Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Classical anticonvulsant drugs have a long history of use in treatment of chronic pain, particularly neuropathic pain (McQuay et al., 1995). Gabapentin (Neurontin), a structural analog of -aminobutyric acid (GABA), is a new anticonvulsant that is effective for alleviating neuropathic pain in animal models as well as in patients (Mellick et al., 1995; Rosner et al., 1996; Hunter et al., 1997). However, its pharmacological actions are different from other substances that interact at GABA synapses because it does not bind to GABA receptors or any known neurotransmitter receptors (Rock et al., 1993; Taylor et al., 1998). There has been no substantial evidence as to the primary site(s) and mechanisms of action that contribute to the analgesic effect of gabapentin on neuropathic pain.

Although the pathogenesis of allodynia (normally innocuous stimuli become painful) after a peripheral nerve injury is not fully known, continuous discharges from ectopic foci are known to contribute to maintenance of neuropathic pain (Sheen and Chung, 1993; Matzner and Devor, 1994; Ossipov et al., 1995). Ectopic discharge activity from injured peripheral afferent nerves has been proposed as a driving force for sensitization of spinal dorsal horn neurons and maintenance of hyperexcitable state of central sensory neurons (Yoon et al., 1996). Recently, systemic gabapentin has been shown to inhibit spinal dorsal horn neurons in nerve injured rats (Chapman et al., 1998). It remains to be determined whether therapeutic doses of gabapentin have any effect on spontaneous discharges originated from injured peripheral nerves. A typical pattern of nerve injury-induced neuroma is repetitive bursting discharge activity, which is similar to electrophysiological alterations of central neurons causing epilepsy (Rock et al., 1993; Wamil and McLean, 1994). Thus, we wondered whether gabapentin possesses an inhibitory effect on the ectopic afferent activity generated from injured peripheral nerves. The aim of the present study was to examine the effect of gabapentin on ectopic discharge activity from the injured sciatic nerve and allodynia in an animal model of neuropathic pain.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Male rats (Harlan Sprague-Dawley) weighing 250 to 300 g were used in this study. Under halothane anesthesia, the right jugular vein was cannulated with PE 50 tubing and the catheter was externalized to the back of the neck. The left sciatic nerve was exposed and isolated at midthigh and one-third to one-half of the nerve was ligated tightly with a 5-0 silk suture, according to the method described previously (Seltzer et al., 1990). The animals were allowed to recover for 14 to 18 days before electrophysiological recording or behavioral testing. The surgical preparations and experimental protocols were approved by the Animal Care and Use Committee at Wake Forest University School of Medicine.

To quantify mechanical sensitivity of the paw, rats were placed in individual plastic boxes on a mesh floor and allowed to acclimate for 30 min. A series of von Frey filaments (filament numbers 3.61-5.46; Stoelting Co., Wood Dale, IL) were applied perpendicularly to the plantar surface of the left paw with sufficient force to bend the filaments for 6 s. Brisk withdrawal or paw flinching were considered as positive responses. In the absence of a response, the filament of next greater force was applied. In the presence of a response, the filament of next lower force was applied. The tactile stimulus producing a 50% likelihood of withdrawal response was calculated by using the "up-down" method as described in detail before (Chaplan et al., 1994; Pan et al., 1998). Each trial was repeated two to three times at approximately 2-min intervals, and the mean value was used as the force to produce withdrawal responses. After baseline thresholds of withdrawal response to von Frey filament stimulation were determined, animals received i.v. injections of saline (n = 6) or gabapentin (n = 8) at 15-min intervals. The actual dose of gabapentin injected was 10, 20, 30, and 30 mg/kg to yield a cumulative dose of 10, 30, 60, and 90 mg/kg. Based on a previous study, the elimination half-life of gabapentin from plasma in rats is about 2 to 3 h after i.v. injection (Vollmer et al., 1986). The mechanical thresholds were determined every 15 to 30 min after each injection. In five separate normal rats, we injected i.v. 90 mg/kg of gabapentin to determine whether it alters the paw-withdrawal response to the application of von Frey filaments (i.e., analgesic effect).

Allodynic conditions were confirmed in all rats before afferent nerve recording experiments. Rats were anesthetized with an i.p. injection of sodium phenobarbital (Nembutal, 45 mg/kg). The left carotid artery was cannulated for monitoring the blood pressure. The trachea was cannulated and the rat was ventilated artificially. Arterial blood gases were analyzed with a blood gas analyzer and maintained within physiological limits. Body temperature was maintained in the range of 37-38°C with a circulating water heating pad and heat lamps throughout the experiment.

The fascia and sheath overlying the left sciatic nerve were removed carefully. The nerve then was draped on a platform and covered with warm mineral oil. Small nerve filaments were teased gently from the nerve segment proximal to the ligated site under an operating microscope (model M900; D.F. Vasconcellos S.A., São Paulo, Brazil). Single-unit afferent nerve activity was recorded with a bipolar stainless electrode. The nerve filaments were dissected gradually until single-unit activity of afferents was isolated. The action potential of the afferent was amplified and processed through an audioamplifier (model AM8; Grass Instrument, W. Warwick, RI) and an oscilloscope (model 450; Gould, Cleveland, OH). The neurogram was recorded on a thermal-sensitive recorder (model K2G; Astro-Med, W. Warwick, RI). The single unit was identified initially by examining the wave form and the spike amplitude on the oscilloscope at a rapid sweep speed as well as by checking the recorded sound frequency related to each spike activity. Furthermore, the signals were digitized at a sampling rate of 20 kHz and recorded into a Pentium computer through an analog-to-digital interface card for subsequent off-line analysis. An amplitude threshold was set for the recorded action potential of nerve fibers. When an event was detected, the associated wave form (6 ms) would be extracted and displayed continuously in a separate software oscilloscope window (DataWave Technology, Inc., Longmont, CO). Single-unit recording was ensured by checking the constancy of the shape and polarity of the displayed spike wave form. Discharge frequency was quantified by using data acquisition and analysis software (DataWave Technology), and a histogram was created for each afferent. Accurate counting of the afferent discharge frequency was verified for each afferent by comparing the constructed histogram with the hard copy, which was recorded simultaneously.

After the spontaneous discharge activity of a single-unit afferent from the injured nerve site was identified, the baseline discharge was recorded for 15 to 30 min. Then, saline or gabapentin was injected i.v. at cumulative doses of 10, 30, 60, and 90 mg/kg (the actual dose of gabapentin injected was 10, 20, 30, and 30 mg/kg), each separated by 15 min. The animals were dosed at an interval identical with that used for the behavioral study. We used the following two criteria to ensure that the recorded activity was ectopic discharges originating from the neuromas: 1) recorded nerve fibers had no receptive field in the peripheral tissue, and 2) at the end of recording, the ectopic discharge activity was increased by direct stimulation of the neuroma but was not altered by transecting the nerve distal to the neuroma site. In addition, after observing the inhibitory effect of gabapentin on the ectopic discharges from the injured afferents, we determined whether gabapentin had any effect on responses of normal A- and C-fibers to mechanical stimulation (these normal afferent fibers usually have no spontaneous discharges). Single-unit activity of afferent fibers were recorded from the left sciatic nerve in separate, normal rats. The conduction velocity of normal afferents and injured afferent fibers was measured by electrical stimulation of the sural nerve and the sciatic nerve just proximal to the ligated site, respectively. Conduction time was determined by measuring the time interval from the signal of electrical stimulation to recording of the evoked afferent's action potential, displayed on the oscilloscope. C- and A-fiber afferents were classified as those with a conduction velocity <2.5 and 2.5 to 15 m/s, respectively. After the receptive fields of afferents were precisely located, afferent responses to topical application of calibrated von Frey filaments were examined before and after i.v. injection of 90 mg/kg of gabapentin. Gabapentin (Parke-Davis Pharmaceutical Research, Ann Arbor, MI) was dissolved in normal saline and injected in a volume of 0.2 ml followed by a 0.1-ml flush with saline.

Data are presented as mean ± S.E.M. Discharge activity of afferents was averaged before and after each gabapentin treatment. Paw-withdrawal thresholds in response to mechanical stimulation before and after nerve ligation and evoked responses of normal afferents by mechanical stimulation before and after gabapentin treatment were compared by using a paired Student's t test. The effects of gabapentin on allodynia and afferent activity were determined by analysis of variance followed by the Dunnett's post hoc test. P < .05 was considered to be statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Behavioral Studies. Paw-withdrawal threshold in response to application of von Frey filaments before sciatic nerve ligation was 32.4 ± 2.1 g. The mechanical threshold decreased significantly (4.1 ± 0.7 g, P < .05) within 7 days after nerve ligation and remained stable for at least 3 weeks in all animals studied. Three animals were excluded from the study because the withdrawal threshold was >8 g after nerve ligation. I.v. injection of saline did not affect significantly the allodynic state (n = 6, Fig. 1). Intravenous injection of 30 to 90 mg/kg gabapentin increased significantly the withdrawal threshold in eight other rats in a dose-dependent manner (Fig. 1). The threshold after i.v. injection of 90 mg/kg gabapentin was slightly higher than that obtained before nerve ligation, but such a difference was not statistically significant. Gabapentin administration was not associated with any overt behavioral changes except this increase in withdrawal threshold. Only at a high dose (90 mg/kg), gabapentin appeared to have a slight calming effect on the animals' exploratory behavior. In addition, i.v. injection of 90 mg/kg gabapentin did not change the mechanical threshold of five normal rats. The paw-withdrawal threshold was 35.5 ± 6.3 and 34.3 ± 6.3 g (P > .05) before and after treatment with gabapentin, respectively.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Effect of i.v. injection of saline (n = 6) or gabapentin (n = 8) on mechanical thresholds determined by paw-withdrawal response to von Frey filaments. The points labeled with doses indicate values obtained 15 min after each injection. Data are presented as mean ± S. E.M. *P < .05 versus pretreatment control.

Electrophysiological Recording Experiments. A total of 25 afferents were recorded from the injured left sciatic nerve in 25 additional rats. These afferents exhibited typical spontaneous bursting discharge activity (Fig. 2), as characterized in detail previously (Matzner and Devor, 1994). The conduction velocity was measured in 17 of 25 afferents studied. There were 13 A-fibers with a conduction velocity ranging from 3.0 to 13.8 m/s. The four C-fibers had a conduction velocity of between 0.6 and 1.5 m/s. Repeat i.v. injection of saline did not affect the ectopic discharge frequency of 10 separate afferents during the entire recording period from 1 to 3 h (Fig. 3). Intavenous injection of 10 mg/kg gabapentin failed to influence the spontaneous discharge frequency of afferents. In a cumulative dose range of 30 to 90 mg/kg, gabapentin significantly inhibited the discharge activity of 15 afferents in a dose-dependent fashion (Figs. 2 and 3). At the cumulative dose of 30, 60, and 90 mg/kg, gabapentin suppressed completely the spontaneous discharge activity of three, four, and eight afferent fibers, respectively. We paid attention to changes in interspike intervals of afferent firing and to the on-off bursting cycle. As the dose increased, gabapentin did not change the interspike intervals. Rather, the average duration of off-periods lengthened progressively as that of on-periods shortened gradually (Fig. 2).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Original representative neurograms (each tracing is 30 s of recording) showing dose-dependent inhibitory effect of i.v. gabapentin on single-unit ectopic discharge activity from an injured sciatic afferent fiber. The neurograms were sampled between 12 and 14 min after injection of each dose.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Effect of i.v. injection of saline (n = 10, A) or gabapentin (n = 15, B) on spontaneous discharge activity recorded from injured afferent fibers. Note that in B, the points represent averaged total discharges during the entire recording period between doses. Data are presented as mean ± S. E.M. *P < .05 versus pretreatment control.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Lack of effect of i.v. injection of 90 mg/kg of gabapentin on the responses of normal afferents (n = 12) elicited by topical application of von Frey filaments (VFH). Data are presented as mean ± S. E.M. Receptive fields of all afferents were located in the left hindpaw.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, we explored the neurophysiological mechanisms of the antiallodynic action of gabapentin in a rat model of neuropathic pain. The major finding of the current study is that the ectopic discharge activity from injured peripheral afferent nerve is suppressed by therapeutic doses of gabapentin. We observed that i.v. gabapentin, at a range of 30 to 90 mg/kg, dose-dependently reversed allodynia caused by partial sciatic nerve ligation. Furthermore, we found that similar doses of gabapentin significantly inhibited the discharge activity recorded from injured afferent fibers but had no effect on the conduction velocity and the response of normal afferents to mechanical stimulation. Therefore, these data provide new electrophysiological evidence suggesting that the peripheral action of gabapentin on ectopic discharge activity from injured afferent fibers may constitute an additional mechanism by which gabapentin produces an antiallodynic effect on neuropathic pain.

A prominent feature of central sensitization is allodynia, a state in which normally innocuous input is perceived as pain (Gracely et al., 1992; Yoon et al., 1996). Previous studies indicate that sustained high-frequency discharge from ectopic sites in the peripheral nerve after nerve injury causes enhanced responsiveness of spinal dorsal horn neurons, which contributes toward the pathogenesis of neuropathic pain states (Kajander and Bennett, 1992; Matzner and Devor, 1994; Yoon et al., 1996). Neuropathic pain syndromes after peripheral nerve injuries are often poorly relieved by two major classes of analgesics: nonsteroidal anti-inflammatory drugs and opioids (MacFarlane et al., 1997). In the search for alternative treatment, anticonvulsants have become the more commonly used interventions (McQuay et al., 1995; MacFarlane et al., 1997). Among these agents, gabapentin has been shown to be effective in animals models of neuropathic pain as well as in chronic pain patients (Mellick et al., 1995; Rosner et al., 1996; Field et al., 1997; Rosenberg et al., 1997). It has a much lower incidence of side effects compared with other anticonvulsants (Rosner et al., 1996; Rosenberg et al., 1997). However, the mechanisms and site(s) of its antinociceptive action are largely unclear. Although it has been demonstrated that systemic as well as intrathecal administration of gabapentin has antinociceptive effects in various pain models (Hunter et al., 1997; Hwang and Yaksh, 1997), previous studies have not examined the action of gabapentin on injured peripheral afferents. Chapman et al. (1998) recently reported that s.c. injection of 30 to 100 mg/kg gabapentin inhibits the spontaneous activity of spinal dorsal horn neurons in rats with L₅/L₆ spinal nerve ligation, suggesting a spinal site of action of this agent. Other studies have found that administration of the monoclonal antibody anti-GD2 ganglioside induces ectopic discharge activity from primary afferent fibers, which causes allodynia in rats (Xiao et al., 1997). Because i.v. injection of 30 to 300 mg/kg gabapentin reverses allodynia caused by this antibody, the peripheral action of gabapentin has been proposed (Gillin and Sorkin, 1998). A recent behavioral study has shown that local injection of gabapentin attenuates Formalin-induced nociception in rats, also indicating a peripheral action of gabapentin (Carlton and Zhou, 1998). As demonstrated in the present study, therapeutic doses of gabapentin are capable of suppressing ectopic discharge activity generated from injured afferent nerve sites. Thus, in addition to its effect on sensitized spinal dorsal horn neurons caused by nerve injury, the effect of gabapentin on ectopic afferent activity may contribute to its antinociceptive action by directly eliminating nociceptive afferent input to the spinal cord. We recognize that the antiallodynic effect of gabapentin may occur over a longer period of time after i.v. injection. Our data indicate that gabapentin has a rapid effect on ectopic discharges, which is consistent with its effect on allodynic behavior. There are two possible explanations for this observation. The allodynia produced in this model (partial sciatic nerve ligation) may be highly dependent on the ectopic afferent barrage. Thus, elimination of abnormal input by gabapentin rapidly reversed the allodynic condition. Alternatively, the effect of i.v. gabapentin on ectopic discharges may not account entirely for its antiallodynic effect. The quick antiallodynic effect of gabapentin may be a result of its combined central and peripheral effects. Data from this study provide a new rationale for the use of systemic gabapentin as an analgesic agent in neuropathic pain.

Both voltage-activated Na⁺ channels and voltage-sensitive Ca⁺⁺ channels are closely related to the generation of ectopic discharge activity of injured nerves (Devor et al., 1992; Matzner and Devor, 1994). It is not yet known which channel is functionally relevant to the inhibitory actions of gabapentin on allodynia and ectopic discharges from injured afferent nerves. Recent studies have shown that gabapentin has a high affinity to the ₂-subunit of voltage-sensitive Ca⁺⁺ channels in the brain tissue (Brown et al., 1998). However, in vitro electrophysiological experiments failed to demonstrate any effect of gabapentin on voltage-sensitive Ca⁺⁺ channels (Rock et al., 1993). The effect of gabapentin on Na⁺ channels is still uncertain. Although one study found that gabapentin can inhibit voltage-activated Na⁺ channels in cultured neurons (Wamil and McLean, 1994), others have reported a lack of effect of gabapentin on neuronal Na⁺ channels (Rock et al., 1993). Our in vivo data indicate that gabapentin selectively silenced the ectopic discharge activity from neuromas but did not affect the conduction velocity and the response of normal afferent fibers to tactile mechanical stimulation. Furthermore, because gabapentin gradually altered the on-off bursting cycle without affecting the interspike intervals, these results suggest that the action of gabapentin is inhibiting the impulse generation (neuroma electrogenesis) rather than blocking the impulse propagation (Matzner and Devor, 1994). This mode of action resembles that of sodium channel blockers on neuroma ectopic discharges as reported previously (Devor et al., 1992; Matzner and Devor, 1994). Thus, these data suggest that gabapentin may have a direct or indirect action on the Na⁺ channels at the injured nerve site. Further studies are needed to determine the involvement of ion channels in the action of gabapentin on the generation of ectopic afferent discharge activity after nerve injury.

In summary, i.v. injection of 30 to 90 mg/kg gabapentin attenuated significantly the allodynia induced by partial sciatic nerve ligation in rats in a dose-dependent manner. By recording single-unit activity of sciatic afferent fibers, we found that the same therapeutic doses of gabapentin dose-dependently inhibited the ectopic discharge activity from injured nerve sites. However, responses of normal afferent fibers to mechanical stimulation and the mechanical threshold of normal rats were not affected by gabapentin. Therefore, this study provides new information that systemic gabapentin produces an antiallodynic effect in neuropathic pain, an action that may be mediated at least in part by inhibition of peripheral ectopic afferent discharge activity from injured nerve sites.