Epinephrine is often added to local anesthetics to decrease systemic absorption by regional vasoconstriction. This reduces the risk of local anesthetic toxicity when large doses are given, for example, during neural plexus or epidural blocks. Because ropivacaine has intrinsic vasoconstrictor properties (1–7), the addition of epinephrine has been considered unnecessary. However, few pharmacokinetic data are available on the effect of the addition of epinephrine to epidural ropivacaine in humans. Therefore, we designed this prospective, randomized, double-blinded study to investigate the effect of the addition of epinephrine on the early systemic absorption of ropivacaine given epidurally to women undergoing abdominal hysterectomy.
After approval from the clinical research ethics committee of the Chinese University of Hong Kong and with written informed consent, we recruited 24 ASA physical status I and II women having elective abdominal hysterectomy under combined epidural and general anesthesia. These women were aged between 30 and 70 yr old, with body weight ranging from 45 to 60 kg, and height ranging from 145 to 165 cm tall. Patients were excluded if they had any bleeding tendency, systemic or local infection, renal or hepatic disease, vascular or vasoactive disorders, known allergy to local anesthetics, were taking β-blockers, or were smokers.
Patients were not given premedication. On arrival into the operating room, routine monitoring was applied, and baseline hemodynamic variables were recorded. An IV catheter was inserted for drug and fluid administration, and IV prehydration of 20 mL/kg of lactated Ringer’s solution was given. A radial artery catheter and a second venous catheter were inserted in the contralateral forearm for blood sampling. With the patient in the lateral position and with aseptic precautions, the epidural space was identified with a 16-gauge Tuohy needle using a loss-of-resistance technique after skin infiltration with lidocaine at the L2-3 to L3-4 interspace. An epidural catheter was then inserted 4 cm into the epidural space. The epidural catheter was aspirated, inspected for blood or cerebrospinal fluid, and secured, and the patient was returned to the supine position.
General anesthesia was then induced with IV thiopental 3–5 mg/kg or propofol 2–3 mg/kg. Neuromuscular block was achieved using atracurium or rocuronium, and the trachea was intubated. Anesthesia was maintained using isoflurane adjusted to an end-tidal concentration of 0.5%–1% in nitrous oxide and oxygen.
Immediately after the induction of general anesthesia, patients were randomly allocated to receive an epidural injection of 1.5 mg/kg of plain ropivacaine (plain group, n = 12) or 1.5 mg/kg of ropivacaine with epinephrine 5 μg/mL (epinephrine group, n = 12) diluted to a volume of 15 mL with saline. The randomization was performed by the drawing of sequentially numbered, opaque sealed envelopes that contained a computer-generated random number. All doses were prepared by an anesthesiologist who took no further part in patient care or data collection. Three milliliters of the solution was initially injected epidurally after a negative aspiration test. Three minutes later, if there was no major hemodynamic disturbance, the remainder of the dose was injected over 3 min. Intraoperatively, hypotension was treated with IV fluids and IV ephedrine as required. Intraoperative blood loss was measured by weighing wet swabs and estimating the volume of blood in the operative suction bottle. Venous blood samples were taken for measurement of hemoglobin concentration at the commencement and completion of surgery.
The time of completion of the epidural injection was considered time zero. Two-milliliter samples of arterial and venous blood were simultaneously collected immediately before the injection of ropivacaine and at times 1, 2.5, 5, 7.5, 10, 15, 20, 25, 30, 40, 50, 60, 75, 90, 120, 150, and 180 min. Five milliliters of blood was withdrawn as dead-space from the sampling catheters before each sample collection. Blood sampling was continued in the postanesthesia care unit as required.
Each blood sample was collected into lithium heparin tubes, mixed gently, and centrifuged at 3000 rpm for 10 min at room temperature. The plasma was then separated and stored at −70°C until batch assay. Total plasma concentrations of ropivacaine were determined using high-performance liquid chromatography with UV detection at 210 nm. The calibration curve for the assay was linear over the range 0.01–3 μg/mL using four standards with a correlation coefficient of r = 0.9986. The lower limit of detection was 0.01 μg/mL. The intraassay coefficients of variation were 1.4% at 0.1 μg/mL and 5.3% at 2 μg/mL, and the interassay coefficients of variation were 4.4% at 0.1 μg/mL and 8.1% at 2 μg/mL. The mean relative extraction efficiency ranged from 78.4% to 88.9% for concentrations between 0.05 and 1 μg/mL.
At the completion of surgery, residual neuromuscular block was antagonized with neostigmine and atropine, and the patient was transferred to the postanesthesia care unit. Postoperative analgesia was provided using a continuous epidural infusion of a dilute ropivacaine-fentanyl mixture that was started only after the final blood samples had been collected.
Prospective power analysis based on previously published data (8) indicated that a sample size of 12 patients per group would have 80% power to detect a 25% difference in peak arterial plasma concentration of ropivacaine at an α-level of 0.05. Results are presented as mean ± sd unless otherwise stated. Arterial and venous plasma concentration-versus-time curves over the 180-min sampling period were plotted. Peak plasma concentration (Cmax) and time to Cmax (Tmax) were obtained directly from the observed plasma data.
Individual Cmax and Tmax were obtained for each patient for both arterial and venous plasma concentrations. Cmax and Tmax were analyzed by repeated-measures analysis of variance (rm-ANOVA), with sampling site (artery versus vein) as a within-subjects factor and group (epinephrine versus plain) as a between-subjects factor. Serial plasma concentrations were analyzed by 3-factor rm-ANOVA with group (epinephrine versus plain) as the between-subjects factor, and time and sampling site (artery versus vein) as the within-subjects factors. The Greenhouse-Geisser correction was applied where appropriate (9). Any significant interaction term was examined by analysis of simple main effects. Differences were considered significant if P < 0.05. Analyses were performed using SPSS (version 10) for Windows (SPSS Inc, Chicago, IL). Differences between arterial and venous ropivacaine concentrations over time were examined by fitting a population averaged generalized linear model to the arterio-venous concentration difference using a normally distributed error term, an exchangeable correlation structure, and a logarithmic link function (10). This model fitted a monoexponential decline to the arterio-venous concentration difference while allowing for correlation of observations within subjects. This model was fitted using the software Stata, version 7 (Stata Corporation, College Station, TX).
All patients completed the study. Patient height and duration of surgery were greater in the epinephrine group compared with the plain group (both P = 0.04); however, other patient characteristics and total blood loss were similar between groups (Table 1).
Plasma Cmax and Tmax are shown in Figures 1 and 2 and Table 2. Cmax and Tmax were obtained directly from the observed plasma data. Because the variability of values for Tmax became greater with increasing magnitude, a logarithmic transformation was applied to these data before analysis. The rm-ANOVA showed no evidence of any interaction between treatment group (epinephrine versus plain) and sampling site (artery versus vein) for Cmax (P = 0.35) or Tmax (P = 0.74), indicating that any effect of epinephrine was similar for both arterial and venous samples.
Cmax was significantly smaller (mean difference, 0.44; 95% confidence interval [CI], 0.12 to 0.76 μg/mL;P = 0.01) in the epinephrine group (arterial, 0.92 ± 0.32 μg/mL; venous, 0.82 ± 0.33 μg/mL) compared with the plain group (1.31 ± 0.39 μg/mL and 1.31 ± 0.50 μg/mL, respectively; both P = 0.01). There was no significant difference between the arterial and venous values for Cmax (mean difference, 0.05; 95% CI, −0.03 to 0.13 μg/mL; smaller in venous than in arterial plasma).
There was a trend towards greater Tmax in the epinephrine group compared with the plain group: geometric mean of 16 ± 2 min for arterial and 23 ± 2 min for venous blood in the epinephrine group versus 9 ± 2 min and 12 ± 3 min, respectively, in the plain group (mean difference factor, 1.9; 95% CI, −0.96 to 3.7). Tmax was greater in venous than in arterial plasma by a factor of 1.4 (95% CI, 1.1 to 1.7;P = 0.005).
Plasma concentration-time profiles of ropivacaine are shown graphically in Figure 3. Analysis using rm-ANOVA revealed both the 3-factor interaction (group × sampling site × time;P = 0.84) and the group × sampling site interaction (P = 0.90) to be nonsignificant. However, the group × time (P = 0.002) and sampling site × time (P = 0.001) interactions were both significant. To interpret these interactions, differences in mean plasma concentrations between groups and between sampling sites were compared separately for the time periods 0–60, 60–120, and 120–180 min. For the interval 0–60 min, both arterial (P = 0.005) and venous (P = 0.008) concentrations were significantly smaller in the epinephrine group compared with the plain group (two-sample t-tests). However, for the periods 60–120 and 120–180 min, neither arterial nor venous plasma concentrations differed significantly between groups.
Similarly, arterial plasma concentrations of ropivacaine were significantly larger than venous concentrations in both the epinephrine (P = 0.001) and plain (P = 0.01) groups for the first 60 min (paired t-tests), with no significant differences thereafter. Figure 4 shows the population averaged exponential decline model fitted to the arterio-venous concentration difference for all patients. From the model, we estimated that the mean arterial concentration was initially 0.16 μg/mL larger than the mean venous concentration; the difference declined over time with a half-life of 24 min. Inclusion of predictors denoting treatment group (epinephrine versus plain) or its interaction with time in the model did not lead to significant improvement (P = 0.66 and P = 0.28, respectively), providing no evidence that epinephrine affected the magnitude or rate of decay of the arterio-venous difference. The estimated effect of epinephrine was a nonsignificant reduction in the arterio-venous difference by a factor of 0.97 (95% CI, 0.67 to 1.42).
Epinephrine is one of the most common additives used to modify the pharmacodynamic and pharmacokinetic properties of local anesthetics. Absorption of local anesthetics from the epidural space is biphasic; by causing local vasoconstriction, epinephrine decreases the fraction absorbed in the fast component, thus reducing the Cmax (11,12). As the absorption rate constants are unchanged, the Tmax may not be prolonged (11,12). Although the effect of epinephrine is generally less pronounced with longer-acting compared with shorter-acting local anesthetics (11), our study showed that for ropivacaine, the addition of epinephrine significantly reduced Cmax and mean plasma concentrations of ropivacaine in the first hour in both arterial and venous plasma after epidural injection. This suggests that, in clinical usage, the addition of epinephrine may decrease the risk of systemic toxicity and increase the margin of safety when giving large doses of epidural ropivacaine, for example, during surgical epidural anesthesia. However, our study did not address the pharmacodynamic effects of adding epinephrine to epidural ropivacaine, such as prolongation of sensory block.
In contrast to our findings, it has not been recommended that epinephrine be added to ropivacaine because ropivacaine has previously been demonstrated to possess intrinsic vasoconstrictor properties. For example, when ropivacaine was intradermally injected, Akerman et al. (1) and Cederholm et al. (13) observed blanching of the skin, and Kopacz et al. (2) and Cederholm et al. (6) reported decreased skin blood flow as measured using laser Doppler flowmetry. Furthermore, Dahl et al. (7) found that an epidural injection of ropivacaine reduced epidural blood flow, measured using a 133xenon clearance technique. In a clinical study, Hickey et al. (14) showed that the addition of epinephrine to ropivacaine used for brachial plexus block did not decrease plasma concentrations of ropivacaine. From these studies, it has been extrapolated that the addition of epinephrine to epidural ropivacaine would have minimal pharmacokinetic advantage.
Why are the results of our study at variance from these previous reports? The vasoconstrictive property of ropivacaine is dependent on the concentration of ropivacaine studied (5,6) and has not been confirmed across all concentrations. There is evidence for a dose-related or biphasic vasoactive effect, with small concentrations invoking vasoconstriction and large concentrations causing no effect or vasodilation (1,5,6). Data on the effect of added epinephrine are also conflicting, with reports varying from vasoconstriction (1–3,13) to no effect (14–17) or attenuation (6) of the vasoconstrictive effect of epinephrine. In patients undergoing reduction mammoplasty, administration of a bupivacaine-epinephrine mixture for preincision infiltration was associated with markedly less intraoperative blood loss compared with ropivacaine alone, suggesting there may be an advantage to adding epinephrine to ropivacaine for vasoconstrictive effect (18). Furthermore, vascular beds vary in their response to drugs and stimuli; hence, as also commented by Kopacz et al. (2), it may not be valid to extrapolate data obtained from examining skin blood flow to effects in the epidural space, which has greater vascularity.
Few previous studies have directly investigated the effect of epinephrine on the absorption of ropivacaine from the epidural or caudal epidural space. Arthur et al. (15) investigated the effect of the addition of epinephrine 5 μg/mL on the systemic absorption of ropivacaine in concentrations of 0.25%, 0.5%, 0.75%, and 1% after epidural injection in dogs. They found that epinephrine-containing solutions resulted in smaller mean peak arterial concentrations of ropivacaine compared with plain solutions, although the difference only reached statistical significance for the 0.5% concentration. The authors concluded that epinephrine did not consistently decrease the Cmax of ropivacaine, but the study may have lacked the power to detect a difference between groups due to the small sample size of only six animals. The same group reported in a subsequent study (16) that the addition of epinephrine 5 μg/mL to ropivacaine 1% injected epidurally in dogs reduced mean whole blood concentrations of ropivacaine compared with plain ropivacaine, although the difference was again not statistically significant. In a more recent study in humans, Leonard et al. (19) investigated the effect of the addition of epinephrine 5 μg/mL on continuous epidural infusion of ropivacaine 10 mg/h during labor and found that maternal plasma concentrations of ropivacaine were smaller in the epinephrine group at one hour but not at delivery. Roelants et al. (20) studied the pharmacokinetics of caudal epidural ropivacaine in children and found, similar to our results, that the addition of epinephrine 5 μg/mL to ropivacaine 2 mg/kg reduced mean Cmax by 34%, but unlike our results increased mean Tmax by 264%. Together with our study, these data suggest that epinephrine has a significant vasoconstrictor effect that is greater than, or additive to, any intrinsic vasoconstrictor activity of ropivacaine in the epidural space.
We chose to administer a weight-proportional dose of 1.5 mg/kg of ropivacaine, given in a fixed volume of 15 mL. In our patients, this resulted in a mean concentration of ropivacaine of 0.53%. It would be of interest to investigate whether the effect of epinephrine is similar with other concentrations of ropivacaine.
Most pharmacokinetic studies of epidural local anesthetic have been performed in either awake volunteers or patients undergoing regional anesthesia alone, whereas our patients underwent combined general and regional anesthesia. Sandler et al. (21) reported that general anesthesia had no confounding effect on the pharmacokinetics of epidural ropivacaine assessed by venous samples. Whether general anesthesia might modify the effect of epinephrine or the arterio-venous concentration difference is not known.
Previous pharmacokinetic studies have shown larger drug concentrations in arterial blood compared with venous blood. This has important implications for the site of blood sampling for determination of plasma drug concentrations in clinical practice (22–24). In our study, we found that the arterio-venous difference in plasma concentration of ropivacaine decreased over the first 60 minutes after the epidural injection, and this was unaffected by the addition of epinephrine. Arterial concentrations of ropivacaine were 30% larger than venous concentrations early in the sampling period, with the difference decreasing exponentially with time. This difference decreased to 10% after approximately 20 minutes. Emanuelsson et al. (8) reported significant arterio-venous concentration differences (up to 50%) for up to 60 minutes after epidural injection of ropivacaine in adult volunteers. This difference was larger and persisted longer than that observed by other investigators after the epidural injection of bupivacaine where arterio-venous plasma concentration differences (10%–30%) were significant in the first 30 minutes in adults (25) and 15 minutes in children (26). This suggests that the peripheral venous plasma concentration of ropivacaine in the first hour after injection may be an unreliable measure of vital organ tissue concentration in the assessment of toxicity secondary to systemic absorption and that arterial sampling is preferable during this time.
In conclusion, we have found that the addition of epinephrine significantly decreased the Cmax and the systemic plasma concentrations of ropivacaine in the first hour after epidural injection in anesthetized women. Therefore, the addition of epinephrine to ropivacaine may be a useful strategy to reduce the risk of systemic toxicity when it is injected epidurally.
The authors wish to thank their research nurses Floria Ng, Mabel Wong, and Charlotte Lam for assistance with data collection.
1. Akerman B, Hellberg IB, Trossvik C. Primary evaluation of the local anaesthetic properties of the amino amide agent ropivacaine (LEA 103). Acta Anaesthesiol Scand 1988; 32: 571–8.
2. Kopacz DJ, Carpenter RL, Mackey DC. Effect of ropivacaine on cutaneous capillary blood flow in pigs. Anesthesiology 1989; 71: 69–74.
3. Guinard JP, Carpenter RL, Owens BD, Nadir B. Comparison between ropivacaine and bupivacaine after subcutaneous injection in pigs: cutaneous blood flow and surgical bleeding. Reg Anesth 1991; 16: 268–71.
4. Nakamura K, Toda H, Kakuyama M, et al. Direct vascular effect of ropivacaine in femoral artery and vein of the dog. Acta Anaesthesiol Scand 1993; 37: 269–73.
5. Gherardini G, Samuelson U, Jernbeck J, et al. Comparison of vascular effects of ropivacaine and lidocaine on isolated rings of human arteries. Acta Anaesthesiol Scand 1995; 39: 765–8.
6. Cederholm I, Evers H, Lofstrom JB. Skin blood flow after intradermal injection of ropivacaine in various concentrations with and without epinephrine evaluated by laser Doppler flowmetry. Reg Anesth 1992; 17: 322–8.
7. Dahl JB, Simonsen L, Mogensen T, et al. The effect of 0.5% ropivacaine on epidural blood flow. Acta Anaesthesiol Scand 1990; 34: 308–10.
8. Emanuelsson BM, Persson J, Alm C, et al. Systemic absorption and block after epidural injection of ropivacaine in healthy volunteers. Anesthesiology 1997; 87: 1309–17.
9. Winer BJ. Statistical principles in experimental design. 2nd ed. New York: McGraw-Hill, 1971.
10. Diggle PJ, Liang KY, Zeger SL. Analysis of longitudinal data. Oxford: Oxford University Press, 1994.
11. Tucker GT, Mather LE. Clinical pharmacokinetics of local anaesthetics. Clin Pharmacokinet 1979; 4: 241–78.
12. Burm AG, van Kleef JW, Gladines MP, et al. Epidural anesthesia with lidocaine and bupivacaine: effects of epinephrine on the plasma concentration profiles. Anesth Analg 1986; 65: 1281–4.
13. Cederholm I, Akerman B, Evers H. Local analgesic and vascular effects of intradermal ropivacaine and bupivacaine in various concentrations with and without addition of adrenaline in man. Acta Anaesthesiol Scand 1994; 38: 322–7.
14. Hickey R, Blanchard J, Hoffman J, et al. Plasma concentrations of ropivacaine given with or without epinephrine for brachial plexus block. Can J Anaesth 1990; 37: 878–82.
15. Arthur GR, Feldman HS, Covino BG. Comparative pharmacokinetics of bupivacaine and ropivacaine, a new amide local anesthetic. Anesth Analg 1988; 67: 1053–8.
16. Hurley RJ, Feldman HS, Latka C, et al. The effects of epinephrine on the anesthetic and hemodynamic properties of ropivacaine and bupivacaine after epidural administration in the dog. Reg Anesth 1991; 16: 303–8.
17. Cederholm I, Anskar S, Bengtsson M. Sensory, motor, and sympathetic block during epidural analgesia with 0.5% and 0.75% ropivacaine with and without epinephrine. Reg Anesth 1994; 19: 18–33.
18. Liddle AM, Hall AP, Arrowsmith J, Smith G. Effect of infiltration with ropivacaine on blood loss during reduction mammoplasty. Br J Anaesth 1998; 81: 974–5.
19. Leonard SA, Flynn R, Kelleher N, Shorten GD. The effect of the addition of epinephrine to ropivacaine for labour epidural analgesia on efficacy, systemic absorption and neonatal outcome. Eur J Anaesthesiol 2000; 17 (suppl 19): A519.
20. Roelants FA, Veyckemans F, Verbeeck RK, Van Obbergh LJ. Pharmacokinetics of caudal ropivacaine with or without epinephrine in children. Anesthesiology 2000; 3A: A1269.
21. Sandler AN, Arlander E, Finucane BT, et al. Pharmacokinetics of three doses of epidural ropivacaine during hysterectomy and comparison with bupivacaine. Can J Anaesth 1998; 45: 843–9.
22. Chiou WL. The phenomenon and rationale of marked dependence of drug concentration on blood sampling site: implications in pharmacokinetics, pharmacodynamics, toxicology and therapeutics (Part I). Clin Pharmacokinet 1989; 17: 175–99.
23. Chiou WL. The phenomenon and rationale of marked dependence of drug concentration on blood sampling site: implications in pharmacokinetics, pharmacodynamics, toxicology and therapeutics (Part II). Clin Pharmacokinet 1989; 17: 275–90.
24. Moore DC, Mather LE, Bridenbaugh PO, et al. Arterial and venous plasma levels of bupivacaine following epidural and intercostal nerve blocks. Anesthesiology 1976; 45: 39–45.
25. Bachmann MB, Biscoping J, Adams HA, et al. The significance of the sampling site in the determination of plasma levels of local anesthetics using 0.75% bupivacaine as an example. Reg Anesth 1990; 13: 16–20.
26. Eyres RL, Bishop W, Oppenheim RC, Brown TC. Plasma bupivacaine concentrations in children during caudal epidural analgesia. Anaesth Intensive Care 1983; 11: 20–2.