Hemodynamic effects of local anesthetics intoxication . Experimental study in swine with levobupivacaine and bupivacaine

I Assistant Professor, Department of Anesthesiology, Laboratory of Experimental Anesthesia, Center for Experimental Medicine and Surgery, School of Medicine, UNICAMP, Campinas, São Paulo, Brazil. II Fellow Master degree, Department of Surgery, UNICAMP, Campinas, São Paulo, Brazil. III Fellow Master degree, Department of Pharmacology, UNICAMP, Campinas, São Paulo, Brazil. IV Biologist, Center for Medicine and Experimental Surgery, UNICAMP, Campinas, São Paulo, Brazil. V Full Professor, Head Department of Surgery, UNICAMP, Campinas, São Paulo, Brazil


Introduction
High doses of local anesthetic agents are sometimes required for surgery following the administration of local and regional anesthesia and there is always the potential risk for toxic reactions.Both the cardiovascular and central nervous systems are the primary target organs of local anesthetic toxicity in case of inadvertent intravascular injection.Bupivacaine is currently one of the most widely used local anesthetics, due to its quality of anesthesia and prolonged duration of action. 1,2Nevertheless, an Anesthesiology editorial on the severe cardiovascular effects of bupivacaine intoxication was published in 1979, 3 and since then research has focused on discovering new longacting local anesthetics with lower toxicity.Although bupivacaine is synthesized in the form of its two dextrorotatory R(+) and levorotatory L(-) isomers 4 , until recently it had only been marketed as a racemic mixture, containing 50% of each of the two enantiomers.However, evidence that the levorotatory isomer caused less toxicity dates back to 1972. 5,6Levobupivacaine, the levorotatory isomer, has recently begun to be more widely studied.In some animal models, the lethal dose of levobupivacaine was shown to be up to 1.6 times higher than that of the racemic mixture 7 !In humans, levobupivacaine would be less potent at producing negative inotropic effects and prolonging the PR and QT intervals of the ECG, typical of racemic mixture intoxication 8 .However, the drawback with this new agent would be its less intense motor block 10 , despite having equal analgesic potency to racemic bupivacaine. 9The aim of this study was to simulate acute intoxication with either agent in pigs, as might accidentally occur during local and regional anesthesia with high doses of local anesthetics and to evaluate its hemodynamic repercussions.

Methods
After approval from the Ethics Committee for Animal Experiments, forty healthy Large White pigs (both sexes, body weight 20-27 kg) underwent the following protocol: 1.The pigs were fasted on the night before the procedure and had free access to water.
2. In the morning of the procedure, they were weighed, an auricular vein was cannulated and anesthesia was induced with a 25 mg.kg -1 IV dose of sodium thiopental 2.5% solution 11 .
3. The body surface area of the animal (BSA) was calculated in square meters using the classic formula from the literature 12 : BSA= (9 x weight in grams 2/3 ) x 10 -4 , introducing the value into the Engstrom AS/3 multiparametric monitor for calculation of body index values.
4. A tracheal tube was inserted and connected to a pneumatically driven ventilator using a partial re-breathing system and CO 2 absorbent.Tidal volume was 15ml.kg - and respiratory rate was adjusted to maintain ETCO 2 at 32 -34 mmHg.Fresh oxygen flow was 1 l/min and hemoglobin oxygen saturation was measured, placing the sensor on the animal tongue, aimed at achieving a value above 97%.ECG was also monitored.
6. Subsequently, an incision was made along the inner aspect of the thigh under local anesthesia (5 ml 1% lidocaine hydrochloride without vasoconstrictor) for catheterization of the femoral artery and continuous measurement of arterial blood pressure.Through the same incision, the femoral vein was dissected and a 7F Swan-Ganz catheter was inserted.The catheter was advanced into a branch of a pulmonary artery and proper catheter placement was confirmed by the pressure waveform obtained.Cardiac output (CO) was then measured by the thermodilution method.In addition, mean arterial pressure (MAP), mean pulmonary artery pressure (mPAP), central venous pressure (CVP) and pulmonary capillary wedge pressure (PCWP) were measured.Using classic formulas in the literature, the monitor also performed calculations of the remaining hemodynamic parameters, e.g.: cardiac index (CI), stroke index (SI), systemic vascular resistance index (SVRI), pulmonary vascular resistance pulmonary index (PVRI), left ventricular stroke work index (LVSWI), right ventricular stroke work index (RVSWI).At this early phase, blood samples were drawn from the animals for dosage of hematocrit and hemoglobin levels.
7. After a resting period of about 30 minutes to ensure stabilization, baseline hemodynamic measurements (T 0 ) were recorded.
8. Then the animals were randomly divided into 2 groups in a double-blinded fashion: bupivacaine group (B) and levobupivacaine group (L).Each group was intravenously injected with a toxic dose of 4 mg.kg - of either local anesthetic 13 during 30 seconds and the experimenter was blinded to the study code.
12. At the completion of the experiment, each animal was sacrificed while under anesthesia by a 10 ml intravenous injection of 19.1% potassium chloride solution.
Subsequently, the double-blinded protocol was disclosed and data was statistically treated.Categorical variables received statistical treatment using the chi-square test.Variables without normal distribution were logtransformed for the tests in order to reduce skewness and variability.To compare the distribution of single numerical variables (measured at a single time point), the Student's ttest was used.To study the behavior of the hemodynamic parameters measured at several time points, analysis of variance (ANOVA) for repeated measures was performed, followed by Duncan's post-hoc test for multiple comparisons to compare the groups at each time point (between groups comparisons), and by the profile test by contrasts to analyze the changes in the parameters of each group (within groups comparisons).The significance level was set at 5%, i.e., p < 0.05.

Results
The following table demonstrates the distribution by sex, mean and standard deviations of weight, hematocrit levels, hemoglobin levels and body surface areas found in both groups.Following intoxication, there was a reduction in cardiac index (Figure 1 and Table 1) in both groups and these values did not return to those similar to resting values until the end of the experiment (p<0.001).The decrease in L was statistically more important than in B and was maintained until T 4 (p<0.001).
to those similar to resting values at T 3 in group B and at T 4 (p<0.001) in group L. Starting from T 1, values in L were significantly lower than those in B until T 4 (p<0.001).

FIGURE 2 -Mean Arteria
There was a significant decrease in heart rate (Figure 3 and Table 3) following intoxication in both groups, although the reduction was more considerable in E and this difference was maintained until the end of the experiment (p<0.001).In both groups, HR did not return to values comparable to resting values until T 6 (p=0.003).
Central venous pressure (Figure 4 and Table 4) increased significantly in both groups and the increase was     6) showed increased values in both groups following intoxication and these values were maintained until the end of the experiment (p<0.001).No difference between the groups was found.7) demonstrated a significant decline.In both groups, these values remained lower than resting values until T 4 (p<0.001).The fall in L, however, was significantly more important at T 1 and T 2 (p=0.019).

FIGURE 7 -Stroke Index
Systemic vascular resistance index (figure and table 8) in group B increased, maintaining values higher than resting values until T 6 .In L, SVRI initially showed a decrease followed by a significant increase beginning at T 2 and maintained values higher than resting values until T 6 (p<0.001).B values were higher than L values at T 1 and T 2 (p=0.004).10) showed a decrease after intoxication in both groups.However, the fall was more significant in L, and this difference was maintained until T 3 (p=0.001).Values returned to those similar to resting values at T 4 in group B and T 5 (p<0.001) in group L.   11) also showed a decrease after intoxication in both groups.The fall was more substantial in L and a significant difference between the groups was maintained from T 1 to T 4 and at T 6 (p=0.001).Values returned to those similar to resting values at T 3 in B, and only at T 5 (p<0.001) in L.

Discussion
The cardiac toxicity of local anesthetics is attributed to a blockade of sodium channels in the heart, leading to a prolonged conduction time with widening of QRS complexes, prolongation of PR interval, AV block and arrythmias 14 .Furthermore, these drugs have the ability to cause cardiac toxicity by altering mitochondrial metabolism in cardiac cells and thus alter inotropism 15 .The S(-) isomer may be less cardiotoxic due to its lower affinity for cardiac sodium channels, in comparison to the R(+)isomer, as demonstrated in guinea pigs 16 .Such data, however, should be viewed with some reservation before it can be extrapolated to humans.The overall pharmacodynamic events seen with levobupivacaine are similar to those seen with bupivacaine but we have found quantitative differences 17 .Other authors have also compared levobupivacaine, racemic bupivacaine and ropivacaine in animals, demonstrating that levobupivacaine had a similar 18 or even greater toxicity. 19,20In agreement with those authors, our results showed that levobupivacaine had greater hemodynamic repercussions on a swine model of acute intoxication, simulating what may occur after accidental intravenous injections of local anesthetics during local and regional anesthesia.Evidence of toxicity was shown in the significantly more important decrease in cardiac index, mean arterial pressure, heart rate and left ventricular stroke work index.In a more homogeneous model, these results confirm those previously obtained with a preparation containing 75% of the levorotatory isomer in dogs 21 .Results obtained in animals should be viewed cautiously and prudence is of the essence when extrapolating these data to humans.However, further comprehensive studies must be encouraged, since results of this nature are not the first to be observed.Accidental injections of high doses of local anesthetics and toxic reactions during local and regional anesthesia have decreased in the last 30 years, falling from 0.2 to 0.01%.Peripheral nerve blocks still account for the majority of these cases (7.5 per 10.000) 22 .Such considerations should intensify new efforts to discover drugs and techniques that enable us to achieve a significantly low morbidity and mortality rate, protecting patients from the undesirable and unpredictable effects of local and regional anesthesia techniques with high doses of local anesthetics.

Conclusion
Levobupivacaine was more cardiotoxic than racemic bupivacaine when large doses were injected intravenously in swine, as may accidentally occur during local and regional anesthesia.

FIGURE 10 -
FIGURE 10 -Left Ventricular Stroke Work Index

FIGURE 11 -
FIGURE 11 -Right Ventricular Stroke Work Index

TABLE 1 -
2ardiac index l.min - .m2The goups were homogeneous regarding distribution by sex (p=0.736),weight (p=1), hematocrit level (p=0.334), hglobin level (p=0.099) and body surface area (p=0.95).There was no difference in any of the hemodynamic parameters measured at rest between the groups.
Mean arterial pressure (Figure2and Table2) decreased significantly in both groups; values returned

TABLE 2 -
Mean arterial pressure mmHg

TABLE 4 -
Central venous pressure cmH 2 O FIGURE 4 -Central Venous PressureMean pulmonary artery pressure (figure and table 5) showed no significant alterations or differences between both groups.FIGURE 5 -Mean Pulmonary Arterial Pressure

TABLE 5 -
Mean pulmonary artery pressure mmHg Pulmonary capillary wedge pressure (figure and table

TABLE 6 -
Pulmonary capillary wedge pressure mmHg FIGURE 6 -Pulmonary Capillary Stroke index (figure and table

table 9 )
showed values higher than resting values from T 2 to T 5 in both groups (p=0.001) . No difference was found between B and L. FIGURE 9 -Pulmonary Vascular Resistance Index Left ventricular stroke work index (figure and table