Bupivacaine increases the rate of motoneuron death following peripheral nerve injury

1 Introduction

Appropriate management of pain after an injury or surgical procedure has been shown to improve patient outcomes and reduce hospital costs by hastening recovery (Hadzic et al., 2004; Hadzic et al., 2005). Both surgeons and anesthesiologists acknowledge that there are many advantages to peripheral nerve blocks, including superior early recovery compared with general anesthesia (decreased pain, nausea, vomiting, time to discharge home, rate of unplanned admission), better pain control than systemic opioids and facilitated aggressive rehabilitation, improving long-term functional outcomes (Hadzic et al., 2004; Hadzic et al., 2005; Capdevila, Ponrouch, & Choquet, 2008; Richman et al., 2006).

While infrequent, perioperative nerve damage as a complication from regional anesthesia can be devastating. Peripheral nerve injury resulting in peripheral neuropathy can manifest as prolonged or permanent numbness or transient or permanent motor weakness. Such injuries can cause both emotional and physical stress to patients and require additional healthcare interventions such as doctor visits, testing, and therapy. In the worst-case scenarios, peripheral neuropathy can lead to permanent disability. The reported incidence of long-term nerve injury after peripheral nerve blocks vary from 0.02% – 15% depending on the definition of injury and length of follow-up, and this makes it extremely difficult to obtain reliable and consistent data about the incidence (Auroy et al., 2002; Brull, McCartney, Chan, & El-Beheiry, 2007; Barrington et al., 2009; Selander, 1986; Liguori, 2004; Liu et al., 2009; Fredrickson & Kilfoyle, 2009a). In the uncommon cases of perioperative nerve injury, the question of mechanism remains unknown and many speculate it is likely multifactorial. We propose that additive effects of local anesthetic neurotoxicity in the setting of pre-existing nerve pathology may account for post-block peripheral nerve injury.

It has been well documented that all local anesthetics are neurotoxic and can cause structural damage to neural tissue at high concentrations (Lambert, Lambert, & Strichartz, 1994; Werdehausen et al., 2009; Radwan, Saito, & Goto, 2002; Perez-Castro et al., 2009; Kalichman, Powell, & Myers, 1989). At doses used clinically, it is assumed that local anesthetics are safe, however at high concentrations or prolonged exposure times (i.e., continuous infusions, epinephrine use), local anesthetics can cause growth cone inhibition and collapse, neuron death, Schwann cell death, demyelination, and infiltration of inflammatory cells (Selander, 1986; Lambert et al., 1994; Werdehausen et al., 2009; Radwan et al., 2002; Perez-Castro et al., 2009; Kalichman et al., 1989; Kroin et al., 2010). Because local anesthetics are frequently applied to sites where peripheral nerves may already be diseased or injured (e.g., chemical, burn, mechanical injury or neurodegenerative disease), understanding their effects on injured neurons may have important implications for clinical practice. The risk of performing regional anesthetic techniques in patients with underlying nerve damage or disease must therefore be considered. The current study examined whether injured neurons are more susceptible to the toxicity of local anesthetics. The purpose of this study was to determine if local anesthetics exacerbate the rate of motoneuron death following axotomy.

2 Materials and methods

Adult male C57B/6 mice were obtained from Jackson Laboratories (Bar Harbor, ME). All mice were housed and manipulated in accordance with institutional and National Institutes of Health guidelines, and approval was obtained from the Animal Care and Use Committee to conduct the following experiments. Animals were provided autoclaved food pellets and water ad libitum, and were housed under a 12 hr/12 hr light/dark cycle in microisolater cages contained within a laminar flow system to maintain a pathogen-free environment. After arrival, mice were permitted 1 week acclimation to their environment before experimental manipulation. All experimental manipulations were performed <4 hr into the light cycle under aseptic conditions.

Mice were anesthetized with 2% isoflurane and the right facial nerve of each animal was exposed at its exit from the stylomastoid foramen, transected with microscissors, and the cut ends displaced to prevent reconnection. The left facial nerve was exposed, but not transected, and served as an internal control. At the time of axotomy, solutions of normal saline, 2% lidocaine, or 0.75% bupivacaine was applied directly to the injury site via syringe. The total dose for each local anesthetic was based on the maximal suggested dose used clinically by weight: Lidocaine (7μg/g) and Bupivacaine (5μg/g). This dose was delivered in a total volume of 50μl. There were a total of 8–10 mice per experimental group. Behavioral assessment for the completeness of the axotomy via inspection of eye-blink reflex and vibrissae movement was accomplished at the time of surgery and each week after the surgery until the animal was euthanized.

Four weeks after axotomy the animals were euthanized by CO₂ asphyxiation. The brains were removed and blocked to include the entire facial nucleus then rapidly frozen. Cryostat sections (25μm) were taken from the entire rostrocaudal extent of the facial motor nucleus, fixed with 4% paraformaldehyde, and stained with thionin to reveal cell bodies. The internal genu of the facial nerve is visible within the brainstem, and its location on the slides was used to align matching axotomized and uninjured control side sections. Surviving thionin-stained motoneurons exhibit densely stained and large cytoplasm, clear nuclei and dense nucleoli. The motoneurons within the facial motor nucleus were counted on a minimum of 8 matched sections per animal.

3 Results

To determine if commonly used local anesthetics affect neuronal survival after a peripheral nerve injury, we placed saline, 2% lidocaine, or 0.75% bupivacaine on the proximal nerve stump following a facial nerve axotomy. The average facial motoneuron survival for mice in the control group (saline treatment) 4 weeks after axotomy was 80%, which is similar to previously published data (Figs. 1 and 2). The average facial motoneuron survival in mice treated with lidocaine was 78%, which is virtually identical to the saline control group. In contrast, the average facial motoneuron survival in mice treated with bupivacaine was 35%, a significant decrease of 45% relative to the saline treated control group (F(2,23) = 234.9, p < 0.01 with a critical value of 5.66).

4 Discussion

Peripheral nerve deficits after regional anesthesia are uncommon, however for affected patients, they can be devastating. Post-operative nerve deficits may have multiple etiologies, including surgical manipulations (ie: traction, large volume irrigation, direct surgical insult), post-operative sequelae (ie: edema, inflammatory mediators), or injury related to regional anesthesia (ie: direct needle trauma, intraneural injection, epinephrine-induced ischemia, local anesthetic neurotoxicity). It would seem obvious to assume that the increased use of ultrasound guidance to perform regional anesthesia should improve the safety of these techniques and therefore decrease the incidence of nerve damage. However, while ultrasound has improved the success, onset and quality of peripheral nerve blocks it is interesting that the incidence of nerve injury related to peripheral nerve blocks has not decreased, suggesting that another mehanism may be involved (Liu et al., 2009; Fredrickson & Kilfoyle, 2009a; Fredrickson, Ball, & Dalgleish, 2009b; Lupu et al., 2010; Mariano et al., 2009; Perlas et al., 2008). The current study examined whether injured neurons are more susceptible to the toxicity of local anesthetics. We have shown that bupivacaine placed on a nerve stump at the time of axotomy significantly exacerbates levels of cell death in the affected motoneurons, compared to axotomy alone. This finding is vital and may impact how clinicians choose to provide care to patients.

We hypothesize that pre-existing neuron pathology may predispose patients to postoperative peripheral nerve deficits after a regional anesthetic procedure, such that the combination of two sub-clinical insults (ie: pathology + local anesthetic neurotoxicity) produces a noticeable deficit or worsening symptoms. This additive mechanism has previously been described as the double crush phenomenon in 1973 by Upton and McComas who noted that 70% of patients with carpal tunnel syndrome also had evidence of a cervical root lesion suggesting that the presence of a proximal nerve compression renders the distal nerve more vulnerable to compression. Subsequently, there have been reports that the incidence of peroneal nerve palsy following total knee replacement is increased in patients with significant valgus, a preoperative neuropathy or excessive tourniquet time (Horlocker, 2011). Thus, an improved understanding of the effects of local anesthetics on neuron survival and axon regeneration may lead to strategies to identify patients at higher risk for permanent neural deficits after peripheral nerve blocks and/or decrease the risk of neural deficit following peripheral nerve blocks.

Our data suggest that local anesthetics may contribute to a “double crush phenomenen” on previously injured peripheral nerves. We propose that local anesthetics, known to be neurotoxic, serve as a pharmacologic crush. The neurotoxic properties of local anesthetics have previously been elucidated by predominantly in vitro cell culture or explanted nerve studies. Local anesthetics, cause rapid cell death of cultured neuroblastoma cells via both necrosis and apoptosis in a concentration-dependent manner and at subclinical concentration (Werdehausen et al., 2009; Perez-Castro et al., 2009). Local anesthetics including ropivacaine, lidocaine, and bupivacaine, also cause growth cone collapse, demyelination, Schwann cell dysfunction and death, axonal destruction and infiltration of inflammatory cells (Radwan et al., 2002; Kalichman et al., 1989; Yang, Abrahams, Hurn, Grafe, & Kirsch, 2011; Puljak, Kojundzic, Hogan, & Sapunar, 2009; Hertl, Hagberg, Hunter, Mackinnon, & Langer, 1998; Whitlock et al., 2010). We chose to investigate lidocaine and bupivacaine at their maximal but clinically relavant concentrations based on known data demonstrating their neurotoxic effects on peripheral nerves. Interestingly, we only found a difference in the mice treated with bupivacaine, and not lidocaine. This difference was surprising given the plethora of clinical data showing neurotoxic effects of lidocaine. Bupivacaine is known to be a longer acting local anesthetic than lidocaine (6–8 hours vs 1–2 hours duration of action). Perhaps in our model, a complete nerve transection, coupled with a longer-lasting nerve block of bupivacaine, could account for the increased motoneuron cell death. With the increasing use of peripheral nerve catheters, patients have an extended length of exposure to local anesthetics and this may alter the rate of neurologic complications. Future directions include the elucidation of a mechanism responsible for bupivacaine-induced motoneuron death.

There are many clinically relevant reports in the literature to support the double crush phenomenon, including regional anesthetics that involved a nerve that is already impaired which may have predisposed the patient to a noticiable post-block deficit. Previous studies have investigated the risk associated with neurologic diseases such as polio, multiple sclerosis, and diabetic neuropathy and regional anesthesia and have concluded that the risk of neurologic injury from regional anesthesia techniques is relatively uncommon albeit higher than the general population (Auroy et al., 2002; Hebl, Horlocker, & Pritchard, 2001; Hebl, Horlocker, & Schroeder, 2006a; Moen, Dahlgren, & Irestedt, 2004; Hebl, Kopp, Schroeder, & Horlocker, 2006b). Many times in practice, these pateints are not considered candidates for neuraxial or peripheral nerve blockade because of the assumed risk of exacerbating preexisting neurologic deficits or developing new neurologic dysfunction. Borgeat et al. (2001) found that 1/3 of patients who had persistent symptoms of nerve injury at 1 month following a brachial plexus block for hand surgery actually had an underlying carpal tunnel or sulcus ulnaris syndrome Hebl et al. (2001) described a pharmacologic double crush syndrome in a patient receiving cisplatin therapy who developed diffuse brachial plexopathy after an interscalene block. The authors suggest that cisplatin exposure, known to cause peripheral neuropathy, made the patient’s nerves more susceptible to local anesthetic toxicity. The same authors later report that patients with preexisting spinal canal pathology have higher incidence of neurologic complications after neuraxial blockade (Hebl, Horlocker, Kopp, & Schroeder, 2010). In support of this, an epidemiologic study evaluating severe neurologic complications after neuraxial block reported that undiagnosed spinal stenosis was a risk factor for cauda equina syndrome and paraparesis (Moen et al., 2004). Finally, a recent prospective study compared sciatic nerve block duration in type -2 diabetic and non-diabetic patients and found delayed recovery of both sensory and motor function in diabetic patients (Cuvillon et al., 2013). For anesthesiologists the aforementioned reports imply that patients with diseased or previously injured nerves, including neuropathy from diabetes, peripheral vascular disease or neurotoxic chemotherapy, may be at increased risk for block-related nerve injury.

5 Conclusion

Patient, surgical, and anesthetic risk factors may all contribute to perioperative nerve injury. The increasing prevalence of risk factors for nerve injury (diabetes and obesity), need to be considered when opting to provide a regional anesthetic. We have shown that bupivacaine placed on a nerve stump at the time of axotomy significantly exacerbates levels of cell death in the affected motoneurons, compared to axotomy alone. Thus, it may be worthy to give consideration to avoiding local anesthetics that are more potent, reducing doses and/or concentration and avoiding or limiting vasoconstrictive additives in these at-risk patient populations.