Opioid-induced Loss of Local Anesthetic Potency in the Rat Sciatic Nerve

PERIPHERAL nerve block with local anesthetics (LAs) has become an integral part of postoperative pain management enabling better pain control with less opioid use and consequently fewer opioid-related side effects, more rapid utilization of physical therapy, and a reduction in the length of hospital stay.^1,2  And while it is generally appreciated that postoperative pain management in opioid-tolerant patients can be particularly difficult,³  preclinical and clinical studies^4–6  suggest that this is not only due to a decrease in the analgesic efficacy of opioids, but may reflect a decrease in the potency of LAs as well. Such a decrease in efficacy not only further limits the therapeutic options available to these patients, but also increases the risk of serious LA-related side effects including local and systemic toxicity.⁷  Thus, an opioid-induced loss of LA potency may pose a significant clinical challenge, particularly in the face of an increase in opioid use as the first-line treatment for perioperative pain.

Given the number of complicating factors that may influence the appearance of a loss of LA potency in opioid-tolerant patients, we determined whether intrinsic changes in peripheral nerves could account for such a loss of LA potency in opioid-tolerant patients in a rat model of opioid tolerance. Our results suggest that induction of morphine analgesic tolerance was sufficient to drive a three-fold decrease in lidocaine potency, a change that lasted for weeks beyond the last morphine injection, was proportional to the degree of tolerance, and was dependent on central opioid receptor activation.

Adult male and female Sprague–Dawley rats (275 to 350 g; Envigo; Harlan Sprague Dawley, USA) were used for all experiments. Rats were housed in an Association for Assessment and Accreditation of Laboratory Animal Care International–accredited animal care facility with unrestricted food and water on a 12:12-h light–dark cycle (lights on 7:00 am). All procedures involving animals were reviewed and approved by the University of Pittsburgh Institutional Animal Care and Use Committee (Pittsburgh, Pennsylvania). Animal care and handling were in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health (Bethesda, Maryland).

All drugs used in the study were purchased from Sigma-Aldrich, USA, and were used at the following doses: morphine 3, 5, or 10 mg/kg, subcutaneously; naloxone 1 mg/kg, intraperitoneally; and naloxone methiodide 1 mg/kg, subcutaneously. Drugs were dissolved in normal saline, which served as a vehicle control. For the experiments to determine the dose dependence of the morphine-induced changes of LA potency, animals received seven daily morphine injections (3, 5, or 10 mg/kg, subcutaneously; n = 6 for each group). For the experiments to determine the onset of the morphine-induced changes in LA potency, animals received three or five daily morphine injections (10 mg/kg, subcutaneously; n = 6). For the experiments to determine the extent to which recovery from morphine tolerance was associated with the recovery from the loss of LA potency, animals received seven daily morphine injections (10 mg/kg, subcutaneously; n = 6), and then morphine analgesic efficacy and LA potency were assessed 7, 21, or 35 days after the last morphine injection. Morphine analgesic efficacy was assessed 30 min after a subcutaneous injection of morphine based on estimated half-life of morphine analgesia.⁸  For the initial 10-mg/kg experiment, and the onset and dose-dependent experiments, sciatic nerves were harvested 1 day after the last morphine injections. For the recovery experiments, lidocaine potency was assessed within 3 h of the morphine or vehicle injection. For the experiments to determine the role of μ-opioid receptor activation in the loss of LA potency, naloxone (1 mg/kg, intraperitoneally) or naloxone methiodide (1 mg/kg, subcutaneously) were coadministered with morphine (10 mg/kg) on each of the seven daily injections. Changes in LA potency were assessed 1 day later. Because experiments were performed in a blinded manner, no formal randomization scheme was employed. However, because rats were pair housed, we attempted to counterbalance which of the two rats in a cage (first or second) received the same drug treatment.

Because the primary goal of this study was the characterization of changes in lidocaine potency in the isolated peripheral nerve, behavioral data were only collected as a means to assess the analgesic efficacy of systemic morphine administration. The heat withdrawal assay was used for this purpose because the analgesic response to morphine in this assay is robust and well characterized. Animals were placed in a plexiglass test chamber (diameter, 18 cm × 22 cm) on a 3-mm-thick glass plate. The glass surface temperature was maintained at 30°C by a feedback-controlled, resistive heating system embedded in the glass (iiTC Life Sciences, USA). Rats were acclimated to the test chamber for 20 to 30 min before testing. The radiant heat source was from a projection bulb placed directly under the plantar surface of the rat’s right hind paw. The paw withdrawal latency to radiant heat stimulation was defined as the time from onset of radiant heat to withdrawal of the rat’s hind paw. The intensity of the radiant heat source was adjusted to result in baseline latencies of 10 to 12 s. This was achieved with a heating rate over the first 10 s of 3°C/s. A terminal temperature of approximately 54°C was maintained until cutoff. A cutoff of 20 s was employed to avoid tissue injury. The results from three trials (with an intertrial interval of 5 min) were averaged to yield a mean withdrawal latency. Tests were conducted 30 min before and after either saline or morphine injection. These settings were consistent with those used previously.⁹  The experimenter collecting the behavioral data was blinded to the treatment. However, because the response to 10 mg/kg was so pronounced, it was difficult to maintain blinding in the initial morphine versus vehicle experiments. Nevertheless, blinding was effectively maintained for the dose response and opioid receptor antagonist experiments.

Sciatic nerve harvest and compound action potential (CAP) recordings were performed as described previously.¹⁰  Briefly, rats were anesthetized with an intraperitoneal injection of 1 ml/kg of a mixture of ketamine (55 mg/ml), xylazine (5.5 mg/ml), and acepromazine (1.1 mg/ml). The sciatic nerve was removed from ischial tuberosity to tibialperoneal bifurcation (average length = 4.7 cm) and then placed in ice-cold oxygenated Locke solution (in millimolar; 136 NaCl, 5.6 KCl, 14.3 NaHCO₃, 1.2 NaH₂PO₄, 2.2 CaCl₂, 1.2 MgCl₂, 11 dextrose, pH 7.2 to 7.4). The nerve was then transferred to a recording chamber in which the nerve was submerged in Locke solution with the distal end connected to platinum stimulation electrodes and the central end to a glass suction recording electrode. The distal end of the nerve was separated from the central end with a grease gap. The recording chamber was continuously superfused (2 to 5 ml/min) with Locke solution at room temperature with or without lidocaine. The experiments were done at room temperature because we sought to analyze the impact of lidocaine on A and C waves in relative isolation, and this is difficult to achieve at increased (e.g., body) temperatures, especially with relatively short nerves (less than 40 mm). The recording at room temperature may have influenced our results, but note that LA potency at increased temperatures, at least in our hands, is comparable to that observed at room temperature.¹⁰  The CAP was evoked with constant voltage electric pulses of 0.2 to 0.5 ms in duration, applied at 0.05 Hz, where the stimulus intensity was approximately two times that needed to evoke a maximal amplitude C-fiber component of the CAP. The stimuli were applied with an electric stimulator (Model S48; Grass Instruments Co., USA). The recording electrode at the other end of the nerve was connected to the input stage of a differential preamplifier (0.1 to 10 kHz; WPI model DAM-80, USA). Voltage data were digitized via a CED 1401 Micro A/D converter and analyzed using CED Spike 2 version 5 for MS Windows (CED, England). The average of six consecutive CAPs were rectified and integrated to quantify A- and C-fiber components as area under the curve. The A-fiber deflection of the CAP (A wave) was distinguished from the C-fiber deflection (C wave) based on the conduction velocity and amplitude (fig. 1A). After establishing a stable baseline over at least 30 min of recording, lidocaine was applied to the nerve at increasing concentrations, where a stable response was achieved before applying the next higher concentration (fig. 1B). To control for time-dependent changes in the CAP, vehicle-treated nerves were studied over the same time course as required for generation of a complete lidocaine concentration–response curve. As an additional control, the extent to which CAP block reversed after wash of the lidocaine was also determined (fig. 1B). The EC₅₀, or concentration of lidocaine needed to block 50% of the CAP, was used as a measure of potency and was determined from concentration response curves generated with the application of lidocaine at 30, 100, 300, 1,000, and 3,000 μM. These data were fitted using a Hill equation: (1 − CAP_lido/CAP_baseline) = 1/([drug]^nH + EC₅₀^nH), where CAP_lido is the CAP in the presence of lidocaine, CAP_baseline is the CAP before the application of lidocaine, and nH is the Hill coefficient.

The same experimenter (Q.L.) performed the behavioral and electrophysiological experiments. The experimenter remained blinded to drug treatment until all electrophysiological experiments had been completed.

The number of animals used in this study was determined with a power analysis based on variability previously observed in the study of rat sciatic nerves with LAs,¹⁰  and our desire to be able to detect at least a 50% change in potency (EC₅₀) with a power of 0.8 and alpha of 0.05, with four groups analyzed with a one-way ANOVA. While our a priori hypothesis was that there would be a decrease in potency, data were nevertheless analyzed with two-tailed tests. The Holm–Sidak test was used for post hoc comparisons. A and C waves were analyzed separately. Pearson correlation coefficient was used to determine the association between the morphine-induced changes in lidocaine EC₅₀ and the morphine analgesic efficacy. The percent maximum possible effect (%MPE) was used to estimate analgesic efficacy and was calculated according to the formula: %MPE = 100 × (postdrug latency − predrug baseline latency)/(20 − predrug baseline latency). Changes in withdrawal latency over time were analyzed using a mixed-design two-way ANOVA with the treatment as the independent factor and the time as the repeated factor. The Newman–Keuls test was used for post hoc comparisons between groups. Statistical analysis was performed with SPSS 19.0 statistical software package (SPSS, USA). Data are reported as mean ± SD unless otherwise indicated.

Consistent with the results of previous studies, we observed a concentration-dependent block of both A and C waves of the sciatic nerve CAP by lidocaine (fig. 1, C and D). EC₅₀ for lidocaine-induced block of the sciatic nerves from naive rats were 186 ± 32 (n = 6) and 201 ± 30 μM (n = 6), for the A and C waves, respectively. In nerves from rats that received seven daily injections of morphine (10 mg/kg, subcutaneously), there was a significant (P < 0.01) decrease in lidocaine potency for both A (608 ± 53 μM; n = 6) and C (613 ± 42 μM; n = 6) waves (fig. 1, C and D). All animals started in this and subsequent experiments were retained through to the CAP analysis. Consequently, it was not necessary to implement a strategy for dealing with missing data during statistical analysis.

We next sought to determine whether the decrease in LA potency was dose-dependent. Two additional groups of rats were studied that received seven daily injections of 3 (n = 6) or 5 (n = 6) mg/kg morphine. Because female rats are more sensitive to morphine than males,¹¹  two groups of female rats were studied to determine whether there was also a sex difference in the impact of morphine on LA potency. The female rats received seven daily injections of either vehicle or 5 mg/kg morphine. To determine the analgesic efficacy of the morphine administered, heat withdrawal latencies were determined before and 30 min after morphine administration. The first morphine injection was associated with a significant (P < 0.01) increase in withdrawal latency to a duration close to cutoff (approximately 20 s) in all rats tested regardless of dose or sex (fig. 2A). However, the morphine-induced increase in withdrawal latency decreased in each group with subsequent morphine injections such that there was a significant (P = 0.025) loss of analgesic efficacy in each group, by day 7, relative to the increase in latency observed on day 1 (fig. 2A). The withdrawal latency observed in the presence of morphine on day 7 was also significantly different (P = 0.031) between male groups treated with morphine, with the longest latency associated with the 3-mg/kg dose and the shortest latency associated with the 10-mg/kg dose (fig. 2A). Furthermore, while the withdrawal latency in male rats receiving 3 and 5 mg/kg was still significantly longer than either baseline withdrawal latency or saline-treated rats, the withdrawal latency in rats receiving 10 mg/kg was not different (P = 0.264) from baseline latency or that in saline-treated rats. Finally, when withdrawal latency was analyzed as a function of sex and time with female rats compared to male rats receiving either 5 or 10 mg/kg, there was a main effect associated with time (P < 0.01), but no significant influence of sex (P = 0.067) or a significant interaction between sex and time. Thus, the development of tolerance was comparable in males and females.

Lidocaine potency was tested on nerves from each group of rats the day after the last injection. There was a significant (P < 0.01) decrease in lidocaine potency (increase in EC₅₀) in all nerves from morphine-treated rats relative to that in nerves from naive or saline-treated rats for both A (fig. 2B) and C (fig. 2C) waves. Furthermore, the morphine-induced decrease in lidocaine potency was dose-dependent in male rats, such that the greatest decrease in potency was observed in the 10 mg/kg group and the smallest in the 3 mg/kg group (fig. 2, B and C). The decrease in lidocaine potency in female rats receiving 5 mg/kg morphine was not significantly (P = 0.261) different than the decrease in lidocaine potency in male rats receiving 5 mg/kg morphine, but significantly (P = 0.024) less than the decrease in male rats receiving 10 mg/kg morphine for both A and C waves (fig. 2, B and C).

To determine the timing over which the morphine-induced loss of LA developed, two additional groups of male rats were studied, receiving 3 (n = 6) or 5 (n = 6) daily injections of 10 mg/kg morphine subcutaneously. The morphine-induced increase in paw withdrawal latency in these 2 groups mirrored that observed in the group receiving seven daily injections, with the maximum increase in latency observed on day 1 and a smaller increase in latency observed with each subsequent injection (fig. 3A). The increase in latency observed after three daily injections was significantly greater than baseline, but significantly smaller than the increase observed on day 1 (fig. 3A). The same was true after five daily injections. However, the increase in latency after the fifth injection in the 5-day group was also significantly less than that after the third injection in the 3-day group.

There was a significant decrease in the potency of lidocaine-induced block of both the A (fig. 3B) and the C (fig. 3C) waves in sciatic nerves from rats receiving three and five daily injections of morphine, relative to that in sciatic nerves from saline-treated and naive rats. However, the magnitude of the decrease in potency was injection number dependent, where the decrease was the greatest after seven daily injections and the smallest after only three daily injections. The differences in potency between each group were also significant (P < 0.01; fig. 3, B and C).

To explore the relationship between the development of morphine tolerance and the loss of LA potency, we analyzed both dose dependence (3, 5, and 10 mg/kg) and onset (three, five, and seven daily injections) analgesia data as a function of the shift in LA potency. To estimate tolerance, paw withdrawal latency data were transformed to a %MPE, where an increase in latency to cutoff was considered full analgesic efficacy and the absence of a morphine-induced change in latency was considered complete tolerance. EC₅₀ data were then plotted relative to the %MPE (fig. 4). The association between %MPE and EC₅₀ was linear and highly significant with an r² = −0.92, P < 0.01 for A wave and r² = −0.93, P < 0.01 for C wave, where a 50% decrease in analgesic efficacy was associated with a more than 100% decrease in lidocaine potency.

We next sought to determine the time course over which the morphine-induced loss of LA potency recovers after cessation of morphine administration and to determine the extent to which the recovery from tolerance was associated with the recovery from the loss of LA potency. To address these issues, additional groups of rats were first treated with seven daily injections of morphine (10 mg/kg), and then the morphine analgesic efficacy and the LA potency were assessed 7 (n = 6), 21 (n = 4), or 35 (n = 5) days after the last morphine injection. Consistent with the first group of rats tested after seven daily injections of morphine, there was no change in paw withdrawal latency after the morphine injection on day 7 relative to baseline. Furthermore, 7, 21, or 35 days after the last morphine injection, there was no behavioral evidence of tolerance, as morphine-induced increase in paw withdrawal latency was comparable (P = 0.173) to that observed in rats that received their first morphine injections (fig. 5A).

The potency of lidocaine-induced block of A and C waves in sciatic nerves from saline-treated rats challenged with morphine hours before nerves were harvested was comparable to that observed in sciatic nerves from naive rats (fig. 5, B and C). However, the decrease in lidocaine potency observed in all groups of morphine-treated rats was significant (P < 0.01) compared to that observed in saline-treated or naive rats (fig. 5, B and C). Furthermore, there was no evidence of a recovery of the decrease in potency, even after 35 days, as the EC₅₀ in the nerves from this group was comparable (P = 0.304) to that tested the day after seven daily injections of morphine (fig. 5, B and C). This was observed for both A and C waves.

To determine whether the morphine-induced loss of LA potency is due to the activation of μ-opioid receptors, and if so, whether these receptors are located in the periphery, four additional groups of animals were studied. These included rats receiving seven daily injections of naloxone 1 mg/kg, intraperitoneally, or naloxone methiodide 1 mg/kg, subcutaneously alone, or each of these antagonists coadministered with morphine (10 mg/kg, subcutaneously; n = 6 per group). Naloxone would block both central and peripheral opioid receptors, whereas the blood–brain barrier impermeant analog of naloxone, naloxone methiodide, would only block peripheral opioid receptors. As expected, neither naloxone nor naloxone methiodide had any detectable influence on paw withdrawal latency (P = 0.168; fig. 6A). Similarly, there was no detectable (P = 0.072) increase in paw withdrawal latency in rats treated with the combination of naloxone and morphine. However, there was a significant increase in paw withdrawal latency in rats coadministered naloxone methiodide and morphine (fig. 6A).

The potency of lidocaine-induced block of the A (fig. 6B) and C (fig. 6C) waves of the CAP in sciatic nerves from rats treated with naloxone or naloxone methiodide alone was comparable (P = 0.085) to that observed in sciatic nerves from naive rats. There was also no change (P = 0.163) in lidocaine potency in nerves from rats treated with the combination of naloxone and morphine (fig. 6, B and C). However, there was a significant decrease in lidocaine potency in nerves from rats treated with the combination of naloxone methiodide and morphine relative to data from naive rats or from the other three groups in this experiment (fig. 6, B and C). Furthermore, the lidocaine potency in nerves from the rats treated with naloxone methiodide and morphine was comparable (P = 0.173) to that in nerves from the rats treated with morphine alone (fig. 6, B and C).

The purpose of this study was to determine whether morphine tolerance is associated with a loss of LA potency in isolated peripheral nerves. Repeated systemic morphine administration resulted in a decrease in LA potency that was comparable in males and females. This decrease in potency was dependent on the dose of morphine and the number of injections, and consequently, it was significantly correlated with the development of tolerance to morphine-induced analgesia. Strikingly, the loss of LA potency persisted beyond the recovery of tolerance. Finally, the loss of LA potency was blocked by naloxone, but not by naloxone methiodide. These results suggest that the apparent decrease in LA potency observed in opioid-tolerant patients, most clearly documented by Hashemian et al.,⁵  may be due, at least in part, to a decrease in the potency of LA-induced peripheral nerve block.

While there are a number of potential mechanisms that could account for the observed decrease in LA potency, we suggest that our experimental design involving an isolated nerve continuously superfused over virtually its entire length argues most compellingly for a mechanism that involves changes intrinsic to the nerve. The use of the isolated nerve eliminated possible changes in blood flow that could affect lidocaine clearance. Similarly, bathing the whole nerve in lidocaine not only minimized the potential impact of changes in drug access to its target, but also eliminated changes in myelination and/or the density or distribution of ion channels (e.g., Na⁺ and K⁺ channels) that can affect charge spread within an axon, and consequently potency estimates based on focal application of lidocaine. The fact that changes in both K⁺ and Na⁺ channel expressions have been documented in response to repeated morphine injection^12,13  suggests that we may have underestimated the impact of morphine administration on the apparent decrease in LA potency that would be observed with focal LA administration.

While nociceptive testing was necessary to establish the association between the development of analgesic tolerance and the decrease in LA potency, the nociceptive testing necessarily introduced a potential confound that it was not the morphine, but the repeated heating of the hind paw responsible for the changes in both A and C fibers. However, arguing against this possibility are (1) a single injection of morphine, which was the only manipulation in which animals were consistently stimulated to cutoff, was not associated with a decrease in LA potency, (2) the decrease in potency was most robustly manifest in rats in which complete tolerance had been generated, a state in which the rats were responding to the thermal stimuli at the same latency observed in naive rats, (3) there was no evidence of persistent sensitization in animals that had been driven to cutoff, as withdrawal latencies were the same the day after morphine injection as they were at baseline, and (4) the stimulation paradigm, including baseline latency, cutoff, number of stimuli employed to establish baseline, and the interstimulus interval, was comparable to that used extensively by other investigators with no evidence of persistent sensitization.

The primary target of LAs is voltage-gated Na⁺ channels (VGSCs). LAs are thought to access a low-affinity site in channels in the resting or closed state, but a high-affinity site(s) in channels in open and inactivated states.¹⁴  Consequently, any processes that lead to an increase in the fraction of channels residing in the low-affinity state will result in a decrease in LA potency. These processes include (1) a hyperpolarization of the membrane potential (relieving the high-affinity state associated with channel inactivation), (2) a posttranslation-induced change in the biophysical properties of the channels, such that more channels reside in the low-affinity state, and (3) a change in the expression of VGSC subunits such that there is either an increase in the expression of channels with an intrinsically lower affinity for LAs or in channels with different biophysical properties, such that there is a shift in the fraction of channels residing in low- versus high-affinity states.¹⁵  Among these changes, the latter two phenomena have been more clearly documented. For example, the VGSC β subunits can influence both fast and slow inactivation, with changes in the phosphorylation of the channel influencing these properties as well.¹⁶  Similarly, the tetrodotoxin-resistant Na⁺ channel (Nav1.8) is intrinsically less sensitive to LA block than tetrodotoxin-sensitive channels in dorsal root ganglion neurons.¹⁷  Ongoing studies are designed to distinguish which, if any, of these possibilities contribute to the morphine-induced decrease in LA potency.

While both the dose dependence and the time course of the morphine-induced decrease in LA potency suggest that the induction of analgesic tolerance is necessary for the loss of LA potency, these two phenomena may be mechanistically unrelated. Morphine clearly produces a number of changes in the body in addition to the suppression of nociceptive signaling, and the differential time course and/or magnitude of the tolerance that develops in response to repeated morphine administration for several of these changes suggests that underlying mechanisms are distinct.¹⁸  Furthermore, it is unlikely that any of the synaptic mechanisms that have been implicated in the development of analgesic tolerance, such as the activation of N-methyl-D-aspartate receptors¹⁹  or the induction of long-term potentiation,²⁰  contribute directly to the morphine-induced decrease in LA potency because even if it is due to presynaptic changes, additional mechanisms would be needed to account for the decrease in LA potency observed in both A and C fibers. Similarly, the observation that the morphine-induced decrease in LA potency is blocked by naloxone, but not by naloxone methiodide, argues against any of the previously described changes in the primary afferents, such as the increase in transient receptor potential vanilloid type 1, extracellular signal-regulated kinase (ERK), the immediate early gene c-Jun, and the mitogen-activated protein kinase p38,²¹  which have been suggested to contribute to the development of analgesic tolerance. However, whether the same signaling molecules mediate both analgesic tolerance and the decrease in LA potency remains to be determined, the morphine-induced activation of cells in the dorsal horn, such as microglia,^22,23  that are able to release mediators capable of acting back on primary afferents could contribute to the development of analgesic tolerance and the decrease in LA potency in both A and C fibers.

A striking finding from the current study was that the morphine-induced decrease in LA potency persisted for 35 days after the last morphine injection. Minimally, the observation that the morphine-induced decrease in LA potency outlasted analgesic tolerance indicates that while these two phenomena may share a common induction mechanism, the mechanisms controlling the persistence of these changes are distinct. There are at least two general mechanisms that could account for the persistence of the decrease in LA potency in primary afferents. First, there is a persistent change in a cell in the central nervous system (CNS), such as a microglial cell in the dorsal horn that enables the cell to maintain the change in primary afferents. Consistent with this possibility, there is evidence of a repeated morphine-induced increase in histone acetylation in the spinal cord.²⁴  Identification of the signaling molecules mediating the maintenance of the change in primary afferents would suggest a way to restore LA potency.

A second mechanism to account for the persistent change in LA potency would be a persistent change in primary afferents secondary to mediators released in the CNS during the repeated morphine administration. Persistent changes in nociceptive afferents have been described in response to a single exposure of inflammatory mediators, such as prostaglandin E2,²⁵  nerve growth factor, tumor necrosis factor alpha, interleukin 6,^25,26  and most recently, the μ-opioid receptor agonist, mu-opioid receptor agonist D-Ala2, N-MePhe4, Gly-ol]-enkephalin.²⁷  These previously described changes are associated with a prolonged hyperalgesia in response to a subsequent inflammatory challenge, rather than a shift in LA potency. Nevertheless, similar to the results obtained in the current study, the persistent changes in nociceptive afferents were not associated with overt changes in nociceptive behavior and were only manifest upon subsequent challenge. While our naloxone results argue against a role of peripheral μ-opioid receptor activation in the loss of LA potency, it will be interesting to determine whether any of the mechanisms implicated in the prolonged hyperalgesic response contribute to the induction and/or maintenance of the decrease in LA potency.

While we focused on the peripheral nerve in the current study, there are at least two observations suggesting that the changes observed in the peripheral nerve may develop in the CNS as well. There is evidence that the duration of anesthesia achieved with intrathecal administration of 5% lidocaine is significantly shorter in opium abusers.⁴  Similarly, the analgesia associated with a single intrathecal injection of lidocaine is decreased in morphine-tolerant rats.⁶  And while a change in the central axons of primary afferents may contribute to these previous results, it will be interesting to determine whether resistance to LAs develops in dorsal horn neurons, or at even higher centers in the CNS, potentially pointing to a deleterious interaction between opioids and other use-dependent Na⁺ channel blockers used for the treatment of pain.

Our results with 3 mg/kg morphine are likely to have the greatest clinical significance because they suggest that even a moderate dose of morphine that did not produce complete tolerance produced a significant decrease in LA potency. This suggests that there may be a decrease in LA potency even in patients on moderate doses of opioids. The result may be the need for higher LA infusion rates in these patients. Minimally, this suggests that standardized postoperative pain management protocols²⁸  would have to be adjusted for this population in order to achieve optimal postoperative pain management. However, such an increase in LA administration would also be associated with an increase in the risk of LA toxicity. Furthermore, because the decrease in LA potency is persistent, it should not be sufficient to simply get patients off opioids before surgery. Rather, it will be necessary to identify mechanisms underlying the shift or more relevantly its maintenance, so as to be able to reverse them as a way of restoring LA potency. In the meantime, our results suggest that analgesic strategies other than peripheral nerve block should be pursued for opioid-tolerant patients.

The authors would like to thank James Ibinson, M.D., and Eser Yilmaz, M.S., Department of Anesthesiology, University of Pittsburgh, Pittsburgh, Pennsylvania, for valuable input during the preparation of this manuscript.

Supported by grants K12A203475 (to Dr. Liu), R01DE018252 (to Dr. Gold), and R01NS083347 (to Dr. Gold) from the National Institutes of Health (Bethesda, Maryland).

The authors declare no competing interests.