Intrathecal Lidocaine Reverses Tactile Allodynia Caused by Nerve Injuries and Potentiates the Antiallodynic Effect of the COX Inhibitor Ketorolac

A total of 26 male Sprague-Dawley rats (Harlan Industries, Indianapolis, IN), weighing 200–250 g, were used in this study. The number of all rats used in different treatments is summarized in table 1. All surgical procedures were in conformity with the Wake Forest University (Winston-Salem, North Carolina) guidelines on the ethical use of animals, and studies were approved by the Animal Care and Use Committee. Animals were implanted with intrathecal catheters according to the method described previously. 11Under halothane anesthesia (2–4% in oxygen-air), a polyethylene catheter (PE-10, 7.5 cm) was inserted intrathecally through a small puncture made in the atlanto-occipital membrane of the cisterna magna to reach the lumbar enlargement of the spinal cord. Animals were allowed 4 to 5 days to recover from the surgery, and those displaying signs of motor dysfunction (forelimb or hind limb paralysis) were excluded from the study. Lidocaine (100, 200, or 300 μg; Abbott Laboratories, Chicago, IL) was injected through the exteriorized portion of the catheter in 15 μl volume followed by a flush with 10 μl saline, 0.9%. Control rats were only injected with the same volume of saline. To determine whether systemically administered lidocaine is able to reverse established tactile allodynia, lidocaine (300 μg dissolved in 150 μl saline) was injected intraperitoneally in four PSNL rats 3 weeks following PSNL. To determine the effect of prior intrathecal lidocaine injection on the antiallodynic effect of the COX inhibitor ketorolac, 1 week following intrathecal lidocaine injection (100–300 μg), 10 μl ketorolac (0.5%, 50 μg; Allergan, Irvine, CA) was intrathecally injected in these SNL rats.

Rats were anesthetized with 2–4% halothane in oxygen-air. For PSNL, the left sciatic nerve was exposed at the high thigh level, and one third to one half of the nerve was ligated with 6-0 silk suture as previously described. 8For SNL, the left L5 and L6 spinal nerves were exposed and ligated with 6-0 silk suture as described before. 12Before and after surgery, all rats were behaviorally tested to determine the paw withdrawal threshold of both hind paws to mechanical stimuli. Animals were placed in a plastic cage with a wire mesh floor and allowed to explore and groom until they settled. A set of von Frey filaments with bending forces ranging from 1.25 to 30 g was applied, in ascending order, to both plantar hind paws (“up-and-down” method 13). A transient (10–20 min) weakness or paralysis of both hind limbs was seen in almost all rats with 300 μg intrathecal lidocaine injection but only in one third of rats with 200 μg intrathecal lidocaine and was completely absent in rats treated with 100 μg intrathecal lidocaine. Rats receiving either intrathecal saline or intraperitoneal lidocaine exhibited no abnormal behavior. Only after complete recovery from this paralysis were these rats tested behaviorally (2 h after intrathecal lidocaine). Each hind paw was measured three times, and the average values were obtained. Two independent individuals who were blinded to the study groups did the behavioral test. Similar results were obtained from the two examiners.

The mean ± SEM values from both hind paws were determined for each group. The mean values after nerve injury or after injection were compared with prelesion baseline values statistically using a one-way repeated measures analysis of variance with Dunnett multiple comparisons (SigmaStat, v. 2.03; Jandel Scientific Inc., San Rafael, CA). The significance level was set at P < 0.05.

Two weeks after PSNL, the withdrawal threshold of both hind paws of all rats was significantly lower than the baseline value (fig. 1, P < 0.05), indicating that tactile allodynia had developed. Then, 15 μl lidocaine (2%, 300 μg) was intrathecally injected in five rats, while saline in the same volume was injected in another five rats that served as controls. Another four rats received 300 μg intraperitoneal lidocaine. Two hours following intrathecal lidocaine, the tactile allodynia in both hind paws of intrathecal lidocaine-injected rats was significantly reversed (fig. 1). This reversal was also observed 3 days after intrathecal lidocaine. By then, intrathecal lidocaine also had an antinociceptive effect on the ipsilateral hind paw (fig. 1, P < 0.05). However, 1 week after injection, the antiallodynic effect disappeared, and tactile allodynia was restored to preinjection level. At all time points following intrathecal saline injection, tactile allodynia was persistent in all rats (fig. 1, P < 0.05). In four PSNL rats with intraperitoneal lidocaine, no attenuation or reversal of tactile allodynia was observed (data not shown).

Four weeks following SNL, all rats exhibited a significant reduction in the withdrawal threshold of the ipsilateral hind paw when compared to the prelesion baseline (figs. 2A–C, P < 0.05). Two hours and 2 days following 100, 200, and 300 μg intrathecal lidocaine injection, tactile allodynia was markedly reversed to the prelesion level. Three days after injection, tactile allodynia reappeared in the ipsilateral hind paw of 200 and 300 μg intrathecal lidocaine-injected SNL rats (figs. 2B and C, P < 0.05). Although the withdrawal threshold in the hind paw of the 100 μg intrathecal lidocaine-injected SNL rats also declined, it was not significantly lower than the prelesion baseline value (fig. 2A). The withdrawal threshold in the contralateral hind paw of all SNL rats was not significantly different from the prelesion level after either SNL or intrathecal lidocaine injection.

Consistent with a previous report, 9intrathecal ketorolac (50 μg) failed to attenuate the tactile allodynia caused by SNL (fig. 3A). Interestingly, 1 week after intrathecal lidocaine, when tactile allodynia reappeared, intrathecal ketorolac reversed tactile allodynia for 4 h in SNL rats that had previously received 200 (fig. 3B) and 300 μg (not shown) lidocaine. One day after intrathecal ketorolac, its antiallodynic effect disappeared. However, intrathecal ketorolac failed to exert any antiallodynic effect on SNL rats that had received 100 μg intrathecal lidocaine previously (fig. 3B). Intrathecal ketorolac (100 μg) alone also failed to alleviate SNL-induced tactile allodynia but also exhibited the antiallodynic effect 1 week after intrathecal lidocaine (data not shown). The magnitude of the antiallodynic effect exerted by 100 μg intrathecal ketorolac was similar to that induced by 50 μg intrathecal ketorolac 1 week after prior intrathecal lidocaine injection (data not shown). Either intrathecal saline injection 1 week following prior lidocaine injection or intrathecal ketorolac (50 μg) injection 1 week following prior intrathecal ketorolac (50 μg) injection failed to alleviate SNL-induced tactile allodynia (data not shown).

The long-lasting effects of intrathecal lidocaine observed in this study were unexpected and carry important fundamental and methodological considerations for the laboratory study of neuropathic pain states as well as potential clinical implications. Although the mechanisms underlying the antiallodynic effects of lidocaine on animals and patients are unknown, a peripheral mechanism has been proposed. Expression of sodium channel subtypes in afferents is altered following nerve injury, 14,15and ectopic discharges are noted from neuroma sites, afferent fibers, and dorsal root ganglion cells. 4,16,17Lidocaine reduces ectopic discharge after systemic administration at concentrations that fail to block nerve conduction, 3perhaps by acting on unique or up-regulated sodium channel subtypes induced by nerve injury. A noncentral site of lidocaine action is further supported by failure of intrathecal lidocaine to reverse mechanical allodynia in SNL rats. 5In the current study, we observed that 100 μg intrathecal lidocaine, which failed to block motor nerve conduction (signs including any weakness or paralysis of the hind limb), also effectively reversed tactile allodynia for more than 3 days in PSNL rats and less than 3 days in SNL rats. Higher doses of intrathecal lidocaine, 200 and 300 μg, triggered a transient paralysis of both hind limbs in both PSNL and SNL rats. Reversal of tactile allodynia persisted long after paralysis of both hind limbs disappeared, indicating that the blockade of nerve impulse conduction in thick myelinated Aβ axons in the lumbar dorsal root is not likely involved. A previous study showed that local infusion of lidocaine onto injured sciatic nerve failed to relieve thermal hyperalgesia in PSNL rats, 18suggesting that the antiallodynic effect induced by intrathecal lidocaine in PSNL rats is not mediated through blocking conductance of the small-diameter Aδ and C axons in the dorsal roots. Therefore, it is more likely that intrathecal lidocaine has a central suppressing effect at the dorsal horn level. Several lines of evidence also support this assumption. Previous in vivo  19and in vitro  20,21studies showed that lidocaine inhibits spinal neuron activity, likely by blocking sodium and potassium currents evoked in dorsal horn neurons. 21PSNL 22and SNL 23,24induces hyperexcitability in the ipsilateral dorsal horn neurons in rats exhibiting tactile allodynia. Thus, intrathecal lidocaine may suppress this hyperexcitability of the dorsal horn neurons, thus achieving its antiallodynic effect.

We also noticed that the antiallodynic effect induced by intrathecal lidocaine lasted longer and was more effective in the PSNL model than in the SNL model since the antinociceptive effect in the ipsilateral hind paw was also observed by day 3 after injection. These observations suggest that the sensitivity to intrathecal lidocaine depends on the injury model. The reasons for the variability of sensitivity in different injury models are currently unknown. This difference may contribute to the different effects of intrathecal ketorolac on the two models (see the section below entitled Prior Intrathecal Lidocaine Potentiates the Antiallodynic Effect of Intrathecal Ketorolac on Spinal Nerve Ligation Rats).

The etiology of the long duration of intrathecal lidocaine in alleviating tactile allodynia following PSNL and SNL is uncertain but is reminiscent of long-lasting effects observed after intravenous lidocaine administration in animal models 5and in some patients with chronic neuropathic pain. 3It is unlikely that lidocaine remains in spinal tissue for more than a few hours after intrathecal injection. 25Although some have speculated that longer-lasting lidocaine metabolites may be responsible for long-lasting effects, 5the current study is not consistent with this interpretation since lidocaine is metabolized in the liver, and lidocaine metabolites would thus occur only after systemic absorption. However, systemic (intraperitoneal) lidocaine in the current study, which would also be metabolized in the liver, had no effect on tactile allodynia at this dose. Since we observed that intrathecal lidocaine likely interacts with eicosanoid systems in the spinal dorsal horn (see the section below entitled Prior Intrathecal Lidocaine Potentiates the Antiallodynic Effect of Intrathecal Ketorolac on Spinal Nerve Ligation Rats), lidocaine's interaction with other pain-related systems could possibly underlie its prolonged antiallodynic effects. Further study is required to test this hypothesis.

It has been shown previously that COX inhibitors, including ketorolac, when intrathecal injected alone, fail to attenuate SNL-induced neuropathic pain but potentiate morphine. 9,10Consistent with these studies, we confirmed here that intrathecal ketorolac alone had no antiallodynic effect on SNL-induced tactile allodynia. Although we observed that intrathecal ketorolac alleviated PSNL-induced tactile allodynia, 26its antiallodynic effect in SNL rats was observed only following prior intrathecal lidocaine. The potentiating effect was only seen following 200 and 300 μg but not 100 μg intrathecal lidocaine, indicating that the potentiation is dose dependent. However, 50 and 100 μg intrathecal ketorolac exerted the same magnitude of antiallodynia following prior intrathecal lidocaine. Intrathecal saline did not display any antiallodynic effect on SNL rats 1 week after intrathecal lidocaine. Our data strongly suggest that intrathecal lidocaine interacts with eicosanoid systems in the spinal cord following nerve injury. In the spinal cord, both neurons and glia produce prostaglandins (PG). 27The production is enhanced by peripheral inflammation. 28,In vitro  studies show that high-dose lidocaine reduces the production of PGE2 in human gastric mucosa. 29Local lidocaine inhibits the eicosanoid formation in burned skin. 30,In vitro  administration of the local anesthetic ropivacaine dose-dependently inhibits zymosan-induced release of eicosanoid from human granulocytes and endothelial cells, 31and bupivacaine inhibits the EP1 receptor at high concentrations. 32Based on these findings, we speculate that intrathecal lidocaine may either inhibit the production and release of PGs or inhibit EP1 receptors in the spinal dorsal horn, thus down-regulating PG systems. This down-regulation of PG systems may persist following high doses of lidocaine (200–300 μg), even after their antiallodynic effect disappears, hence potentiating the antiallodynic effect of intrathecal ketorolac. However, following prior intrathecal injection of ketorolac, the second intrathecal injection of ketorolac did not alleviate the tactile allodynia. This result indicates that the potentiation of ketorolac's antiallodynic effect by prior intrathecal lidocaine may be mediated through mechanisms other than the inhibition of PG production in the spinal cord. Further studies are certainly warranted to address this issue.

Single intrathecal lidocaine alleviates established tactile allodynia for up to 3 days in PSNL rats and less than 3 days in SNL rats. The alleviation is likely mediated through a central mechanism. Our data also suggest that nerve injury models differ in their sensitivity to intrathecal lidocaine. The use of a test dose of lidocaine to confirm intrathecal catheter location can have long-lasting effects in some models of nerve injury. Thus, the test should be done only at the end of the experiments. The analgesic property of systemic lidocaine in chronic pain states involving nerve damage has made it useful in treating neuropathic pain patients. 3However, its side effects, including dizziness, tinnitus, tremor, and paresthesias, may limit its systemic application. Local anesthetics are the class of drugs most commonly administered with opioids by chronic intrathecal infusion in chronic pain patients. 33,34Intrathecal injection of 100 μg lidocaine, which failed to block motor nerve conduction, yet reversed neuropathic pain as reported in the present study, may provide another approach to alleviate chronic pain in some patients. The dose-dependent potentiation of intrathecal ketorolac by intrathecal lidocaine also provides a new avenue in the future treatment of neuropathic pain.