Postoperative Analgesia from Intrathecal Neostigmine in Sheep

Abstract

Spinal neostigmine produces analgesia in chronically prepared rats, but not in sheep.However, since pain itself activates bulbospinal inhibitory pathways, neostigmine may be more effective in the postoperative period. We examined in sheep the antinociceptive effect of intrathecal neostigmine in the acute postoperative period and determined the muscarinic receptor subtype activated by neostigmine. A cervical intrathecal catheter was inserted via a laminotomy in 14 sheep that then received, in random order 1 mg of spinal neostigmine or saline on postoperative Day 1 and the other injection on postoperative Day 2. Three additional sheep received, on separate days, intrathecal neostigmine alone or with the muscarinic receptor subtype-specific antagonists pirenzepine (M1) 2 mg or AFDX-116 (M2) 2 mg. Antinociception was tested using a mechanical stimulus after each injection. Baseline withdrawal threshold did not change postoperatively. Intrathecal neostigmine, but not saline caused antinociception on both of the first two postoperative days. In contrast, intrathecal neostigmine caused no antinociception in another similar study performed at least 5 days after surgery. Pirenzepine, but not AFDX-116, abolished antinociception from neostigmine, suggesting an action on M1 subtype muscarinic receptors. Intrathecal neostigmine is antinociceptive in sheep during the acute postoperative period, and these data suggest that spinal cholinergic tone, and hence intrathecal neostigmine's analgesic effect, may be enhanced during the acute postoperative period.

(Anesth Analg 1995;80:1140-4)

Intrathecal and epidural injection of opioids is commonly employed to achieve selective analgesia without other sensory blockade ^[1]. Because of concern over intraspinal opioid-induced side effects, especially delayed respiratory depression, other classes of drugs are being examined for spinal analgesia. This report describes analgesia from intrathecal injection of the cholinesterase inhibitor, neostigmine, in sheep.

Cholinergic agonists cause behavioral analgesia in animals after intrathecal injection ^[2], presumably by mimicking the release of acetylcholine (ACh) from intrinsic spinal cholinergic neurons located deep in the dorsal horn which send a dense network of fibers to the superficial dorsal horn ^[3,4]. Analgesia from ACh and synthetic cholinergic agonists is inhibited by muscarinic, but not nicotinic antagonists, indicating an action on muscarinic receptors ^[2]. The subtype(s) of muscarinic receptors activated to cause analgesia has not, however, yet been elucidated.

Intrathecal injection of a cholinesterase inhibitor along with ACh potentiates and prolongs behavioral analgesia from ACh ^[2]. Whether cholinesterase inhibitors alone cause analgesia would depend on the degree of tonic activity of cholinergic neurons, since one would expect little effect of cholinesterase inhibitors in the absence of spontaneous release of ACh. Rats exhibit tonic spinal cholinergic activity ^[5], and intrathecally administered cholinesterase inhibitors alone cause analgesia in this species ^[6]. In contrast, intrathecally administered neostigmine alone causes no analgesia in chronically prepared sheep, although it does enhance analgesia from other drugs which stimulate ACh release ^[7].

Pain itself stimulates modulatory systems in the spinal cord and midbrain. Some of these systems enhance the perception of pain, and the hypothesis that "preemptive analgesia" will block the activation of these systems and diminish postoperative pain is actively being tested clinically ^[8]. Other systems are activated by a painful stimulus to diminish pain perception ^[9-11] and may explain in part the lack of pain occasionally observed in the initial period after traumatic injury.

We hypothesized that tonic spinal cholinergic activity will increase in sheep in the immediate postoperative period, and could result in analgesia from intrathecally administered neostigmine alone during this period. The purposes of this study were, therefore, to determine behavioral analgesia from intrathecal neostigmine in the first 2 postoperative days after laminotomy surgery, and to determine the muscarinic receptor subtype activated by neostigmine to cause analgesia.

Methods

After approval from our Animal Care and Use Committee, a total of 17 ewes of mixed Western breeds, weighing between 45 and 60 kg, were studied. After a 24-h fast, anesthesia was induced with ketamine, 20-30 mg/kg, intramuscularly, and maintained with 0.5%-1.5% inhaled halothane. The animal was turned prone, a laminotomy was performed at the C1-2 interspace and a 21-gauge Portex Trademark catheter (Braun Medical, Allentown, PA) was inserted under direct vision 8 cm caudad into the intrathecal space. After wound closure and securing the intrathecal catheter, anesthesia was discontinued and the animal recovered in a metabolic cart. Prophylactic analgesics were not administered, although flunixin, 1.1 mg/kg intravenously was available for behavioral evidence of pain. In all cases, animals were standing and eating within 1 h of completion of surgery, and no animal demonstrated behavioral evidence of pain requiring treatment.

Animals were tested prior to drug injection, then at 30-min intervals for 2 h after drug injection using a noxious mechanical stimulus to the forelimb. As previously described ^[7,12], this method uses a device strapped to the forelimb through which progressive pressure is applied to a blunt metal pin which protrudes against the skin until the animal lifts that leg, at which time the pressure is recorded and immediately released. A cutoff pressure of 20 N is not exceeded in order to avoid tissue damage during periods of intense analgesia.

To test the analgesic action of intrathecal neostigmine, 14 sheep were randomized into two groups of seven ewes each. One group received neostigmine methylsulfate (1 mg) on the first postoperative day and an equal volume of preservative-free saline (0.5 mL) on the second postoperative day. The other group received these injections in the reverse order: saline on postoperative Day 1 and neostigmine on postoperative Day 2. Antinociception in these animals was compared to that of 14 other sheep that were tested 60 min after receiving intrathecal neostigmine 1 mg as part of another study during the same time in this laboratory ^[13]. Sheep in that study had recovered a minimum of 5 days after cervical laminotomy for insertion of the intrathecal catheter.

Three sheep in the current study were noted on subsequent testing to exhibit continued antinociception from intrathecal neostigmine beyond the first 2 days. To test the muscarinic receptor subtype(s) activated by intrathecal neostigmine, these three sheep received, in random order and beginning at least 5 days after surgery, intrathecal neostigmine 1 mg with saline, or with the specific M1 muscarinic antagonist pirenzepine 2 mg, or with the M2 antagonist AFDX-116 2 mg. Experiments were separated by at least 48 h and continued ability of intrathecal injection of neostigmine to produce antinociception was confirmed at the end of the all experiments in the series.

Neostigmine methylsulfate was purchased from International Medication Systems, Ltd. (El Monte, CA). Pirenzepine was purchased from Sigma Chemical Co. (St. Louis, MO). AFDX-116 was a gift from Boehringer-Ingleheim (Ridgefield, CT). All drugs were dissolved in preservative-free normal saline and injected in a 0.5-mL volume followed by 0.5 mL normal saline flush (2 times catheter deadspace). Drug doses used in the current study were based on previous studies in sheep, in which neostigmine caused hemodynamic effects after intrathecal injection which was antagonized by AFDX-116 ^[14].

Data are expressed as mean +/- SEM. Antinociception data were converted to percent maximum possible effect (%MPE) according to the standard formula: Equation 1

Effect of treatment compared to baseline was determined by one-way analysis of variance (ANOVA) followed by Dunnett's test. Treatments were compared by two-way ANOVA for repeated measures, excepting the comparison of antinociception between the animals in the current study and those recovering more than 5 days after surgery. Since these data were not normally distributed they were analyzed by the Kruskal-Wallis test followed by Dunn's test. P < 0.05 was considered significant.

Results

Baseline withdrawal threshold did not differ preoperatively (7.0 +/- 0.5 N) from those on postoperative Day 1 (7.6 +/- 0.7 N) or postoperative Day 2 (7.0 +/- 0.7 N). Intrathecal saline caused no effect on mechanical thresholds on either postoperative day Figure 1 and Figure 2. In contrast, intrathecal neostigmine increased the threshold to withdrawal response on both days Figure 1 and Figure 2. The maximum effect of intrathecal neostigmine occurred 90 min after injection (range, 30-90 min). There was a suggestion of a more profound effect of intrathecal neostigmine on postoperative Day 1 than on postoperative Day 2, since the maximal effect after injection was numerically greater on postoperative Day 1 (56% +/- 19% MPE on postoperative Day 1 vs 34% +/- 14% MPE on postoperative Day 2) and since withdrawal thresholds remained significantly above baseline 120 min after intrathecal neostigmine injection on postoperative Day 1, but not on postoperative Day 2. However, the antinociception versus time curves after intrathecal neostigmine on postoperative Days 1 and 2 did not differ by two-way ANOVA. Neostigmine's effect 60 min after injection was greater on the first postoperative day than in animals receiving neostigmine at least 5 days after surgery Figure 3.

Figure 1: Antinociception, expressed as percent maximum possible effect (%MPE) after intrathecal injection on the first postoperative day at time 0 of saline (fill square) or neostigmine 1 mg (open circle). Each point represents the mean +/- SEM of seven animals. *P < 0.05 vs baseline; **P < 0.01 vs baseline.

Figure 2: Antinociception, expressed as percent maximum possible effect (%MPE) after intrathecal injection on the second postoperative day at time 0 of saline (fill square) or neostigmine 1 mg (open circle). Each point represents the mean +/- SEM of seven animals. **P < 0.01 vs baseline.

Figure 3: Withdrawal thresholds at baseline (left) and 60 min after intrathecal injection of neostigmine 1 mg (right) in animals on the first and second postoperative days (n = 7, current study) and in those studied at least 5 days after surgery [n = 14 ^[13]]. Each box represents the 25th-75th percentile of values, with the median shown as the horizontal line within the box, and the full range of values as the vertical lines extending above and below the box. *P < 0.05 vs neostigmine on postoperative Day 1.

Intrathecal injection of the M1 muscarinic receptor antagonist, pirenzepine, completely antagonized the antinociceptive effect of intrathecal neostigmine, whereas the M2 muscarinic antagonist, AFDX-116, was without effect Figure 4.

Figure 4: Antinociception, expressed as percent maximum possible effect (%MPE) after intrathecal injection of neostigmine 1 mg, with saline (fill square), pirenzepine 2 mg (fill triangle), or AFDX-116 2 mg (open circle). Each point represents the mean +/- SEM of three animals. *Neostigmine-saline and neostigmine-pirenzepine curves differ at all times after baseline by two-way analysis of variance.

Discussion

Intrathecal neostigmine 1 mg injected at least 5 days after surgery, failed to produce antinociception in previous studies in sheep using the same mechanical testing system as used in the current study ^[7,13]. The observation that this same neostigmine dose did cause antinociception in the acute postoperative period raises important mechanistic and clinical implications.

Pain itself activates a pain-inhibitory system at the level of the spinal cord. Initially described under the term, "diffuse noxious inhibitory control" by Le Bars et al. ^[9], painful stimulation in one part of the body was shown to decrease nociceptive responses throughout the body. This effect is due to a spinalsupraspinal-spinal loop ^[15], and the descending inhibitory systems activated may include a monoaminergic pathway ^[11].

Gordh et al. ^[6] recently hypothesized that spinally released norepinephrine from this pathway activates intrinsic spinal cholinergic neurons to cause ACh release, which produces analgesia. Thus, pharmacologic stimulation of descending noradrenergic pathways produces behavioral analgesia which is blocked by intrathecal injection of the noradrenergic antagonist, phentolamine, and also by intrathecal injection of the cholinergic antagonist, atropine ^[16]. In addition, intrathecal injection of alpha_2-adrenergic agonists, which mimic the effect of spinally released norepinephrine, increases ACh concentrations in lumbar cerebrospinal fluid (CSF) in sheep and humans ^[7], and causes behavioral analgesia in rats which is antagonized by atropine ^[6].

One would expect that analgesia from intrathecally administered cholinesterase inhibitors would be influenced by the degree of activity of spinal cholinergic neurons. As such, intrathecal neostigmine, injected at least 5 days after surgery, causes no analgesia alone in sheep, but enhances intrathecal clonidine-induced increases in CSF ACh and antinociception ^[7]. Based on the current study and previous examinations of diffuse noxious inhibitory control, we hypothesize that postoperative pain itself enhances spinal cholinergic tone and hence the analgesic effect of intrathecal neostigmine Figure 5. This hypothesis is directly supported by Zhuo and Gebhart ^[11] who demonstrated that noxious visceral stimulation produced antinociception to a thermal test in rats, and this antinociception was enhanced by intrathecal injection of the cholinesterase inhibitor, physostigmine, and partially blocked by the muscarinic antagonist, atropine.

Figure 5: Role of descending inhibition in response to pain. Painful stimuli, such as in the postoperative state, enter the spinal cord and are relayed from the dorsal horn to central sites, where they activate a variety of nuclei in the brainstem, 1. This results in increased activity of descending serotoninergic and noradrenergic pathways, 2, which have direct inhibitory effects within the spinal cord dorsal horn, 3, and excite spinal cholinergic interneurons, 4, which also inhibit nociceptive transmission.

This study raises a variety of questions appropriate for future study. It is conceivable that the apparent potency of intrathecal neostigmine for postoperative analgesia is influenced by the nature of surgery, being greater after more extensive and painful procedures than after minor ones. In addition, the shift in potency of intrathecal neostigmine after surgery may depend on the basal spinal cholinergic tone before surgery. For example, this shift was obvious in sheep, which demonstrate no analgesia from intrathecal neostigmine 5 days after surgery. However, whether neostigmine's apparent potency would increase more or less after surgery in species which demonstrate clear analgesia from intrathecal neostigmine in the unstimulated state [e.g., rats ^[6] and humans; unpublished observations] is unclear. Also, whether intrathecal neostigmine increases ACh concentrations in CSF or in the dorsal spinal cord more in the postoperative period than at a time remote from surgery is not known, and deserves examination.

The second aim of the current study was to define the muscarinic receptor subtype responsible for analgesia after intrathecal neostigmine. Previous ligand binding studies have demonstrated both M1 and M2 muscarinic receptor subtypes in the superficial dorsal horn ^[17], and previous behavioral studies in rats have yielded evidence consistent with analgesia from each subtype ^[18]. In contrast, the current study in sheep is consistent with a predominant or exclusive role of M1 receptor subtypes in spinal cholinergic analgesia in this species. Interestingly, the spinal hemodynamic effects of intrathecally administered neostigmine in sheep are due to actions on M2 muscarinic receptors in the intermediolateral cell column ^[14], suggesting that cholinergically mediated analgesia and hemodynamic effects could be separated in this species with receptor-selective agonists.

In summary, intrathecal neostigmine, but not saline, causes clear antinociception in the first 2 postoperative days after cervical laminotomy in sheep. Since intrathecal neostigmine failed to cause antinociception when injected at times more remote from this surgery, the antinociception observed in the current study is most likely due to the increased spinal cholinergic tone during the acute postoperative period. The muscarinic receptor subtype responsible for this antinociception in sheep is most likely of the M1 type.