Population pharmacokinetics of bupivacaine in combined lumbar and sciatic nerve block.

OBJECTIVES: The primary aim of this study was to establish the population pharmacokinetic (PPK) model of bupivacaine after combined lumbar plexus and sciatic nerve blocks and secondary aim is to assess the effect of patient's characteristics including age, body weight and sex on pharmacokinetic parameters.
MATERIALS AND METHODS: A total of 31 patients scheduled for elective lower extremity surgery with combined lumbar and sciatic nerve block using plain bupivacaine 0.5% were included. The total bupivacaine plasma concentrations were measured before injection and after two blocks placement and at selected time points. Monitoring of bupivacaine was made by high performance liquid chromatography (HPLC) with ultraviolet detection. Non-linear mixed effects modeling was used to analyze the PPK of bupivacaine.
RESULTS: One compartment model with first order absorption, two input compartments and a central elimination was selected. The Shapiro-Wilks test of normality for normalized prediction distribution errors for this model (P = 0.156) showed this as a valid model. The selected model predicts a population clearance of 930 ml/min (residual standard error [RSE] = 15.48%, IC 95% = 930 ± 282.24) with inter individual variability of 75.29%. The central volume of distribution was 134 l (RSE = 12.76%, IC = 134 ± 33.51 L) with inter individual variability of 63.40%. The absorption of bupivacaine in two sites Ka1 and Ka2 were 0.00462/min for the lumbar site and 0.292/min for the sciatic site. Age, body weight and sex have no effect on the bupivacaine pharmacokinetics in this studied population.
CONCLUSION: The developed model helps us to assess the systemic absorption of bupivacaine at two injections sites.

Introduction

The efficacy of combined lumbar plexus and sciatic nerve blocks has been proven to be an adequate analgesic and regional anesthetic in the cases that involve surgeries to the lower extremity.[1] This combination, however, requires the administration of large volumes (up to 60 ml) and large doses of long acting amide anesthetics[2] to provide a faster onset of nerve block and to slightly delay the request of pain killers post-surgery.[3] Moreover, due to the proximity of these nerves to deep muscular beds, abnormally rapid absorption or inadvertent intravascular injection could result in systemic toxicity[4] including life-threatening central nervous system and cardiovascular toxicity.[5] Hence, prevention of such complications is of most importance and could be achieved by the supervision of skilled practitioners.[6]

Bupivacaine has been the most widely used long acting anesthetic since the late sixties when it was readily used to achieve regional lumbar plexus and sciatic nerve blocks.[7] However, after an alarming study published in 1979, reporting the severe cardiovascular toxicity induced by bupivacaine,[8] many amide local anesthetics with lesser known toxicity have been developed and compared with bupivacaine including ropivacaine and levobupivacaine, the S-enantiomer of racemic bupivacaine. Although these drugs might present a safer alternative to bupivacaine, the use of the latter is still widely prevalent as it provides a longer lasting effect.[29] Furthermore, in Tunisia, bupivacaine remains the only local anesthetic used for regional anesthesia in adults, mostly due to financial issues and might also be due to patient's satisfaction with the use of bupivacaine.[10] Although numerous studies have reported the pharmacokinetic profile of bupivacaine (given with a vasoconstrictor) in other types of anesthesia, data on bupivacaine plasma concentrations in combined lumbar plexus sciatic nerve blocks is scarce.[11] Therefore, there is an urge to assess the different pharmacokinetic parameters of bupivacaine in the Tunisian population that will eventually lead to the development of an individualized dosage scheme, to put an end to the critical debate on the maximum recommended dosage for local anesthetics[1213] specifically in regards to bupivacaine in combined lumbar to plexus sciatic nerve blocks.

We therefore conducted a randomized trial for lower extremity surgery using combined lumbar plexus and sciatic nerve blocks where the aim was to establish the population pharmacokinetic (PPK) model of bupivacaine (without adjuvant) in the Tunisian population after combined lumbar plexus and sciatic nerve blocks and assess the effect of patient's characteristics including age, body weight and sex on pharmacokinetic parameters, using non-linear mixed effect approach.

Materials and Methods

Chemicals and Reagents

Bupivacaine (BUPIVACAINE® 0,5%), a racemic mixture of its two enantiomers: R (+) bupivacaine and S (−) bupivacaine, was purchased from UNIMED Tunisia. Midazolam (Ipnodis® 5 mg/ml) was provided by MEDIS (Tunisia). Chloroform was obtained from Merck. n-hexane 99% high performance liquid chromatography (HPLC) grade, Isopropanol2 HPLC grade, were obtained from Scharlau (Tunisia). Heparinated plastic tubes were purchased from Plastiques Gosselin Hazebrouck, France. Drug free plasma for analytical development and validation was obtained from a blood bank (H.U.C Charles Nicolle, Tunisia).

Patients

The protocol of this study was reviewed and approved by the committee of medical ethics at Aziza Othmana Hospital, Tunisia.

A total of 31 patients were enrolled in this prospective study for lower extremity surgery.

Patients were labeled as physical status I or II according to the American society of Anesthesiologists. They were given written consents to sign. Exclusion criteria included contraindication to regional anesthesia, known allergy to bupivacaine, hepatic or renal insufficiency and patients treated with drugs which may alter the metabolism of bupivacaine.

Block Placement

An experienced anesthesiology team performed all blocks according to the ethical protocol. All patients were treated in the pre-operative holding area. After application of routine monitors and supplemental oxygen, the patient was sedated using divided doses of intravenous midazolam, 1-5 mg. Using a nerve stimulator connected to a 20-gauge, 100-mm insulated needle, a posterior lumbar plexus block was performed according to the technique initially described by Chayen et al.[14] After a negative aspiration for blood, 30 ml of 0.5% plain bupivacaine (150 mg) was incrementally injected (for 5 min; 30 mg/min). The sciatic nerve block was performed using the same needle and technique with the approach originally described by Labat.[15] Therefore, a negative aspiration for blood, 30 ml of 0.5% plain bupivacaine (150 mg) was incrementally injected (for 5 min; 30 mg/min). Both blocks (lumbar and sciatic nerve) were considered successful if the patient lost motor function within 30 min, defined by the inability to flex the leg at the hip for the first and by an inability to plantar flex or dorsiflex the foot for the second. The time between blocks (from injection of the lumbar-plexus block to injection of the sciatic nerve block was fixed at 30 min. The last needle-removal time (after the second block) was considered time 0.

Three patients had discomfort during surgery and therefore received bolus of ketamine 0.5 mg/kg and/or midazolam 1-2 mg.

Blood Sampling

Venous blood samples of 3 ml were obtained before bupivacaine injection and immediately after blocks placement and at 5, 10, 30, 45, 60, 120 and 240 min after the final injection (in the sciatic nerve site). Blood was collected through a dedicated intravenous catheter placed before surgery in the external jugular vein. The plasma separated by centrifugation of blood samples and stored at −20°C until use. Plasma concentration (μg/ml) was measured at each time point.

Total Plasma Bupivacaine Measurement

Bupivacaine was extracted from plasma according to a previously published method with slight modifications.[16] Briefly, Bupivacaine and internal standard was extracted with a mixture of n-hexane-ispropanol2-chloroform. Bupivacaine quantification in samples was carried out using high performance liquid chromatographic method with ultraviolet absorbance detection (λ =207 nm). In plasma, the lower limit of quantification of bupivacaine was set at 0.042 μg/ml. The linearity was checked from 0.03 to 8 μg/ml (r2 >0.998) with no visible interferences were observed. Technique used presents good precision and accuracy, the within-day co-efficient of variation (CV) of the assay varied from 3.3% to 4.71% respectively. The between-day CV varied from 1.43% to 5.97% respectively. Stability was evaluated according to Food and Drug Administration (recovery was within 98 ± 6.5%).

Statistical Analysis

Normalized prediction distribution errors (npde) (Library npde, version 1.2, November 21st, 2007 manufactured by The Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA) was used for model validation; this analysis was made using R software - package (version 2.9.2) manufactured by The R Foundation for Statistical Computing ISBN 3-900051-07-0.

Non-compartmental analysis

Peak bupivacaine concentration (Cmax) and the time to reach the peak (Tmax) were recorded for each patient and are reported as the mean ± standard deviation (SD).

Non-linear mixed effects modeling (NONMEM) analysis (population approach for parametric analysis)

To assess the bupivacaine pharmacokinetic profile, analysis of plasma concentration data were performed by NONMEM.

Linear pharmacokinetics with the first-order absorption and elimination were considered.

One and two compartment pharmacokinetic models were fitted to the data by nonlinear regression analysis. The choice between one and two-compartment model was made with the Akaike criterion.

One compartment model with two first order absorption rate constants (Ka1 and Ka2 respectively for the lumbar plexus and the sciatic nerve) and central elimination was selected.

The structural model was established initially. The pharmacokinetic parameters were systemic clearance (CL), central volume of distribution (VC = V3), absorption rate constant in the first injection compartment (Ka1 = K13), absorption rate constant in the second injection compartment (Ka2 = K23) [Figure 1].

Figure 1

Diagrammatic representation of the selected model with one compartment. Dose 1: First dose of bupivacaine injected in the lumbar plexus, Dose 2: Second dose of bupivacaine injected in the sciatic nerve site, Ka1: Absorption rate constant (first order) from the first depot compartment. Ka2: Absorption rate constant (first order) from the second depot compartment, Ke: Elimination rate constant from the central compartment. VC: Central volume of distribution

In the second step, we developed the covariate model. Covariates tested were body weight, age and sex.

A proportional inter subject variability parameter was used θi = TV (θi)* exp (ηθi), were θi is the typical value of the parameter and ηθi is the associated intersubject variability parameter with mean zero and variance ω².

Etas (η) were added to CL, to the two absorption constants and to the central volume of distribution.

An exponential term was used to model the residual error.

The population parameters; between-subject variance and residual variance were estimated using the first order conditional estimate with interaction method, the residual intraindividual variability was modeled using the PREDPP subroutine ADVAN 5 of NONMEM VI (level 1.1).

The quality of fit of the pharmacokinetic model was sought by NONMEM's objective function (OF) using the likelihood ratio test and by visual examination of plots of observations concentrations versus predicted ones. The significance of covariates was assessed by the precision of parameter estimates; by the reduction in inter patient and residual variability and change in OF. The decrease in OF considered significant is of 6.63 (P = 0.01).

Only covariates providing a significant change in the OF, when introduced in the model, were retained in the analysis. The final population parameters were estimated considering the relationship with the covariates.

Model evaluation

The stability and performance of the final model was assessed through an internal validation method using the npde for the evaluation of non-linear mixed-models with 3000 simulated.

Results

Patient characteristics are presented in Table 1.

Table 1

Characteristics of patients receiving bupivacaine (n=31)

The observed individual data and the fitted concentration-time curve obtained with population variables (in thick line) is presented in Figure 2.

Figure 2

Observed plasma concentration-time curves of bupivacaine after combined lumbar-plexus and sciatic nerve blocks with total bupivacaine of 300 mg (150 mg in each block, n = 31). Fitted concentration-time curve obtained with population variables is superimposed and presented in thick line

The mean maximal bupivacaine concentration Cmax was 1.77 ± 1.1 μg/ml, it ranges from 0.23 to 4.8 μg/ml and achieved at maximum time Tmax of 37.8 ± 27.6 min after the second injection. They present a large variability between patients, the CV% was >50%. With these concentrations, we did not observe signs of neurologic or cardiac toxicity in any patient. The CV in measured concentrations at each sampling time point was >50% (data not shown). The linear one compartment model with first order absorption, two input compartments and central elimination was selected. Population value of Cmax and Tmax were respectively 1.347 ± 0.754 μg/ml and 29.35 ± 13.52 min after the second injection. T1/2 of bupivacaine was100.43 ± 82.5 min. The selected model predicts a population clearance of 930 ml/min with interindividual variability of 75.29%. The central volume of distribution was 134 L with interindividual variability of 63.40%. The precision (residual standard error [RSE]) of the estimation of these parameters were respectively 15.48% and 12.76% [Table 2]. The two constants of absorption (Ka1 and Ka2) in the lumbar and sciatic nerve sites of injection were respectively 0.00462/min and 0.292/min. A good precision of estimation of absorption of Ka1 and Ka2 was observed; it was 41.55% and 28.42%. Body weight, age and sex have no discernible effects on bupivacaine pharmacokinetics in the studied population.

Table 2

Population pharmacokinetic parameters of the final model with two absorption kinetics of bupivacaine (n=31)

Plot illustrating the npde are shown in Figure 3. The Wilcoxon signed rank test was different from 0 (0.506) indicating a non 0 mean, the fisher variance test was 0.43 indicating a variance different from 1 and the Shapiro-Wilks test of normality was no significant (0.156) indicating that normality of data was not rejected and the distribution present a normal distribution in the final model. This result confirms the validity of selected model.

Figure 3

Graphs plotted by the package normalized prediction distribution errors (“npde”) for R, for the final model. Quantile-quantile plot of the npde versus the expected standard normal distribution (upper left). Histogram of the npde with the density of the standard normal distribution overplayed (upper right). Scatter plot of npde versus observed X (lower left). Scatter plot of npde versus predicted Y (lower right)

Discussion

Bupivacaine, a long acting local anesthetic is used to provide analgesia in surgical patients. Like other local anesthetic drugs, it has a relatively narrow safety margin.[17] A broad interpatient variability was seen, which was not predictable on the basis of clinical characteristics such as the success of the surgical block.

Results from non-compartmental analysis demonstrate an observed mean maximum concentration (1.77 ± 1.1 μg/ml) higher than the reported lowest blood concentrations of bupivacaine (1.1 μg/ml), at which; toxicity (a convulsive episode) has been noted.[18] Despite these high bupivacaine plasma levels we have not detected any signs of neurological or cardiac toxicity in any patient enrolled in our study. This concentration occurred at 37.8 ± 27.6 min after the final injection at the sciatic nerve site. These parameters were important because they gave us an idea about the toxic potential of the bi-block with high doses of amide local anesthetics.

In this study, we have assessed bupivacaine pharmacokinetic profile using non-linear mixed effect approach after combined lumbar and sciatic nerve blocks. We used NONMEM approach to investigate the demographic fixed effect parameters on pharmacokinetic parameters. Population value of Cmax ± SD and Tmax ± SD (1.347 ± 0.754 μg/ml and 29.35 ± 13.52 min) were near to values determined from non-compartmental analysis, which confirms the necessity of vigilance in uses of local anesthetics.

Population values of CL and central volume of distribution were 930 ml/min and 134 L. A good precision of estimation of these parameters was observed, the RSE were respectively 15.48% and 12.76%. This confirms accuracy of final parameters estimates. These parameters present a large interindividual variability (>50%). Total plasma clearance and central volume of distribution of bupivacaine was described previously after epidural administration of bupivacaine in 12 surgical patients,[19] it was averaged 520 ml/min for clearance and 68 L for steady state volume of the central compartment. And following a 10 min intravenous infusion of the race mate (dose 30 mg) in 10 healthy males;[20] the total plasma clearance of R (+)-bupivacaine (395 ± 76 ml/min) was greater than that of S (−)-bupivacaine (317 ± 67 m1/min). The volumes of distribution of R (+)-bupivacaine at steady state (84 ± 29 1) and were larger than that of S (−)-bupivacaine 71 ± 34 1. No significant effect of age, body weight and sex was found on the pharmacokinetic parameters.

The maximum recommended bupivacaine dose (without vasoconstrictor) described is 150-175 mg and if a vasoconstrictor is added, the dose recommended reaches 150-225 mg.[13] Despite the guidance of experts, surprisingly large doses of local anesthetics are sometimes administered by anesthesiologists; in particular, according to the type of block in which multiple injections were practiced. Hence, combination of peripherals blocks requires the administration of large doses of long acting amide anesthetics (up to 60 ml). It was reported that systemic toxicity was directly related to the local anesthetic potency, large dose and rate of administration.[21]

Plasma concentrations data were best fitted using one compartment with two first order absorption; corresponding to lumbar plexus and sciatic nerve sites; P value of Shapiro-Wilks test of normality for npde analysis was >0.05 testifying the validity of the model. None of the previous studies cited in literature reported such data. Although, our results confirm those described by Mouksassi et al.[22] who used a dose of 200 mg of 0.5% bupivacaine in a combined femoral (classical anterior approach) and sciatic (Labatt approach) nerve block in eight patients.

Absorption kinetics of local anesthetics cannot be derived immediately from plasma concentration-time curves after peri neural administration because they exhibit flip flop kinetics. Throughout this study, we were able to quantify and qualify the systemic absorption of bupivacaine from both injection sites. Absorption rate constants of bupivacaine in the lumbar plexus (Ka1 = K13 = 0.0046/min) and in sciatic nerve site (Ka2 = K23 = 0.292/min).

This was due to the proximity of these nerves to deep muscular beds. These parameters present a large interindividual variability; it was respectively 164.92% and 103.44%. The absorption in the sciatic site is slower than in lumbar site. These findings suggest vigilance in the use of this anesthetic when lumbar and sciatic blocks are combined. The rate of absorption into the lumbar plexus or sciatic nerve requires a slow injection of the drug and needs and increases the time between the injections at the lumbar and sciatic sites (>30 min). The systemic absorption and disposition of local anesthetics was described previously after epidural anesthesia.[2324252627] These studies confirm that the systemic absorption of local anesthetics from epidural space was biphasic; a fast and slow absorption process was considered. Indeed Veering et al. analyzed absorption kinetic of bupivacaine in 19 patients, using stable isotope method, its 0.29 ± 0.09 for fraction absorbed during first rapid absorption phase and 0.67 ± 0.11 for the fraction absorbed during the second slow absorption phase. Burm et al. have found similar values; the fraction of doses absorbed in the fast and slow processes was 0.28 and 0.66 and they characterized the bupivacaine absorption by the determination of absorption half-lives in these process, which were 7 min for the fast and 362 min for the slow. The description of the quality of the estimate by the diagnostic graphiques suggests a good fitting of the final model. This pharmacokinetic model was validated using npde package for statistic software R. The simulations confirm that our final model was valid.

Conclusion

We developed a structural PPK model for plain bupivacaine (without adjuvant) in Tunisian population after combined lumbar plexus and sciatic nerve blocks, using non-linear mixed effect approach. The systemic absorption of bupivacaine is rapid in sciatic nerve site. The time between two injections is a key factor in achieving high concentrations of local anesthetics. By increasing time between two injections, we can avoid the cumulation of drug in the blood.