Local Anesthetics Impair Human Granulocyte Phagocytosis Activity, Oxidative Burst, and CD11b Expression in Response to Staphylococcus aureus 

Three different concentrations of lidocaine (Xylocaine®; Astra, Wedel, Germany), bupivacaine (Carbostesin®, Astra, Wedel, Germany), and ropivacaine (Naropin®; Astra, Wedel, Germany) were chosen for investigation of their effect on granulocyte phagocytosis activity, oxidative burst, and CD11b expression. Preservative-free preparations of the local anesthetics were diluted in isotonic saline to a total volume of 100 μl each. Final working concentrations of local anesthetic in whole blood were as follows: lidocaine, 9.2, 92.2 and 1,846 μm; bupivacaine, 6.1, 61.6, and 770 μm; and ropivacaine, 6.4, 64, and 801 μm. The pH was 7.4 in all experiments. Eleven experiments were conducted for each local anesthetic in each concentration.

After obtaining institutional and ethical committee approval (Medical Faculty, Eberhard-Karls University, Tuebingen, Germany) and informed consent, blood samples (10 ml) were acquired from 14 healthy volunteers (aged 32.5 ± 7 yr) via  an 18-gauge needle in a large antecubital vein. Blood was heparinized with 5 U/ml heparin sodium (Vetren 200®; Byk Gulden, Konstanz, Germany). One hundred microliters of the samples was used for leukocyte count (using trypan blue solution with 3% acetic acid for lysis of erythrocytes). The remaining blood was divided into volumes of 2 ml and mixed with different concentrations of local anesthetic or with isotonic saline (not bacteriostatic) as control. Blood samples were incubated with the different local anesthetic concentrations for 30 min at 37°C.

S. aureus  (ATCC 25923; American Type Culture Collection, Rockville, MD) was cultivated in Tryptone Soya Broth (Oxoid, Wesel, Germany) for 3 h and centrifuged (375 g ) for 5 min. The pellet was resuspended in phosphate-buffered saline without calcium or magnesium (Sigma, Deisenhofen, Germany) and then washed twice to remove the growth medium. For the detection of intracellular bacteria by flow cytometry, the pellet was incubated with Calcein-AM (Calcein-AM special packing; Molecular Probes, Eugene, OR) for 50 min at 37°C. Calcein-AM, an acetoxymethyl ester with a high and stable green fluorescence signal, penetrates the cell wall, is cleaved by cytoplasmic esterases to a hydrophilic product, and is thereby trapped in the cell. Stained bacteria were washed twice in phosphate-buffered saline, adjusted to a final concentration of 2 × 10⁹colony-forming units (CFU) per milliliter, and stored in aliquots at −20°C until use. Bacterial count was performed by plating out appropriate dilutions of bacterial preparations on mannitol salt agar (Oxoid, Wesel, Germany) and counting CFU after 24 h of incubation at 37°C. Repeated assays were performed to ascertain stable fluorescence intensity and bacterial viability during all experiments. For the oxidative burst and CD11b assay, unstained bacteria were diluted to 2 × 10⁹CFU/ml and stored at −20°C.

To provide a constant 5 to 1 ratio of bacteria to leukocytes for each assay, bacteria were diluted based on whole blood leukocyte count. The culture medium was RPMI (RPMI 1640; Sigma, Deisenhofen, Germany) containing 10% autologous serum (to preopsonize bacteria and to provide for faster phagocytosis). Diluted bacteria and local anesthetic–treated whole blood samples were separately incubated for 30 min at 37°C. Bacteria were then added to the blood samples and incubated at 37°C for 10, 30, and 60 min before phagocytosis was stopped by adding N -ethylmaleimide (10 mm; Sigma, Deisenhofen, Germany). Between steps, the samples were stored on ice.

For flow cytometric detection, the granulocytes were stained with the monoclonal antibody against CD13 (Myeloid Cell, CD13 RPE; DAKO, Glostrup, Denmark), and the erythrocytes were lysed (Lysing Solution; Becton Dickinson, Heidelberg, Germany). The samples then were centrifuged (375 g  for 5 min), resuspended in phosphate-buffered saline, and kept on ice until flow cytometric analysis.

For determination of oxidative burst activity and CD11b expression of leukocytes, unstained bacteria were mixed with leukocytes to a final ratio of 5 to 1 and preopsonized, as in the phagocytosis assay. Blood samples were mixed with local anesthetic and treated as described previously. To determine the generation of reactive oxygen species, dihydroethidium (2.5 μg/ml; Sigma, Deisenhofen, Germany) was added. Dihydroethidium is reduced to hydroethidium by reactive oxygen metabolites, emitting a stable fluorescence signal of 550–720 nm wavelength. Because of the broad emission spectrum of dihydroethidium during flow cytometric analysis (detection in channels 2 and 3), a staining with three fluorescence dyes is not feasible. As the uptake of bacteria was already quantified in the phagocytosis assay, unstained bacteria were used (in the same ratio as in the phagocytosis assay) to determine CD11b expression during phagocytosis.

After phagocytosis was stopped (10 and 30 min), as described previously, the samples were stained with a monoclonal antibody against CD11b (Mouse anti-Human CD11b; Caltag Laboratories, San Francisco, CA) for 20 min at room temperature, the erythrocytes were lysed (as in the phagocytosis assay), and the samples were prepared for flow cytometric analysis.

Flow cytometric analysis was performed using FACS SORT (Cell Quest software package, version 3.1f; Becton Dickinson, Heidelberg, Germany). Unstained samples served as a control for the background fluorescence of dihydroethidium, collected through the 620-nm band pass filter. Isotype controls for each monoclonal antibody were used to define the cutoff for positive fluorescence.

Phagocytosis activity was determined in two different ways: as the ratio between the number of neutrophils ingesting bacteria and the total amount of neutrophils and as the fluorescence intensity of intracellular bacteria; the latter served as a semiquantitative measurement of bacterial uptake. Ten thousand granulocytes were analyzed in each sample.

Statistical analysis was performed using the JMP® statistical software package (SAS Institute Inc., Cary, NC). Analysis of variance (ANOVA) was performed, including blood donors as random effect and time, local anesthetic concentration, and their respective interaction as fixed factors. For each time point, differences between the local anesthetic concentration and the respective control were detected by Student t  test with Bonferroni correction for multiple comparisons. Alpha was set at 0.05.

Comparisons among the local anesthetics were performed by ANOVA as well, including blood donors as random effect and local anesthetic concentrations and the different local anesthetics as fixed facts. To avoid multiple comparison problems, time effects were not evaluated. Normal distribution of the ANOVA residuals was assured (Shapiro–Wilk W test).

Ropivacaine, at all concentrations and at all time points investigated, had no effect on the percentage of granulocytes ingesting bacteria. Lidocaine and bupivacaine inhibited phagocytosis but with different time courses. After 10 min of incubation, the percentage of granulocytes ingesting bacteria was decreased to 80% by bupivacaine (770 μm;P < 0.002 compared with control) but not by lidocaine. After 30 min of incubation, the percentage of granulocytes that ingested S. aureus  was decreased to 71% by lidocaine (1,846 μm;P < 0.0002) and to 85% by bupivacaine (770 μm;P < 0.001;fig. 1). After 60 min of incubation, only lidocaine (1,846 μm) significantly reduced the number of granulocytes that ingested bacteria (to 71%;P < 0.0001). Hence, the effects of the three local anesthetics at these concentrations were significantly different from one another (P < 0.02). At lower concentrations, none of the local anesthetics affected the number of granulocytes that ingested bacteria.

In addition, the uptake of labeled S. aureus  was semiquantitatively determined by measuring the intracellular fluorescence signal. The fluorescence intensity of S. aureus  was significantly decreased to 68% by lidocaine at 60 min (1,846 μm;P < 0.003). For bupivacaine (770 μm), a trend toward fewer intracellular bacteria could be observed but did not reach statistical significance (fig. 2). As a result, the effects of lidocaine and bupivacaine at these concentrations were significantly different from those of ropivacaine (P < 0.0001).

Molecular Probes, Eugene, OR). Data after 10, 30, and 60 min are presented as mean and SD. Significant inhibition (*P < 0.003) of bacterial uptake occurred at 60 min of lidocaine exposure.

A significant reduction in oxidative burst—to 61 and to 69%—was observed at 30 min for lidocaine (1,846 μm;P < 0.0001) and bupivacaine (770 μm;P < 0.0001), respectively. Ropivacaine did not have any effect on oxidative burst (fig. 3). However, due to substantial variability, no significant differences among the local anesthetics were observed.

CD11b expression was reduced at 10 and 30 min by lidocaine (1,846 μm; to 50 and 52%) and by bupivacaine (770 μm; to 65 and 70%), respectively (P < 0.006). For ropivacaine (801 μm), a trend toward reduced expression of CD11b to 79% (10 min) and to 78% (30 min) was observed but did not reach statistical significance (fig. 4). The effect of lidocaine was significantly different from that of the other local anesthetics (P < 0.0001), whereas no differences in effect were observed between bupivacaine and ropivacaine.

The present study investigated the effects of lidocaine, bupivacaine, and ropivacaine on granulocyte phagocytosis activity, oxidative burst, and CD11b expression. All three parameters were significantly attenuated by lidocaine or bupivacaine in their highest concentrations (1,846 and 770 μm, respectively). The uptake of intracellular bacteria was significantly affected following lidocaine exposure only. Treatment with ropivacaine showed only a nonsignificant trend toward reduced CD11b expression in the highest concentration (801 μm).

Lidocaine previously has been shown to inhibit phagocytosis of radiolabeled S. aureus , 14heat-killed Escherichia coli , 16or latex particles 15in isolated leukocytes. This study confirms these findings for a whole blood assay and adds new evidence suggesting comparable inhibitory effects of bupivacaine but lack of ropivacaine interference with bacterial ingestion. In addition, we observed a significant decrease in the number of leukocytes actively participating in phagocytosis after lidocaine exposure. Ropivacaine, in contrast, affected neither bacterial ingestion nor the overall number of active phagocytes.

The effects of lidocaine on oxidative burst stimulated by bacterial surface antigens, 16yeast particles, 13,18or intracellular stimuli, like phorbol myristate acetate, have been well documented. 13,15,17,22Our data show that lidocaine similarly affects respiratory burst activated by S. aureus  phagocytosis. In contrast, the effects of bupivacaine and ropivacaine on oxidative burst are variably reported. Mikawa et al.  19did not observe any effect of bupivacaine on zymosan-stimulated oxidative burst, whereas Sinclair et al.  18reported suppression by bupivacaine in concentrations comparable to those used in this study. Cederholm et al.  23could not show significant effects of bupivacaine or ropivacaine on respiratory burst induced by phorbol myristate acetate, whereas Hollmann et al.  22and Fischer et al.  24demonstrated inhibitory activity of both local anesthetics on chemotaxis and oxygen production in granulocytes primed with platelet-activating factor and activated with N -formyl-methionine-leucine-phenylalanine. The differences between these studies may be a result of the various stimuli and incubation times used. Hollmann et al.  25,26recently demonstrated that local anesthetics influence intracellular signaling by interfering with G^qproteins, resulting in the blockage of neutrophil priming. In our study, differentiation between priming and activation effects was not possible because viable S. aureus  was applied as a single stimulus.

CD11b, an adhesion molecule in the β²-integrin family, is rapidly upregulated after neutrophil activation and plays a crucial role by allowing firm adhesion of rolling leukocytes to the endothelium, 27,28thereby initiating the first step in extravasation of leukocytes to sites of inflammation. Moreover, CD11b is the predominant leukocyte surface-binding site for opsonized S. aureus  and is responsible for both recognition and binding. Bound S. aureus  is rapidly internalized into phagocytes (via  phagosomes) and degraded (via  oxidative burst and enzymatically) after interaction with the CD11b receptor. The most pronounced local anesthetic effect observed in our study was on CD11b expression. Inhibition of CD11b expression by lidocaine and ropivacaine has already been reported for N -formyl-methionine-leucine-phenylalanine–stimulated or tumor necrosis factor α–stimulated leukocytes. 20,29In our study, CD11b expression was decreased by lidocaine and bupivacaine during phagocytosis of viable S. aureus . This would be expected to result in compromised ingestion of viable S. aureus  by leukocytes.

We observed significant differences in effect among the various anesthetics. This could imply that either nonequipotent concentrations were used or that various local anesthetics have different pharmacologic effects on leukocyte function. Statements regarding equipotency are difficult to make since we investigated a novel phenomenon for which anesthetic sensitivity has not been described. Clearly, “equipotent” concentrations, as described for analgesia, are irrelevant in the present setting since local anesthesia is largely mediated by the blockage of sodium channels, which are not expressed in leukocytes. 30Therefore, we decided to study a wide range of concentrations, anchored in clinical relevance. For each of the local anesthetics, we chose the lowest concentration to correspond to plasma concentrations observed following regional anesthetic procedures. 31–35In addition, we studied a concentration 10-fold greater than and a concentration corresponding to 100 times the peak plasma levels observed after regional anesthesia. As a result, we studied a greater than 100-fold range of concentrations. Unfortunately, almost no data exist on a more useful measure in the current context: tissue concentrations of local anesthetic at the site of injection (e.g. , after regional blocks, which are performed with high volumes and concentrated local anesthetics). After subarachnoid block, millimolar concentrations are obtained near the spinal cord. 36After brachial plexus block, tissue ropivacaine concentrations were approximately 50 times greater than those subsequently observed in plasma. 34,35By extrapolating, it can be assumed that tissue concentrations are at least 10-fold greater than systemic concentrations. In addition, local anesthetics are frequently applied over the period of days or weeks. In vitro  studies describe an increasing susceptibility of leukocytes to local anesthetics with increasing exposure time. 37Thus, even lower local anesthetic concentrations may result in compromised leukocyte function in cases involving prolonged exposure times. Together, these data indicate that the concentrations used in this study may, indeed, be relevant in the clinical setting.

Although the results of this in vitro  study cannot be extrapolated to local anesthetic actions in vivo  and in clinical practice, our data at least suggest the possibility of clinically relevant effects on leukocyte functions at sites with high local anesthetic concentrations. However, the limitations of the in vitro  model must be emphasized since under experimental conditions, only effects on an isolated system are observed, whereas in vivo , leukocytes and S. aureus  interact with a variety of cellular and noncellular targets, including the tissue immune system and circulating blood cells other than neutrophils. Nonetheless, our findings may be of clinical relevance in subpopulations of patients with preexisting compromised immune status, diabetes mellitus, malignancies, use of antiinflammatory medication, or long-term analgesia for chronic pain syndromes, in whom a 10-fold increase in the incidence of epidural abscess, as compared with healthy subjects, has been reported. 38Our data suggest that ropivacaine, which showed limited effects on leukocyte function, may be the preferred local anesthetic for prolonged regional anesthesia in such patients.

The authors are deeply indebted to Marcel E. Durieux, M.D., Ph.D. (Professor and Chairman, Department of Anesthesiology, University Hospital, Maastricht, The Netherlands), for his expertise in discussing and interpreting the data and his dedicated support in writing and revising the manuscript. The authors also thank Joan Robertson-Hoehne, M.D. (Family Practitioner, Stuttgart, Germany), for copyediting the manuscript.