Erythropoietin Protects against Local Anesthetic Myotoxicity during Continuous Regional Analgesia

This study, including care of the animals involved, was conducted according to the official edict presented by the French Ministry of Agriculture (Paris, France) and the recommendations of the Declaration of Helsinki. Therefore, these experiments were conducted in an authorized laboratory and under the supervision of an authorized researcher (K.N.-G.).

Bupivacaine hydrochloride, 0.5% (15 mm), was purchased from AstraZeneca (Rueil-Malmaison, France) for rat administration and from Sigma-Aldrich (St. Louis, MO; B5274) for cell experiments. rhEPO was obtained from Janssen-Cilag (Issy-Les-Moulineaux, France). All other reagents were purchased from Sigma, with the exception of the ATP monitoring kit (ATP Bioluminescence Assay Kit HS II; Roche Diagnostics GmbH, Mannheim, Germany) and the primary antibodies (OXPHOS Kit, MS601 MitoProfile®; Mitosciences, Eugene, OR).

Experiments were conducted on adult male Wistar rats, 10–12 weeks old, weighing 200–240 g. Rats were housed in a regulated facility with a 12-h light/12-h dark cycle, fed with chow, and allowed free access to tap water. After anesthesia with intraperitoneal pentobarbital sodium (40 mg/kg) and subcutaneous injection of lidocaine (10 mg), a plexus catheter (20 gauge, 0.9 mm OD; Pajunk, Geisingen, Germany) was inserted under the inguinal ligament near the left femoral nerve sheath (fig. 1). It was fixed with stitches on the quadriceps muscle, passed under the skin, and exited at the neck. Incisions were subsequently closed by suturing.

Animals were randomly assigned to four different groups according to the type of combined injections: daily intraperitoneal injection with 5,000 U/kg rhEPO (E) or saline (S) coupled to a perineural catheter injection of 1 ml/kg bupivacaine, 0.25% (EB and SB groups), or saline (ES and SS groups). Rats received one daily intraperitoneal injection for 3 days and seven perineural injections 8 h apart. This induced a decrease in pinprick sensation in the cutaneous distribution of the femoral nerve but not complete motor blockade in the first hour after each bupivacaine injection. Rats were killed by cervical dislocation 8 h after the last perineural injection, when the bupivacaine concentration in muscle was below the threshold of detection (< 0.3 μg/g tissue).10Psoas muscle was quickly dissected adjacent to the femoral nerve, with the former tip region of the catheter located in the middle of the tissue block.

Human skeletal muscle myoblasts (HSMMs) were obtained from Cambrex Bio Science Walkersville, Inc. (Walkersville, MD) and grown according to the manufacturer’s recommendations in skeletal muscle basal medium 2 supplemented with skeletal muscle growth medium 2 Single-Quots (human epidermal growth factor, dexamethasone, l-glutamine, stain buffer, and gentamicin–amphotericin B). For bioenergetic and microscopic analysis, cells were analyzed during the exponential growth phase at 70% confluence. Bupivacaine treatment was performed by incubating cells for 1, 8, 24, or 48 h with 13 different concentrations of bupivacaine ranging from 0.1 to 5,000 μm. When indicated, pretreatment with 1 U/ml rhEPO was performed 8 or 24 h before bupivacaine treatment.

Cytotoxicity of bupivacaine was analyzed using both a dimethyl thiazol diphenyl tetrazolium bromide (MTT) and neutral red assay. HSMM cells were seeded at a density of 2.10⁴/ml in 96-well plates and cultured until subconfluence in medium with or without bupivacaine and/or rhEPO. Staining with MTT or neutral red was performed as described by Mosmann12and Borenfreund and Puerner,13respectively. The MTT assay measures the reduction of the tetrazolium compound, primarily by respiratory chain activity of mitochondria using tetrazolium as an artificial electron acceptor. Absorbance was measured in a multiwell scanning spectrophotometer (Dynex MRXII; Chantilly, VA) at a wavelength of 570 nm for MTT and 540 nm for neutral red, with a reference set at 630 nm. Tests were performed in quadruplicate and repeated six times. The results were expressed as a percentage of control absorbance.

To assess mitochondrial respiration in rats, we used a permeabilized muscle fiber technique14with the respiratory substrates 10 mm malate plus 10 mm glutamate or 25 mm succinate plus rotenone (1 mg/ml dimethyl sulfoxide and ethanol 1:1) as indicated. Respiration was expressed as ng atom O/min/mg wet weight of the muscle fiber. Concomitant ATP synthesis measurements were performed by luminometry and expressed as nmol ATP/min/mg wet weight of the muscle fiber.15This allowed for calculation of the efficiency of ATP production (ATP/O ratio).

For the HSMMs, endogenous respiratory rate was assayed in intact cells using a high-precision polarographic technique. Respiration was measured at 37°C on an Oroboros® oxygraph (Oroboros Instruments, Innsbruck, Austria) with 1 × 10⁶cells/ml in skeletal muscle growth medium 2.16The experiment began with routine respiration, without additional substrates or effectors. After observing steady state respiratory flux, bupivacaine (0.1 μm–5 mm) was added. In addition, some HSMMs were pretreated with erythropoietin (1 U/ml) before conducting the experiment. Respiratory rates were expressed as ng atom O/min/1 × 10⁶cells.

Mitochondrial membrane potential (ΔΨ) was measured using the lipophilic cation JC-1 according to the manufacturer’s protocol in skeletal muscle growth medium 2.17ΔΨ was expressed as the ratio of red (ex 490 nm/em 590 nm) to green (ex 490 nm/em 527 nm) fluorescence, measured on a Xenius spectrofluorometer (SAFAS, Monaco, France). The mitochondrial uncoupler carbonyl cyanide-p-trifluoromethoxyphenyl-hydrazone (15 mm) was used as positive control for mitochondrial depolarization. HSMMs were grown with or without bupivacaine (1 μm, 1 mm, and 5 mm) for 24 h and/or rhEPO (1 U/ml, 8 h before bupivacaine treatment).

To determine the effects of bupivacaine and rhEPO treatment on the mitochondrial respiratory chain, we measured the individual enzymatic activity of complexes I–IV and citrate synthase in tissue homogenates prepared from rat psoas muscle using spectrophotometric procedures as previously described.10Briefly, approximately 100 mg psoas muscle was minced and homogenized with a glass potter homogenizer (Kimble/Kontes, Vineland, NJ) in ice-cold medium (10% wt/vol) containing 225 mm mannitol, 75 mm sucrose, 10 mm Tris-HCl, and 0.10 mm EDTA at pH 7.2. The homogenate was then centrifuged for 20 min at 650g . The supernatant was collected, and the protein concentration was determined.18Citrate synthase activity was measured, as described by Srere,19in the presence of 4% Triton (vol/vol) by monitoring the formation of thionitrobenzoate dianion from the reaction of coenzyme A and 5,5′-dithiobis (2-nitrobenzoic acid) at 412 nm and 30°C. Complex I activity, reduced nicotinamide adenine dinucleotide ubiquinone reductase, was measured as described by Birch-Machin et al.  20The oxidation of reduced nicotinamide adenine dinucleotide by complex I was recorded using the ubiquinone analog decylubiquinone as an electron acceptor. The decrease in absorption resulting from reduced nicotinamide adenine dinucleotide oxidation was measured at 340 nm at 30°C. Complex I activity was calculated from the difference in the rate before and after the addition of rotenone (2 μm), a specific inhibitor of complex I. Complex II (succinate dehydrogenase)–specific activity was measured by monitoring the reduction of 2,6-dichlorophenol indophenol at 600 nm at 30°C in the presence of phenazine methosulphate.21The oxidation of ubiquinol (UQ¹H²) by complex III (decylubiquinol cytochrome c  reductase) was determined using cytochrome c  (III) as an electron acceptor.20The reduction of cytochrome c  (III) was recorded at 550 nm at 30°C. Complex IV (cytochrome c  oxidase) activity was measured as described by Wharton and Tzagoloff22using cytochrome c  (II) as a substrate. The oxidation of cytochrome c  was monitored at 550 nm at 30°C. Enzyme activities were expressed as nmol substrate/min/mg protein.

The content of respiratory chain complexes was measured by Western blot as described previously.17Tissue samples from psoas muscle (SS and ES groups) were diluted in sodium dodecyl sulfate polyacrylamide gel tricine sample buffer (Bio-Rad Laboratories, Hercules, CA) containing 2% β-mercaptoethanol by incubation for 30 min at 37°C and separated on a 10–22% sodium dodecyl sulfate polyacrylamide gradient mini-gel (Bio-Rad Laboratories) at 150 V. Proteins were transferred electrophoretically to a 0.45-μm polyvinylidine difluoride membrane for 2 h at 100 mA in N -cyclohexyl-3-aminopropanesulfonic acid buffer (3.3 g N -cyclohexyl-3-aminopropanesulfonic acid; 1.5 l methanol, 10%; pH 11) on ice. Membranes were blocked overnight in 5% milk–phosphate-buffered saline +0.02% azide, and incubated for 3 h with the primary antibodies. After three washes with phosphate-buffered saline −0.05% Tween 20, membranes were incubated for 2 h with horseradish peroxidase–conjugated goat anti-mouse secondary antibody (Bio-Rad Laboratories) diluted in 5% milk– phosphate-buffered saline. The secondary antibody was detected using chemiluminescent ECL Plus reagent (Amersham Biosciences, GE Healthcare, Uppsala, Sweden). The signal was quantified by densitometric analysis using Image J software (National Institutes of Health, Bethesda, MD).

Changes in cytosolic levels of reactive oxygen species were monitored using the CM-H²-DCFDA fluorescent probe (Invitrogen) in HSMMs grown with or without bupivacaine (1 μm, 1 mm, and 5 mm for 24 h) and/or rhEPO (1 U/ml, 8 h before bupivacaine treatment). Fluorescence was measured at steady state in skeletal muscle basal medium 2 with a spectrofluorometer (SAFAS, Monaco, France) using excitation and emission wavelengths set at 495 and 520 nm, respectively. Adding H²O²(100 μm) to the cuvette was used as a positive control. Results are expressed as a percentage of the control fluorescence.

Profiles of mitochondrial sections were analyzed by transmission electron microscopy on a Hitachi H-7650 microscope (Tokyo, Japan) as described previously,23,24and catheter insertion sites in rat psoas muscle were analyzed by scanning electron microscopy. A series of at least five different samples was taken from four different muscles from each group of rats (SS, ES, SB, EB).

In HSMMs, the morphology of the mitochondrial network was studied by fluorescence microscopy using mitotracker green (150 nm, 20 min at 37°C) on a Nikon E 200 microscope (Kawasaki, Japan) with a 60×, 1.4 numerical aperture water immersion objective. A series of images was acquired using a Q-Imaging Retiga Exi fast 1394 digital camera driven by Fluoup (Explora Nova, La Rochelle, France). Image analysis was performed with Morpho pro version 2.8 (Explora Nova). HSMMs were grown with or without bupivacaine (1 μm, 1 mm, and 5 mm for 24 h) and/or rhEPO (1 U/ml, 8 h before bupivacaine treatment). Fifteen images were taken from three different cell culture dishes. We defined three main morphologic configurations: filamentous, outgrowth, and fragmented. One hundred twenty cells taken from three independent experiments were analyzed randomly for each condition. The results are expressed as a percentage of the counted cells for each configuration.

Magnetic resonance imaging experiments were conducted on a 4.7-T Biospec horizontal system (Bruker, Ettlingen, Germany), equipped with a 12-cm gradient system, capable of 200 mT/m maximum strength and 180 μs rise time. Measurements were performed with a rat-dedicated probe birdcage resonator (80 mm in diameter and 120 mm long) tuned to 200.3 MHz. Wistar rats were anesthetized with isoflurane (1.5% in air). A solution of 5 mm of magnetic resonance imaging contrast agent (Gd-DOTA, 5 mm, DOTAREM; Guerbet, Aulnay-sous-bois, France) was then injected by the peripheral nerve catheter, and diffusion was observed in the psoas muscle using a dedicated fast low angle shot sequence.

For mitochondrial respiration and enzyme activities in the rat muscle model, quantitative data are reported as a median [25th and 75th percentiles], because of a nonnormal distribution. Data from the groups (SS, SB, ES, EB) were then compared using a Kruskal–Wallis test; Mann–Whitney tests were performed by comparing the SS group with the other three groups (the P  value required for statistical significance was determined by dividing 0.05 by the number of comparisons; therefore, P = 0.05/3 = 0.0166).

For the cell culture model, data are expressed as mean ± SD, and one-way or two-way analysis of variance (α= 0.05) was performed as appropriate, in addition to a Student–Newman–Keuls post hoc  test (with P < 0.05 considered significant).

Tests were performed using SigmaStat 3.1 (Systat Software Inc., San Jose, CA).

Seventy rats were anesthetized according to the protocol summarized in figure 1A. No self-mutilation after catheter placement was observed. The localization of catheters and their tips inside living animals was verified by light microscopy (fig. 1B) and confirmed by magnetic resonance imaging analyses (fig. 1C), as well as by scanning electron microscopy observations (fig. 1D). The catheters were inserted into perimysium connective tissue and between muscle fibers without destruction to reach the vicinity of the femoral nerve, where bupivacaine was released. Five rats with catheter displacement were excluded from the analysis.

Measurements of coupled oxygen consumption rate and ATP synthesis were performed in permeabilized fibers using glutamate plus malate or succinate as substrates. Bupivacaine induced a significant decrease in adenosine diphosphate–stimulated oxygen consumption along with a significant inhibitory effect on ATP synthesis. The efficiency of oxidative phosphorylation (ATP/O ratio) was also reduced. Erythropoietin cotreatment (5,000 U/kg/24 h, EB group) prevented the inhibitory effect of bupivacaine on mitochondrial bioenergetics (table 1).

The activities of the respiratory chain complexes (I–IV) and citrate synthase were measured in the different groups (table 2). Results indicated a significant and global inhibition of respiratory chain activity by bupivacaine, with a prominent effect on complex I. This was associated with a reduction in mitochondrial content, as suggested by the decrease in citrate synthase activity. These toxic effects were fully blocked, however, by cotreatment with rhEPO (table 2). To explore the underlying mechanisms involved in the recovery of citrate synthase activity after rhEPO treatment, we looked at the possible activation of mitochondrial biogenesis by performing Western blot experiments (fig. 2). The results obtained from rat muscles treated with rhEPO showed that rhEPO induced no increase in the expression level of respiratory chain complexes (CI, CII, CIII, and F¹F⁰ATP synthase).

Electron micrographs of longitudinal psoas muscle tissue sections showed that mitochondrial morphology was also altered by bupivacaine (fig. 3). Subsarcolemmal aggregates of swollen mitochondria (fig. 3E) were observed with partial loss of interfibrillar mitochondria (fig. 3K). We also observed membranes thought to be autophagosomes containing either intact or degraded mitochondria (figs. 3K and J, respectively).25The internal organization of mitochondria was also severely affected, forming an onion-like structure (figs. 3G and H) that included rearrangements of matricial space and a loss of cristae. These unusual mitochondrial morphologies were observed in all samples from bupivacaine-treated muscle. Cotreatment with rhEPO prevented this mitochondrial structural disorganization (figs. 3M–P). This also suggests an inhibition of mitochondrial degradation processes by rhEPO.

Bupivacaine treatment (0.1 μm–5 mm) of intact HSMM cells resulted in dose-dependent alteration of endogenous respiration with uncoupling, followed by significant inhibition (fig. 4). These changes were not prevented by the direct addition of rhEPO to the cuvette of the oxygraph (data not shown).

The impact of this inhibition of respiratory rate on the MTT reduction assay was investigated (fig. 5A) and exhibited a threshold profile. Similar results were obtained with the neutral red assay (data not shown). Above 1 mm, bupivacaine exerted a dose- and time-dependent effect that reached 17 ± 5% and 18 ± 6% cell viability with 5 mm bupivacaine after 48 h, as measured by the MTT and neutral red assays, respectively. MTT reduction was decreased by bupivacaine and was partially recovered by rhEPO pretreatment and cotreatment (fig. 5), while treatment with 0.1–100 U/ml rhEPO alone for 24 and 48 h had no effect on MTT reduction (data not shown). No significant difference in this protective effect was observed for rhEPO concentrations ranging from 0.25 to 100 U/ml for 3 mm bupivacaine and 48-h incubation (fig. 5B), and we chose rhEPO at 1 U/ml for subsequent experiments. MTT reduction was significantly less diminished when rhEPO (1 U/ml) was added concomitantly with 5 mm bupivacaine for 24 h (57 ± 4 vs.  31 ± 2% with and without rhEPO, respectively; P < 0.05 by Student–Newman–Keuls post hoc  test) or 48 h (44 ± 10 vs.  17 ± 5% with and without rhEPO, respectively; P < 0.05 by Student–Newman–Keuls post hoc  test) as compared with 5 mm bupivacaine alone. MTT reduction was also not decreased by increasing bupivacaine concentration (fig. 5C). Interestingly, the duration of pretreatment with 1 U/ml rhEPO determined the extent of MTT reduction (fig. 5C).

Treatment of HSMMs for 24 h (fig. 6) with variable doses of bupivacaine (1 μm–5 mm) resulted in a progressive decrease in ΔΨ (figs. 6Ε–Η and 7) with complete depolarization using 1 mm bupivacaine. Interestingly, 1 U/ml rhEPO for an 8-h pretreatment and cotreatment protected against this phenomenon (figs. 6M–P). In contrast, we observed no change in cytosolic reactive oxygen species levels in cells treated with bupivacaine or cotreated with rhEPO (data not shown).

Bupivacaine treatment of HSMMs also led to progressive fragmentation of the mitochondrial network (figs. 6B–D and 8), which was initiated with 1 μm bupivacaine (24 h) and led to complete fragmentation at a dose of 5 mm (fig. 6D). Such changes in the mitochondrial network were delayed by pretreatment and cotreatment with 1 U/ml rhEPO for 8 h (figs. 6I–L, 8, and 9).

Local anesthetics (e.g. , bupivacaine) are widely used for postoperative regional analgesia, even though some studies have demonstrated a potential muscle toxicity risk.4,5,26Our study showed that this risk can be primarily explained by the impairment of mitochondrial structure and function6,7and can be partially blocked by treatment with rhEPO.

Our results indicate that bupivacaine myotoxicity combines two deleterious synergistic effects, leading to a reduction in the activity of respiratory chain complexes I–V and diminished ATP synthesis. This had already been suggested from previous studies with isolated mitochondria.6,7Our work validates those findings in a physiologic and clinically relevant rat model of reiterative exposure to bupivacaine. We also reproduced our findings using human myoblasts in culture. In contrast with previous studies using tumor-derived cell lines to investigate the toxicity of lidocaine and bupivacaine,9our cell model was obtained from primary human myoblasts. The use of cancer cells might have been misleading because of their already abnormal bioenergetic properties and altered mitochondrial features. The general energy crisis caused by bupivacaine as observed in rats and HSMMs could be caused by the high liposolubility of bupivacaine and its accumulation in mitochondrial membranes. The toxic mechanisms of bupivacaine could also involve slipping of either the respiratory chain and/or the F¹F⁰ATP synthase, because we observed a reduction of ATP/O. At the molecular level, the tertiary amine of bupivacaine could potentially interact with respiratory chain complexes and/or the F¹F⁰ATP synthase to inhibit activity.27Hence, mitochondrial structural changes induced by bupivacaine could be caused by a direct interaction with F¹F⁰, given the close link between this enzyme complex and cristae formation.28,29Indeed, the onion-like structure of mitochondria observed in our study is typically associated with abnormal F¹F⁰oligomerization.30All rats that received bupivacaine without rhEPO showed significant inhibition of ATP production in the muscular region surrounding the catheter and exhibited changes in mitochondrial structure. These unusual morphologies could be consistent with autophagy, as indicated by the systematic observation of membrane structures presumed to be autophagosomes containing abnormal mitochondria. Abnormal mitochondrial autophagy was also recently described in various pathologic conditions, but it remains poorly understood.31 

Current hypotheses propose that a decrease in ΔΨ can activate mitophagy.32Our cell culture model confirmed that bupivacaine exposure led to a decrease in ΔΨ,6an effect that is well known with lidocaine.33,34This observation might explain the fragmentation of the mitochondrial network, because fusion and fission mechanisms depend on ΔΨ.35Moreover, a marked decrease in mitochondrial membrane potential with 0.001 and 1 mm bupivacaine at 24 h was shown, whereas no effect of these concentrations on MTT reduction was observed at 24 h. This might be due to the existence of nonmitochondrial sites of MTT reduction, or endocytosis and extrusion of MTT,36or the fact that MTT reduction is not directly dependent on ΔΨ, because there is no evidence that ΔΨ controls respiratory chain complex II activity. The results of the MTT assay used in our study to evaluate the impact of bupivacaine on cell metabolism should be considered with caution. This assay does not give a rigorous measure of cell proliferation rate or cell viability, because it is based on the intracellular reduction of the added tetrazolium salt MTT to a colored water-insoluble formazan salt by the mitochondrial enzyme succinate dehydrogenase. Therefore, it cannot distinguish between cytotoxic molecules and mitotoxic compounds. Therefore, the inhibition of MTT reduction induced by bupivacaine could be interpreted as an alteration of mitochondrial metabolism and/or an inhibition of cell viability. Moreover, inhibition of mitochondrial respiratory chain complexes (genetically or pharmacologically) is usually compensated by different mechanisms so that MTT reduction can be maintained despite a large OXPHOS deficiency.37An important threshold of respiratory chain inhibition must be passed to observe changes in cellular homeostasis. This phenomenon was recently highlighted in cells treated with rotenone, a complex I inhibitor.35 

The cell culture model allowed measurement of the toxic dose of bupivacaine, giving values between 1 and 3 mm. This toxic dose, however, cannot be extrapolated to in vivo  experiments or clinical practice. Indeed, these concentrations are likely much higher than the bupivacaine concentration described in psoas muscle at 1 h after the last injection.10This shows that there is a margin of safety above which the inhibition of cellular energy metabolism by bupivacaine triggers cell death, as reported in different experimental and pathologic situations with mitochondrial pathologic impairment.37 

The molecular mechanisms of the protective effects of rhEPO against bupivacaine-induced myotoxicity include mitochondrial ΔΨ maintenance, as observed in our cell culture model. rhEPO might delay both fragmentation of the mitochondrial network and the initiation of either apoptosis and/or mitoptosis.38,39In addition to this supposed antimitopotic effect, rhEPO restored the efficiency of OXPHOS in rats treated with bupivacaine and prevented the reduction of OXPHOS activity and mitochondrial content, suggesting that preservation of mitochondrial membrane organization stops the apoptotic pathway. This would be a novel physiologic role for rhEPO at the level of mitochondrial energetics and structure (see model fig. 9). With regard to the mechanism, citrate synthase measurements indicated that rhEPO prevented the decrease in mitochondrial content, implying a role in mitochondrial biogenesis or degradation. rhEPO alone did not improve citrate synthase activity, indicating no activation of mitochondrial biogenesis. Western blot analysis of respiratory chain complexes (fig. 2) validated this finding. Therefore, rhEPO could prevent the degradation of mitochondria triggered by bupivacaine, perhaps explaining the recovery of both the ATP/O ratio and citrate synthase. This indicates that rhEPO could interact with mitochondrial degradation processes to prevent the decrease in citrate synthase activity. However, no effect of rhEPO was reported on regulation of the mitochondrial energy system and biogenesis, which involve mitogen-activated protein kinase signaling, the peroxisome proliferator–activated receptor γ coactivator-1αpathway, and various posttranslational modifications (e.g. , phosphorylation, nitrosylation) of complexes I and IV. Most studies of rhEPO and mitochondria have described activation of phosphoinositide-3 kinase/Akt and extracellular signal–regulated kinase 1/2 in the regulation of mitochondrial apoptotic processes.40,41Alternatively, our results suggest that rhEPO might prevent bupivacaine-induced degradation of mitochondria (i.e. , recovery of citrate synthase activity and absence of autophagosomes) through a possible interaction with the regulatory pathway of mitochondrial autophagy. Indeed, among the different kinases activated by rhEPO,42extracellular signal–regulated kinase/mitogen-activated protein kinase plays a role in autophagy.43 

Our study is the first to show that rhEPO can protect against LA-induced myotoxicity. Interestingly, rhEPO is often used in the perioperative period for its hematopoietic effects.44In our study, pretreatment with 5,000 U/kg rhEPO partially prevented myotoxicity in both rats and cultured human myoblasts. The rhEPO doses used in these experiments were an order of magnitude greater than those typically used clinically to treat anemia.45Interestingly, rhEPO had no effect in our cell model on preventing bupivacaine-induced inhibition of endogenous cell respiration, suggesting a time-dependent protective effect. Likewise, higher doses at 8,000 U/kg were shown to be beneficial for neuroprotection in patients with progressive multiple sclerosis,46with no deleterious secondary effects being reported in that preliminary study. A recent review highlights that the cardioprotective effects of rhEPO are species specific and have not been seen in pig and sheep models, and that initial human trials have been disappointing, with definitive study results still pending.42Moreover, our findings suggest that the mitochondrial protective effects of rhEPO should be carefully evaluated in the general context of mitochondrial pathologic dysfunction.47,48rhEPO utilization at high doses or for long periods of time, however, requires rigorous and continuous evaluation during treatment to assess the risk of hypertension,49venous thromboembolism, and mortality, which are routinely described in the elderly and in patients with cancer.50The risk of rhEPO side effects combined to the extremely high cost of rhEPO has to be considered when minor clinical implications such as bupivacaine-induced myotoxicity are being discussed.45,51 

In conclusion, our findings demonstrate that bupivacaine-induced myotoxicity is associated with changes in mitochondrial structure and function in vivo . We show that rhEPO could protect against bupivacaine-induced myotoxicity, suggesting the existence of a protective mechanism in mitochondria. rhEPO not only reduced unusual morphologies consistent with autophagy but also partially rescued the bioenergetic suppression caused by bupivacaine. We examined both the dose dependence and time dependence of bupivacaine and rhEPO interactions with mitochondrial function. The clinical impact of our results remains to be evaluated in practice.

The authors thank Louis Rachid Salmi, M.D., Ph.D. (Senior Research Associate, INSERM U593, Université Victor Segalen Bordeaux 2, Bordeaux, France), for statistical support; and Ray Cooke, Ph.D. (Senior Research Associate, Département des langues vivantes, Université Victor Segalen Bordeaux 2), and Devin Oglesbee, Ph.D. (Senior Research Associate, Department of Laboratory Medicine and Pathology, Mayo Clinic College of Medicine, Rochester, Minnesota), for proofreading the manuscript.