The authors sought to confirm a chance observation that intravenous lipid treatment increases the dose of bupivacaine required to produce asystole in rats. The authors also measured the partitioning of bupivacaine between the lipid and aqueous phases of a plasma-lipid emulsion mixture.
Anesthetized Sprague-Dawley rats were used in pretreatment (protocol 1) and resuscitation (protocol 2) experiments. In protocol 1, animals were pretreated with saline or 10%, 20%, or 30% Intralipid (n = 6 for all groups), then received 0.75% bupivacaine hydrochloride at a rate of 10 ml x kg x min(-1) to asystole. In protocol 2, mortality was compared over a range of bolus doses of bupivacaine after resuscitation with either saline or 30% Intralipid (n = 6 for all groups). The lipid:aqueous partitioning of bupivacaine in a mixture of plasma and Intralipid was measured using radiolabeled bupivacaine.
Median doses of bupivacaine (in milligrams per kilogram) producing asystole in protocol 1 were for 17.7 for saline, 27.6 for 10% Intralipid, 49.7 for 20% Intralipid, and 82.0 for 30% Intralipid (P < 0.001 for differences between all groups). Differences in mean +/- SE concentrations of bupivacaine in plasma (in micrograms per milliliter) were significant (P < 0.05) for the difference between saline (93.3 +/- 7.6) and 30% Intralipid (212 +/- 45). In protocol 2, lipid infusion increased the dose of bupivacaine required to cause death in 50% of animals by 48%, from 12.5 to 18.5 mg/kg. The mean lipid:aqueous ratio of concentrations of bupivacaine in a plasma-Intralipid mixture was 11.9 +/- 1.77 (n = 3).
Lipid infusion shifts the dose-response to bupivacaine-induced asystole in rats. Partitioning of bupivacaine into the newly created lipid phase may partially explain this effect. These results suggest a potential application for lipid infusion in treating cardiotoxicity resulting from bupivacaine.
CARDIOTOXICITY resulting from bupivacaine has been a subject of intense laboratory investigation for nearly two decades. Areas of particular scientific interest include the mechanism(s)[1–9]and treatment [10–17]of this potentially fatal complication of regional anesthesia. 
We made the chance experimental finding that pretreatment with a lipid infusion increases the dose of bupivacaine required to induce asystole in rats. We then refined this experimental model to specifically test whether lipid pretreatment shifts the dose-response curve for lethal intravenous infusion of bupivacaine. We also measured the effects of lipid-based resuscitation on survival after an intravenous bolus of bupivacaine. Improved survival in this model could suggest a potential means to improve treatment of such toxicity.
Materials and Methods
Racemic bupivacaine hydrochloride was purchased from Sigma Chemical Co. (St. Louis, MO). Tritiated bupivacaine for the partition experiments was purchased from Moravek Biochemicals (Brea, CA). Ten percent and 20% Intralipid[registered sign] were purchased from Baxter Healthcare (Deerfield, IL); 30% Intralipid was a gift of the company. All solutions were warmed before infusion.
The following protocols were approved by the Animal Care Committee and Biologic Resources Laboratory at the University of Illinois. Male Sprague-Dawley rats weighing between 250 and 370 g were used in all experiments. Animals in both experimental models were first anesthetized in a bell jar with isoflurane to allow intubation. All animals were then mechanically ventilated with 1.75% isoflurane in 100% oxygen, using a Harvard rodent ventilator model 680 (South Natick, MA), a tidal volume of 3 ml, and a starting rate of 40 breaths/min. Catheters were inserted into the right internal jugular vein, the right carotid artery, and the right internal iliac vein. Electrocardiogram was monitored via three subcutaneous needle electrodes throughout the entire experiment in all rats. Arterial blood gas measurements were made after induction of general anesthesia and again just before infusions to confirm a carbon dioxide tension between 30 and 35 mmHg and a pH between 7.35 and 7.45.
Protocol 1: Pretreatment with Lipid
All animals were allowed to stabilize for 15 min at 1.75% isoflurane and 100% oxygen. Arterial blood pressure and electrocardiogram were continuously monitored in all experiments. There were six animals in each group. Control animals (group 1) received saline as pretreatment at a rate of 3 ml [center dot] kg [center dot] min sup -1 for 5 min via the internal jugular vein. Three groups of test animals were pretreated with Intralipid, either 10%(group 2), 20%(group 3), or 30%(group 4), also at 3 ml [center dot] kg [center dot] min sup -1 for 5 min via the internal jugular vein. Immediately after pretreatment with either saline or lipid, all animals received an infusion of 0.75% bupivacaine via the internal iliac catheter at a rate of 10 mg [center dot] kg [center dot] min sup -1, to an end point of 10 s of asystole. Blood was then drawn from the aorta into a heparinized syringe for determinations of bupivacaine concentration in plasma. The cumulative lethal dose of bupivacaine was calculated in milligrams per kilogram for all animals.
Concentrations of Bupivacaine in Plasma. Concentrations of bupivacaine in plasma were determined by high-performance liquid chromatography using a mepivacaine internal standard and hexane extraction of all samples. The method of hexane extraction was validated with bupivacaine-spiked samples and yielded > 95% recovery of bupivacaine from both normal and lipemic plasma. Thus, concentrations of bupivacaine in plasma reflect total bupivacaine content in both the aqueous and lipid phases of the specimen. The chromatograph consisted of a M510 pump, a WISP 712 autoinjector, a column oven, and a M490E ultraviolet-visible programmable multiwavelength detector controlled by a Millennium Chromatography Manager (Waters Associates, Milford, MA). Separations took place on a Symmetry C18 column, 5 micro meter, 150 x 3.9 mm ID (Waters Associates). The mobile phase consisted of 25% acetonitrile in 25 mM phosphate buffer adjusted to pH 3.0. The flow rate was 1 ml/min, and column temperature was maintained at 30 [degree sign] Celsius. Retention time was 1.8 min for the internal standard and 4.0 min for bupivacaine. Run time was 6 min. Detection of eluting compounds was at 215 nm.
Protocol 2: Resuscitation with Lipid
All animals in this protocol were anesthetized, instrumented, and stabilized at 1.75% isoflurane as described previously. Arterial blood pressure and electrocardiogram were monitored continuously in each experiment. Each rat received an intravenous dose of bupivacaine over 10 s by Harvard infusion pump via the iliac catheter (see later for doses and the rational for the chosen dose range of bupivacaine). Immediately after the bolus dose, isoflurane was stopped, mechanical ventilation continued was with 100% oxygen, and all animals received an infusion of either saline or 30% Intralipid via the internal jugular catheter. In both cases, the infusion was a 7.5-ml/kg bolus dose over 30 s, followed by 3 ml [center dot] kg [center dot] min sup -1 for 2 min. Chest compressions were given during the period of lipid or saline infusion for any animal experiencing > 15 s of asystole. Survival was scored 5 min after the bolus dose of bupivacaine and required heart rate > 100 beats/min and systolic blood pressure > 60 mmHg. Isoflurane (1.75%) was restarted whenever the blood pressure or heart rate met the survival criteria.
Preliminary experiments with this protocol established the bolus dose ranges of bupivacaine necessary to achieve groups with 100% survival, 100% mortality, and at least one intervening dose for both control and lipid treatment. These were 10.0, 12.5, and 15.0 mg/kg for the controls and 15.0, 17.5, 20.0 mg/kg, and 22.5 mg/kg for lipid-treated animals (Figure 1).
Lipid:Aqueous Partition Coefficient of Bupivacaine. The bupivacaine lipid:aqueous partition coefficient was determined in a mixture of Intralipid and rat plasma. Blood obtained by direct heart puncture during halothane-induced anesthesia was centrifuged and the plasma separated. Equal volumes of 30% Intralipid and plasma ([nearly =] 2.0 ml each) were combined and vortexed. Approximately 1.0 micro Ci of tritiated bupivacaine, specific activity 0.81 Ci/mol, was added to the mixture (final concentration of bupivacaine, 93 micro gram/ml). This was vortexed again and separated into aliquots of 1 ml. These aliquots were allowed to sit undisturbed for 1 h at 38 [degree sign] Celsius and then were centrifuged at 10,000 x g for 10 min. High-speed centrifugation separated each of these mixtures into a clear aqueous phase ([nearly =] 0.85 ml), under a lipid phase ([nearly =] 0.15 ml). The latter comprised a clear layer beneath a thin white cap, which were removed together and then redissolved in saline to a total of 1 ml. Aliquots of this solution and the aqueous plasma phase were then analyzed for tritiated bupivacaine content by liquid scintillation counting. The lipid:aqueous bupivacaine partition coefficient is given by the ratio of concentrations of bupivacaine in the combined lipid phase (after correction for saline dilution) to those found in the aqueous phase. This experiment was performed in triplicate.
Dose of bupivacaine and concentrations in plasma were analyzed by Kruskal-Wallis one-way analysis of variance on ranks. Post hoc testing of both data sets was performed by the Student-Newman-Keuls method for multiple comparisons (SigmaStat; Jandel Scientific, San Rafael, CA). Cumulative dose data for bupivacaine were nonparametic, and median values were compared by differences of ranks. Data regarding concentrations of bupivacaine in plasma were parametric, and differences in mean values were evaluated. Probit analysis (CalcuSyn; Biosoft, Cambridge, UK) was used to compare the values for the dose of bupivacaine required to cause death in 50% of animals (LD50) in the saline and lipid resuscitation protocols. We further evaluated the difference in survival of the two groups at 15 mg/kg bupivacaine using a z test of proportions. Statistical significance in all experiments was set at P <or= to 0.05.
Effect of Pretreatment with Lipid on the Lethal Dose of Bupivacaine Administered by Continuous Infusion
The lethal dose of bupivacaine among all animals in protocol 1 ranged from 12.7 mg/kg (in a control) to 111 mg/kg in an animal receiving 30% lipid. Median lethal doses of bupivacaine were (in milligrams per kilogram, 25th percentile-75th percentile) 17.8, 13.2–20.3 (group 1); 27.6, 22.2–31.7 (group 2); 49.8, 41.3–57.8 (group 3); and 82.0, 71.3–101 (group 4). Statistical significance for differences in median lethal dose of bupivacaine was achieved between all groups (P < 0.001).
Arterial Plasma Concentrations of Bupivacaine at Asystole
The mean concentrations of bupivacaine in plasma at the time of asystole for protocol 1 are (in micrograms per milliliter, +/- SEM) 93.3 +/- 7.6 (group 1); 115 +/- 15 (group 2); 177 +/- 31 (group 3); and 212 +/- 45 (group 4). Statistical significance was achieved for the difference in mean concentrations between groups 1 and 4.
Effect of a Lipid-based Resuscitation on Survival from an Intravenous Bolus Dose of Bupivacaine
Probit analysis of the data from protocol 1 (Figure 1) yields the following bupivacaine LD50values for the two treatment groups, with lower and upper 95% confidence intervals (in milligrams per kilogram): 12.5, 11.8–13.4 (saline); and 18.5, 17.8–19.3 (lipid). A z test of proportions at 15 mg/kg bupivacaine showed significance in the difference in survival between the two groups at this dose (P < 0.004).
Partition Coefficient of Bupivacaine
The lipid:aqueous ratio of concentrations of bupivacaine (+/- SE) is 11.9 +/- 1.77. When equal volumes of 30% Intralipid and plasma are combined, the actual lipid volume is [nearly =] 15% of total, and the percent of total bupivacaine dissolved in the lipid phase of this mixture (+/- SE) is 75.3 +/- 1.32%.
We demonstrate that either pretreatment or resuscitation with a lipid infusion substantially shifts the dose-response to bupivacaine-induced asystole in rats. These observations suggest the potential for a novel treatment of bupivacaine-induced cardiotoxicity.
Intralipid is an emulsion in water of soybean oil, predominantly neutral triglycerides, made isotonic with glycerin. Egg yolk phospholipid, in a concentration of [nearly =] 1%, is the emulsifying agent. The emulsion comprises particles of [nearly =] 0.5 micro meter in diameter. In blood, these fat droplets form a lipid compartment, separate from the plasma aqueous phase, into which a lipophilic substance such as bupivacaine might dissolve. Thus, infused bupivacaine might be drawn into this “lipid sink,” reducing the bupivacaine aqueous plasma concentration to lower than that seen in control animals. The bupivacaine lipid:aqueous partition coefficient of 11.9 strongly supports this model of Intralipid action and suggests that the lipid phase has a high capacity for retaining bupivacaine.
The lipid compartment might exhibit additional, complex pharmacodynamic properties that contribute to the observed benefit from lipid infusion. Lipid might, for instance, preferentially shunt bupivacaine to organs that sequester lipid. We found that relatively low concentrations of bupivacaine in plasma were seen occasionally after high doses of bupivacaine, which suggests that bupivacaine in the lipid compartment is removed from the circulation over time.
Overproduction of nitric oxide might provide another explanation for the beneficial effect of lipid infusion in the treatment of toxicity resulting from bupivacaine. Heavner et al. demonstrated that inhibiting nitric oxide synthesis reduces the dose of bupivacaine required for producing arrhythmias or asystole in rats. Chelly et al. showed that the Intralipid emulsion in preparations of propofol promotes synthesis of nitric oxide. Taken together, these phenomena suggest that an Intralipid infusion might cause an increase in production of nitric oxide, which could reduce cardiac toxicity resulting from bupivacaine.
Lipid infusion might also have beneficial metabolic effects. Bupivacaine impairs production of adenosine triphosphate, [21–23]and Eledjam et al. showed, in isolated myocardial strips, that preincubation with adenosine triphosphate prevents depression of contractility by bupivacaine. Van de Velde et al. found that infusion of 20% Intralipid improves contractility in a dog model of the stunned heart and postulated that this results from improved fatty acid oxidation. Thus, triglyceride infusion could, theoretically, attenuate a component of cardiotoxicity resulting from bupivacaine's to inhibition of adenosine triphosphate synthesis.
The resuscitation protocol provides a stringent test of efficacy in treating bupivacaine-induced cardiovascular collapse. The short, fixed injection interval (10 s) approximates the clinical occurrence of a rapid intravascular injection of bupivacaine; as in a clinical scenario, isoflurane is discontinued to reduce the cardiodepressant effects of a potent inhalational agent on resuscitation. We delayed chest compressions for 15 s of asystole and scored survival 5 min after the injection of bupivacaine. Thus, a difference in survival between control and treated animals requires rapid reversal of the cardiotoxic effects of a potentially fatal dose of bupivacaine.
We found a 48% increase in the LD50value for bupivacaine when resuscitation includes lipid infusion. At 15 mg/kg, survival reflects salvage from what appears to be an otherwise fatal dose of bupivacaine. Nevertheless, the resuscitation data support limited interpretation. We only evaluated recovery of cardiovascular performance in the short term and have no outcome data assessing long-term survival, cardiac performance, or neurologic damage. We did not compare lipid infusion with other treatments of toxicity resulting from bupivacaine. This treatment is not yet optimized regarding either dose or concentration of the lipid infusion. It is also not clear whether this phenomenon is specific to toxicity resulting from bupivacaine in the rat or whether it can be applied to other species or to the treatment of toxicity from other lipophilic drugs. Neither the pretreatment nor resuscitation experiments shed light on the mechanisms of toxicity resulting from bupivacaine.
We report that lipid infusion reduces bupivacaine-associated cardiotoxicity. Partition experiments suggest that the primary benefit of lipid infusion results from a lipid sink effect. Other mechanisms may be active, however. These preliminary observations suggest a potential means for improved therapy of a serious anesthetic complication.
The authors acknowledge David Visintine for expert technical assistance and Dr. Guy Edelman and Dr. William Hoffman for their statistical analysis of the data.