparent volume of distribution, mean total body clearance, and slow "washout" half-life) were obtained by fitting a conventional three compartment open model to the arterial blood drug concentration-time data by using a weighted nonlinear least squares procedure and standard methods. A weighting factor of concentration (for exponentiation) 1.3 was used as it was found to give a nonbiased distribution of residuals.
Analysis of graded CNS effects of the drug was performed from videotaped records by observing a well defined set of behaviors that were ranked according to morbidity. Logistic analysis was applied to the quantitation of behaviors reflecting the CNS effects of the drug, giving a graded scale designated as a central effect index (CEI) (L. A. Ladd and L. E. Mather, unpublished data, 2000). Values for each behavior, excitement (=20), hypertonia of neck determined by poll position (=26), metapnea (=28), tremor (=30), hypertonia of limb (=32), head bobbing (=33), neck extension (=35), dorsiflexion (=36), defecation (=40), splayed legs (=45), pacing of forelimbs (=50), pacing of hindlimbs (=55), paddling (=65), convulsion (=80), and death (=100), were derived. The CEI score at any time was determined by discerning all currently defined behaviors and finding the highest score among them. The data were expressed as the mean area under the CEI curve and the mean peak CEI for each dose. The onset and duration of convulsive behaviors were defined as the presence of and time between presence and disappearance of behaviors with ranked CEI value
CNS Effects Convulsive behavior was observed with an onset time that decreased with increasing doses (Figure 1). Individual data for the fatal doses are given in Table 1. The overall dose infused to the onset of this behavior ("convulsive dose") was 127 ± 23 mg (mean ± SD, n = 18 doses). The corresponding arterial blood levobupivacaine concentrations at the onset ("convulsive arterial blood concentration") were 16.3 ± 4.0 µg/mL. One animal (ID 102) underwent excitatory effects but did not clearly present convulsive behavior. When analyzed by using severity of CNS effects using the CEI technique, both the peak CEI and the area under the CEI curve were found to increase with dose (Figure 1).
Hemodynamic and Electrocardiological Effects The effects of levobupivacaine on the main hemodynamic and ECG measurements are shown in Figure 2. Clear dose-response relationships were not apparent because the effects were generally secondary to the CNS excitatory effects. Tachycardia occurred at the onset of CNS excitation at all doses. Increased MABP was found with the onset of CNS excitation and generally returned to baseline values within 2025 min. An initial decrease in CO, apparently directly associated with drug administration, was reversed soon after the onset of CNS excitation and was accompanied by parallel changes in CABF. SV was markedly decreased by drug infusions at all doses and gradually returned to baseline within 30 min. Doses of 200 mg did not markedly change SS; larger doses caused depression, after the onset of convulsion, of up to 70%.
QRS widening occurred after every dose. A mean maximum change of 30%40% was found with no clear dose-response relationship, although the duration of effect tended to increase with increasing dose. The P-R interval was decreased up to 65% of the baseline value with the onset of convulsive behavior. No significant changes were found in Q-Tc. Ventricular arrhythmias included supraventricular tachycardia, VTac, and MF-VTac with most doses (Figure 3); sporadic episodes of bigeminy, trigeminy, and VPCs also were noted.
Cause of Death Death was related to cardiac intoxication in every case. Irreversible sudden onset ventricular fibrillation (VF) was the cause of death in three animals from doses of 300 mg (two animals) and 350 mg (one animal); this was inevitably associated with rapid decline in cardiac performance, including sudden decreases in MABP, CO, SV, and SS with death occurring between 2 to 3 min after commencement of the (3-min) drug infusion. Electromechanical dissociation (EMD) was the cause of death in one animal at 250 mg, one animal at 300 mg, and two animals at 350 mg. EMD and VTac-associated failure to fill was the cause of death in one animal. Death from EMD occurred at between 3 and 5 min after commencement of the infusion. Ventricular tachycardia and associated failure to fill was the cause of death in three animals at a dose of 250 mg and occurred at least 10 min after commencement of the infusion. The overall estimated fatal dose of levobupivacaine was 277 ± 51 mg.
Effects on Acid-Base Balance
Pharmacokinetic Analysis The maximum arterial blood levobupivacaine concentrations occurred at the end of the infusion period and were essentially proportional to dose; however, there was some small variability presumably caused by the profound CVS and CNS effects occurring at this time. Maximum coronary sinus blood levobupivacaine concentrations occurred at 4 to 5 min (Figure 5).
Pharmacokinetic analysis in seven animals surviving 10 doses between 200 and 300 mg indicated that the initial dilution volume was 4.5 ± 1.6 L, mean total body clearance was 1.7 ± 0.4 L/min, total apparent volume of distribution was 97 ± 22 L, and slow "washout" half-life was 70 ± 29 min (Table 2). The fit of the model to the drug blood concentration data overall was very good (Figure 5); departures between the model and data, not unexpectedly, were greatest during the period of convulsive behavior. Analysis of arterial blood and tissue concentrations of levobupivacaine (Figure 5) showed that the concentrations in myocardial and brain tissue were several times greater than corresponding blood concentrations from fatal doses. No significant concentration difference (two-way analysis of variance) was found between the various heart (P = 0.36, 49df.) or brain (P = 0.07, 29df.) regions. Not surprisingly, there were very strong correlations between some tissue concentration pairs: especially, the ventricular tissues (all r > 0.96, P < 10-5), and between cortex and cerebellum (r = 0.92, P < 10-4). When stratified by time of death, there was evidence that the drug concentrations in myocardial ventricular tissue were greater in the animals dying from EMD than from VF (Table 1, Figure 5). Neither blood nor tissue levobupivacaine concentrations were correlated with animal body weight.
In this study, levobupivacaine was infused IV as a single dose for three minutes into conscious animals. Compared with bolus injections or continuous IV infusions, this paradigm reasonably simulates the clinical condition of local anesthetic-induced systemic toxicity and allows observation of events during the infusion. Furthermore, these data can be compared with those previously obtained in this laboratory by using a similar experimental design. The use of conscious intact animals precludes any effect from comedication with a general anesthetic. The overall pharmacodynamic events with levobupivacaine are similar to those with bupivacaine (3,11,15); however, there were quantitative differences. The estimated fatal dose of levobupivacaine was 277 ± 51 mg, compared with 156 ± 31 mg for bupivacaine and 263 ± 43 for ropivacaine found in our previous studies with the same paradigm (3). The value for levobupivacaine is significantly greater than that of bupivacaine (P < 0.001), and it is not significantly different to that of ropivacaine (P = 0.61) (one-way analysis of variance, followed by least significance difference test). Hence, these data help establish that there is a clear margin of safety for levobupivacaine over bupivacaine. Convulsive behavior occurred in all animals. The mean "convulsive dose" of levobupivacaine (127 ± 23 mg) is greater than that found previously for bupivacaine [69 ± 12 mg (3); and 85 ± 11 mg (11)] and is similar to that found previously for levobupivacaine [103 ± 18 mg (11)]. It needs to be taken into account that the "convulsive doses" will be smaller if the duration of infusion is decreased (8). The profound CNS effects obscured observation of the direct cardiovascular effects. However, it was clear that two patterns of fatal cardiac effects predominated. The sudden onset of VF occurred in three animals, and EMD with or without VTac (and associated failure to fill) occurred in the remaining animals over longer time courses. The mechanism of death induced by bupivacaine in sheep is mainly VF (3,11,16). Levobupivacaine infusions at this large-dose range produced increased QRS width of the ECG and generated ventricular arrhythmias, principally VTac, MF-VTac, and VPCs. The arrhythmias were similar to those previously found from a smaller dose range of levobupivacaine (11). However, apart from dose, the major difference between bupivacaine and levobupivacaine seems to be that the probability of a fatal malignant arrhythmia from levobupivacaine is less than from bupivacaine (11). Like bupivacaine, levobupivacaine is considered to cause arrhythmias mainly via a direct inhibition of myocardial conduction system. However, it has been found that the direct injection of bupivacaine into the nucleus tractus solitarius in rats (17) and the lateral cerebral ventricle in cats (18) also elicited ventricular arrhythmias, indicating the importance of indirect CNS-mediated arrhythmogenesis by these drugs. Our previous studies with direct infusions of bupivacaine and levobupivacaine into left coronary arteries failed to demonstrate any difference between them in both the nature of, and the dose-response relationship for, arrhythmogenesis (D. H-T. Chang et al., unpublished data, 2000), whereas the CNS toxicity of bupivacaine is significantly greater than that of levobupivacaine. Therefore, the lesser cardiotoxicity of levobupivacaine than bupivacaine may be a result of an enantiomer-selective lesser indirect medullary-mediated arrhythmogenic effect. The initial reduction in myocardial contractility is probably a result of the direct myocardial effects of levobupivacaine. The increases in HR, MABP, CO, and CABF coinciding with the onset of convulsion are similar to those found in previous studies and are considered to be associated with activation of the autonomic system by convulsion (1113,15,19,20). The augmented sympathetic activity may have antagonized the negative inotropic effect by the activation of myocardial ß-adrenoceptors. This viewpoint is supported by the data from the intracoronary injection of levobupivacaine, in which systemic blood concentration of levobupivacaine was not large enough to elicit significant autonomic activity, where consistent myocardial depression up to 60% was observed (D. H-T. Chang et al., unpublished data, 2000). The changes in acid-base balance were greater than those with smaller IV doses of levobupivacaine or bupivacaine in sheep where no animal developed hypoxemia or acidosis from levobupivacaine but where one animal developed similar changes (acidosis and hypoxemia) before death from 150 mg bupivacaine (11). Thus, it is likely that relevant acid-base changes only happen at the larger (potentially fatal) dose range where the acidosis results from the increased muscle activity during convulsion. Convulsion-associated splenic contraction is probably responsible for the observed increases of hemoglobin concentration. The whole body pharmacokinetic findings were remarkable only to the extent that the values were not significantly different from those found with smaller doses. Hence, linearity of levobupivacaine pharmacokinetics over a large-dose range, from essentially trivial 6.25 mg (21) to near fatal (present study), has now been found. Coronary sinus levobupivacaine concentrations closely tracked the arterial concentrations with a 1 to 2 min lag, but were much smaller because of (rapid) uptake into the myocardium. Caution should be applied to interpretation of arterial blood drug concentrations associated with particular behavioral end-points (notably, convulsions and fatality) because of their rapidly changing nature and their disequilibrium between drug concentrations in arterial blood and tissues occurring over the brief period of observation. The levobupivacaine tissue concentrations (Figure 5) were 23 times greater than the concurrent arterial blood concentrations. The tissue drug concentrations from the fatal doses, not surprisingly, show a general trend to washin-washout with time and for the "peak" concentration to occur earlier in the myocardium than the brain. Individual variability in regional drug concentrations presumably has an anatomical-physiological basis, apart from any pharmacological response to the drug. There were no significant regional differences within the two tissue types. However, the right atrial tissue concentrations were the greatest in the three animals that died before the cessation of IV infusion: this may be a result of postmortem contamination of tissue samples with right atrial blood. Animals dying of EMD ± FF, and surviving longer, had larger ventricular tissue concentrations. The differences in response between the animals may be caused by individual differences in susceptibility toward VF and spontaneous reversion. However, it is also possible that the larger doses of levobupivacaine decreased the myocyte excitability, causing EMD instead of VF. The brain, in particular, medullary, levobupivacaine concentrations may be significant because of their possible link to centrally induced cardiac arrhythmias (17,18). In summary, this study found a significantly greater safety margin of levobupivacaine compared with bupivacaine in terms of the fatal IV dose. Like bupivacaine, levobupivacaine can produce fatal ventricular arrhythmias; however, a less malignant disturbance of cardiac electrical conductivity (EMD or VTac associated failure to fill) was found with levobupivacaine than is usually found with bupivacaine (mainly VF), and it occurred later than VF. Clinical evidence suggests that levobupivacaine has an equal anesthetic potency to bupivacaine (2224) with lower propensity for cardiac disturbances (25). Therefore, this preclinical study gives further confidence to the development of levobupivacaine as a single enantiomer replacement for bupivacaine.
Supported by the National Health and Medical Research Council of Australia and Chiroscience R&D Ltd, Cambridge UK. The authors acknowledge the technical assistance of Dr S. E. Copeland, Mr. M. Iglesias, Mr. R. Kearns, Ms. S. Gu, and Ms. B. Fryirs.
Presented, in part, as an oral communication to the European Society for Regional Anaesthesia, 16th Annual Meeting, London, England, September 19, 1997.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|