Characterization of the Antihyperalgesic Action of a Novel Peripheral Mu-opioid Receptor Agonist-Loperamide 

These investigations were performed in accordance with the protocol approved by the institutional Animal Care Committee, University of California, San Diego.

Male Holtzman-Sprague-Dawley rats (weight, 275-350 g; Harlan Industries, Indianapolis, IN) were housed in cages with free access to food and water at all times and were maintained on a 12-h light/dark cycle.

The following drugs and formulations were used in the study: loperamide (ADL 2-1294B 5% cream; Adolor Corporation, Malvern, PA), vehicle cream (Adolor Corporation), naloxone hydrochloride (molecular weight = 327.37; Du Pont Pharmaceuticals, Garden City, NY), and morphine pellets (Rowell Laboratories, Inc., Baudette, MN). [10]ADL 2-1294B was mixed with the vehicle cream to form concentrations of 1.7% and 0.5%. The quantity of cream applied at each concentration of loperamide was 100 mg, unless otherwise stated. To assess clearance of loperamide,¹⁴C-labeled loperamide (¹⁴) C-labeled ADL 2-1294B 5% cream; Adolor Corporation) was used (specific activity, 12,000 disintegration per minute [DPM]/mg). Naloxone was dissolved in 0.9% sterile preservative-free saline to form a solution with a concentration of 1 mg/ml.

The hindpaw thermal nociceptive threshold was assessed with a device modeled after that described by Hargreaves et al. [11](George Ozaki, Department of Anesthesiology, University of California, San Diego, La Jolla, CA). [13]Rats were placed individually in acrylic cubicles (9 x 22 x 25 cm) on an elevated floor of clear glass (2 mm thick). A radiant heat source (halogen projector lamp CXL/CXP 50 W 8 V; Ushio, Tokyo, Japan), contained in a movable holder, was placed beneath the glass floor. To reduce the variability in plate surface temperature caused by room temperature, the interior of the box under the animal had a heat source to keep the under-plate temperature at 30 [degree sign]C. The mirror mounted at the stimulus source gave direct visualization of the paw and allowed the light to be focused precisely on the paw. The bulb current (4.8 Angstrom) was chosen to evoke an average response latency in naive untreated rats of 10 +/− 1 s. Details of the implementation of this device are given elsewhere. [12,13] 

To initiate a test, the rat was placed in the acrylic cubicles and was allowed [tilde operator] 30 min to acclimate. The heat source was then positioned to focus on the heel portion of the plantar surface of one hindpaw, which was in contact with the glass. The stimulus was activated, which initiated a timing circuit. The time interval between the application of the light and the hindpaw withdrawal response, which was automatically sensed, [13]was defined as the paw withdrawal latency (PWL). In the absence of a response, the stimulus was automatically terminated at 20 s to avoid tissue injury, and that time was assigned as the response latency.

After the measurement of control PWLs on both paws, the rat was placed in an induction box with halothane being delivered. With the loss of spontaneous movement, the anesthesia was maintained with 2% halothane via face mask. The plantar surface of the hind paw was then placed on a hot plate with a surface temperature of 52.0 +/− 1 [degree sign]C with a 10-g sandpouch placed on the dorsal portion of the paw to maintain equal pressure to the heel area of the paw on the hot plate. The exposure time was set at 45 s, as our previous data suggest that 45 s was sufficient to produce thermal hyperalgesia without producing blisters in the subsequent 24 h. [9]The injury site was examined at 24 h to ascertain that no blisters had developed.

Loperamide (0.5%, 1.7%, 5.0%; 100 mg) or vehicle cream was applied to the paw during halothane-induced anesthesia. The formulation was rubbed lightly into the skin with an index finger for 1 min, and excess formulation left on the surface of the skin. Anesthesia was maintained for 10 min. The concentration of halothane was adjusted to keep the rat immobile during the procedure (1-2%). Spontaneous recovery from anesthesia normally was seen within 5-10 min of the termination of the anesthesia. Consequently, by the rat placing his paw, the excess formulation was removed from the paw.

In this experiment, drug or vehicle was applied to one paw. The treated paw is described as a “normal” paw (neither paw of the rat was injured), “injured” paw (drug applied ipsilateral to the thermally injured paw), or “uninjured paw”(drug applied contralateral to the injured paw). In all experiments, control PWL was measured before injury or any treatment. Postinjury/treatment PWLs were assessed every 30 min for 3 h after injury/treatment on both paws.

Effect of Loperamide after Exposure to Thermal Injury-induced Hyperalgesia. After measurement of control PWL, thermal injury was induced as already described. The posttreatment group received loperamide (0.5%, 1.7%, 5.0%) or vehicle 1 min after the injury to the injured paw during continuous anesthesia for 10 min. On recovery from anesthesia, the rat was again placed in an acrylic cubicle. To ensure that the effect of the drug was local and not systemic, in a different set of animals, the highest concentration of the drug was given on the uninjured paw and the effect compared with the treatment on the injured paw. Rats with drug treatment also were assessed regarding behavior changes, measured by the presence of cornea and pinna reflexes and any change in ambulation.

Effect of Loperamide on Nociceptive Responses of the Normal Paw. After measurement of control PWL, the rat was anesthetized with halothane. During anesthesia, loperamide (0.5%, 1.7%, 5.0%) or vehicle were applied to the right paw for 10 min. On recovery from anesthesia, the rat was again placed in the acrylic cubicle for further testing.

Effect of Loperamide before Thermal Injury. To assess the duration of the effect of the drug, loperamide (5.0%) was applied 4, 2, or 1 h before the injury, during anesthesia. Control PWLs were determined before the application of the drug and before the injury. Thermal injury was induced at the proposed time followed by the assessment of PWL every 30 min for 3 h. The effect seen with rats with 1-h pretreatment with vehicle was compared with the effect of pretreatment with drug.

Reversal of the Antihyperalgesic Effect of Loperamide with Naloxone. To define the mechanism of the effect of the drug, naloxone was given intraperitoneally to reverse the effect of peripherally applied loperamide (5.0%) posttreatment. The group pretreated with naloxone received intraperitoneal injection of 1 mg/kg naloxone 5 min before the application of the drug. Three minutes after intraperitoneally administered naloxone, thermal injury was induced during halothane-induced anesthesia followed by application of loperamide (5.0%). The group posttreated with naloxone, after thermal injury and treatment, received naloxone intraperitoneally (1 mg/kg), 20 min after injury. The effect of naloxone itself was tested also. Naloxone (1 mg/kg) was given intraperitoneally, followed by the induction of thermal injury, and the vehicle treatment and the degree of thermal hyperalgesia were compared with the nonnaloxone-treated, vehicle-applied group.

Reversal of the Antinociceptive Effect of Loperamide with Naloxone. After measurement of the control PWL, 1 or 5 mg/kg naloxone was given intraperitoneally, followed 5 min later by the application of loperamide (5.0%), which was applied during anesthesia.

Tolerance to Loperamide. To ensure that the duration of the pretreatment drug effect determined earlier was not resulting from formation of tolerance, the drug was applied 4 h before and 1 min after the injury. After the baseline PWL measurement, loperamide (5.0%) was applied, during anesthesia, as a pretreatment. Four hours later, thermal injury was induced, followed again by the treatment with loperamide (5.0%).

Cross-tolerance with Systemic Morphine. To determine whether there was any cross-tolerance with systemic morphine, interaction between the effect of systemic morphine and loperamide was examined. Rats received subcutaneous implantation of one (or two) morphine pellet(s) on day 1, and two morphine pellets on day 3. The number of pellets given on the first day was decreased from two to one, because of the high incidence of death in rats given two pellets. The morphine pellet is a slow-release formulation containing 75 mg morphine. [10]Subcutaneous implantation of the pellets was performed during halothane-induced anesthesia. A small incision of the skin was made on the stomach, and the pellets were implanted in the subcutaneous pocket. To demonstrate the magnitude of tolerance developed in rats receiving one and two morphine pellets on days 1 and 3, respectively, 5 mg/kg morphine was given intraperitoneally on day 5, and thermal nociceptive responses were measured at 15 and 30 min and every 30 min thereafter for 2 h after injection. The effect was compared with another group of sham-treated rats. As this treatment showed a significant tolerance to systemic morphine, a separate group of rats given the same treatment as just described was tested for cross-tolerance with loperamide. On day 5, thermal injury was induced, and either the vehicle or loperamide (1.7%, 5.0%) was applied after injury. Paw withdrawal latencies were assessed as described for normal rats.

Pharmacokinetic Study with¹⁴C-labeled Loperamide. After induction of thermal injury,¹⁴C-labeled loperamide (30 mg) was applied to a restricted circular area of the injured skin with an approximate diameter of 6 mm, during continued anesthesia with halothane in oxygen-enriched air. After 10 min of exposure to the drug, anesthesia was terminated, except for the 15- and 30-min study. At either 15 min after injury or on spontaneous recovery from anesthesia, excess drug was wiped off lightly from the skin with a dry cloth, as would have resulted with the ambulation of the rat. This procedure was necessary to prevent contamination. The rat was then placed in a cage for postanesthesia recovery and was allowed to move freely in the cage. At 1, 2, 4, 8, and 24 h after injury, anesthesia was re-induced. A blood sample (1 ml) was taken by intracardiac puncture, after which intracardiac high-dose barbiturate was given to kill the animal. Skin biopsy specimens of the paw were then taken from the treated area of the injured paw and from the same area on the contralateral paw with a device that permitted cutting out an area of the skin with an approximate diameter of 4 mm. The skin sample, which included the skin and the subcutaneous tissue, was then weighed. For the 15- and 30-min study, anesthesia was delivered continuously during the entire procedure, and the rat was never awakened. The skin samples were placed in scintillation vials with 5 ml 95% ethyl alcohol overnight. The next day, liquid scintillation cocktail (EcoLite(+); ICN Biomedicals, Inc., CA) was added to the vials, and the samples were placed in a liquid scintillation analyzer (TRI-CARB, 1900CA; Packard, CA). Serum was obtained from the blood samples and, with the addition of liquid scintillation cocktail, was also counted.

Results are presented as mean +/− SEM as either the raw latency in seconds or as a net efficacy, represented as area under the curve (AUC). Raw PWLs were analyzed by two-way repeated analysis of variance (one between [loperamide concentration] and one within [time] factor) followed by Bonferroni/Dunn's post hoc correction for multiple comparisons. To compute AUC, the algesic index (AI) was first calculated as the percent change from baseline values; AI =[(Posttreatment result - Baseline)/Baseline] x 100. The AUC was then calculated using the trapezoidal rule over the entire time course (Baseline [0]- 180 min). The AUC values were compared with the use of one-way analysis of variance, followed by Bonferroni/Dunn's post hoc correction for multiple comparisons. Differences that achieved significance values corresponding to a probability level of < 0.05 were considered to be significant.

No blister formation was observed in any of the rats; however, 37% of the rats showed a remnant of erythema 24 h after the injury. Contralateral PWLs, despite any treatment or injury, showed no significant change from baseline values and are not included in the results.

Thermal injury induced a prominent decrease in the PWL of the injured paw from its baseline values (thermal hyperalgesia), which was blocked by loperamide posttreatment (F = 2.00, P < 0.02). Control PWL of the vehicle-treated group, in the absence of a thermal injury, was 10.2 +/− 0.4 s (n = 5). After injury and treatment with vehicle, significant thermal hyperalgesia was seen, which peaked at 30 min after injury (PWL 30 min, 5.6 +/− 0.4 s;Figure 1A). As shown in Figure 1A, posttreatment with loperamide blocked the occurrence of thermal hyperalgesia in a dose-dependent manner. Application of 1.7% loperamide blocked the significant decrease in PWL, otherwise seen in vehicle- and 0.5% loperamide-treated rats. The PWL 30 min with 5.0% loperamide treatment was significantly increased compared with that of vehicle and 0.5% loperamide treatment. This antihyperalgesic effect was dose-dependent as seen with the AUC calculation (F = 8.145, P < 0.002;Figure 1B). Treatment of the uninjured paw with 5.0% loperamide after the injury had no effect on the thermal hyperalgesia observed in the injured paw. The AUC values of the injured paw (n = 5) were as follows: 5.0% loperamide on uninjured paw, -2,656 +/− 1,016 versus 5.0% loperamide on injured paw, 2,961 +/− 1,255 (P < 0.05), versus vehicle on injured paw, -3,376 +/− 963 (P > 0.05).

Peripheral application of 1.7% and 5.0% loperamide, but not 0.5%, on a normal paw yielded increases in the PWLs of the treated paw from baseline values (F = 2.20, P < 0.008), significant at 30 min after application (Figure 2A). Analysis of the AUC reveals a significant dose-dependent antinociceptive effect with a plateau effect at 5%(F = 5.25, P < 0.02;Figure 2B).

Pretreatment with 5.0% loperamide at 4, 2, and 1 h before injury, and 1-h pretreatment with vehicle, all yielded a slight increase in the PWL immediately before the injury. The increase was not significant with any treatment, however. One-hour vehicle and 4-h loperamide pretreatment showed significant thermal injury-induced thermal hyperalgesia. One- and 2-h pretreatment with loperamide significantly prevented the development of thermal hyperalgesia (F = 9.46, P < 0.001;Figure 3).

No rats with topical drug treatment displayed any behavior changes, as measured by the loss of cornea or pinna reflexes or change of ambulation, indicating no systemic effect of the drug.

The antihyperalgesic effect of 5.0% loperamide on thermal injury - induced hyperalgesia was reversed by systemic naloxone (1 mg/kg given intraperitoneally) when given before the treatment with loperamide (F = 2.67, P < 0.006;Figure 4A), showing significant thermal hyperalgesia at 30 and 60 min after injury. When naloxone was given intraperitoneally after the treatment with loperamide, however, no significant reversal of the antihyperalgesic effect was observed. When analyzed over the entire 180 min (AUC calculation), pre- and posttreatment with 1 mg/kg naloxone given intraperitoneally showed significant reversal of the antihyperalgesic effect of 5.0% loperamide (F = 7.43, P < 0.008;Figure 4B). The rank order of potency of reversal of this antihyperalgesic effect with naloxone was: pretreatment > posttreatment. Naloxone (1 mg/kg) pretreatment with the vehicle treatment showed no significant difference from the thermal hyperalgesia observed in nonnaloxone, vehicle-treated rats (AUC [n = 5]:-4,795 +/− 719 for intraperitoneal naloxone with vehicle treatment vs. -3,376 +/− 963 with vehicle treatment alone [P > 0.2, t test].

Although the antihyperalgesic effect of 5.0% loperamide was significantly reversed by 1 mg/kg naloxone given intraperitoneally as shown earlier, neither 1 nor 5 mg/kg pretreatment with intraperitoneal naloxone reversed the antinociceptive effect of 5.0% loperamide on normal paw (PWL: F = 1.33, P > 0.2; AUC [n = 5]: 4,053 +/− 845 [5% loperamide], 2,395 +/− 838 [1 mg/kg naloxone with 5.0% loperamide], 3,555 +/− 1,405 [5 mg/kg naloxone with 5.0% loperamide]: F = 0.64, P > 0.5).

Application of the 5.0% loperamide 4 h before the injury had no effect on the antihyperalgesic effect of 5.0% loperamide treatment after injury. The thermal hyperalgesia was significantly reversed (AUC: 1,155 +/− 1,312 pre- and posttreatment vs. -4,181 +/− 887 -4-h pretreatment alone; P < 0.01, t test), showing no detectable tolerance on the antihyperalgesic effect of the drug with 4 h of pretreatment.

Control PWLs of the first set of naive rats were 10.05 +/− 0.23 (n = 11). After subcutaneous implantation of one and two pellets on days 1 and 3, respectively, PWLs on day 5 were 11.69 +/− 0.34 (n = 6), which was slightly increased from day 1 and also from sham-treated rats (9.81 +/− 0.33 s [n = 5], P < 0.05). No apparent behavioral disturbances were seen, however, as determined by the placing and stepping, righting, and corneal and pinna reflexes. Intraperitoneally administered morphine (5 mg/kg) on day 5 produced significant analgesia in sham-treated rats (AUC, 3,651 +/− 392). In morphine pellet-treated rats, however, no analgesia was observed (AUC, 118 +/− 445, P < 0.05, t test). These results indicate development of significant tolerance to systemic morphine in rats treated with morphine pellets. In a second group of rats treated with morphine pellets, baseline PWL on day 5 was 11.50 +/− 0.22 s (n = 21). Thermal injury induced significant thermal hyperalgesia, which was blocked by treatment with loperamide (F = 3.34, P < 0.001;Figure 5A). Treatment with vehicle induced thermal hyperalgesia in the morphine-tolerant rats to a similar degree as in normal rats (Figure 5B). Rats with thermal injury followed by treatment with 1.7% loperamide, unlike normal rats, also developed significant thermal hyperalgesia. Posttreatment with 5% loperamide, however, blocked the development of thermal hyperalgesia. Comparison of the dose-response curves in a range of 1.7-5.0% loperamide revealed a significant treatment effect between morphine-tolerant rats and normal rats (F = 12.60, P < 0.003;Figure 5B), suggesting a rightward shift of the dose-response curve.

Application of 30 mg¹⁴C-labeled loperamide, with a specific activity of 12,000 DPM/mg, results in 360,000 DPM applied to an injured paw. The average weight of the skin taken from the paw was 7.2 mg (range, 5.0-9.0 mg). As shown in Figure 6, radioactivity in the skin showed a significant increase after the application of the drug with its peak concentration at 30 min after application, reflecting [tilde operator]0.4% of the applied radioactivity. The decrease in radioactivity revealed an immediate rapid component with a halflife of 2.3 h. Residual radioactivity was noted at 24 h, suggesting a slow component with a halflife of [tilde operator]27 h. No detectable increase in radioactivity was observed in the serum samples or the contralateral paw at any time point (data not shown).

In the current study, loperamide cream applied after injury to the injured paw showed a dose-dependent antihyperalgesic effect. Thermal injury-induced thermal hyperalgesia is a model of hyperalgesia that produces significant and constant thermal hyperalgesia with a stimulus very similar to our clinical situation. This model was chosen for the current study because of the negligible tissue damage, hence less vascular absorption of the drug from the applied site. Technically, the exposure time of the paw to the drug was restricted to 10 min. The resulting effect of this exposure was only observed when the drug was applied ipsilaterally to the injured paw, confirming its local action. Similar antihyperalgesic action results have been observed in the inflamed knee joint of the rat. In this model, the somatosympathetic response evoked by compression of the inflamed knee was blocked in a naloxone-reversible fashion by intraarticular loperamide. [13]a

The fact that pretreatment with naloxone significantly antagonized the antihyperalgesic effect, especially 30 min after injury, when the drug effect is at its maximum, allows us to conclude that the main antihyperalgesic effect of loperamide is opioid receptor-specific. Although pre- and posttreatment with naloxone served to significantly antagonize the antihyperalgesic effect of loperamide, the pretreatment action appeared to be more efficacious. This observation is consistent with the drug with high affinity to the receptor and to the lipid membranes in the receptor environment. This was also observed with other opioids with high affinity to receptors resulting in nonnaloxone-reversible agonist action. Thus, the alkylating agent, [small alpha, Greek]-chlornaltrexamine ([small alpha, Greek]-CNA) has agonist action in the guinea pig ileum, which is prevented by pretreatment but not reversed by posttreatment with naloxone. [14] 

As the thermal injury-induced thermal hyperalgesia resolves by itself within 3 h, the duration of the drug effect could not be determined by the posttreatment experiment. Based on the pretreatment data, the behavioral effects of the topical agent were estimated to persist between 2 and 4 h, which was unexpectedly short. It is important, however, that this duration was close to the calculated elimination halflife of 2.3 h of the drug in the inflamed skin. These two findings suggest that the local concentration of the drug necessary in the skin to induce an antihyperalgesic action was in the range of 200-370 ng/mg tissue. One possibility accounting for the shorter duration of drug action is the development of tolerance. However, our results reveal that even after 4 h of pretreatment, the antihyperalgesic effect of loperamide applied after injury was effective. Whether longer exposure of the agent leads to tolerance is not known.

With few exceptions, [15]the study of peripheral opioids has revealed their effects to be limited to a hyperalgesic state secondary to injury and inflammation. [2]Our study reveals that there is a modest but significant antinociceptive effect of loperamide applied to the normal paw. This effect was not reversible with naloxone, suggesting a possible local anesthetic-like action. A local anesthetic effect of opioids has been demonstrated in many of the piperidine derivatives, such as meperidine. [16]Loperamide, having structural similarity to meperidine, also may possess such properties.

For the lack of higher-concentration drug, we could not perform a full dose-response analysis with morphine-tolerant rats. Our current data, however, suggest a rightward shift of the dose-response curve in morphine-tolerant rats compared with that of normal rats, which reveals partial cross-tolerance with systemic morphine. Given the complete lack of systemic morphine effect obtained by the treatment with morphine pellets, the partial blockade of peripheral opioid action was unexpected. Several issues may be considered. First, the action of loperamide may not result only from the occupation of specific peripheral [micro sign] opioid receptors. As described earlier, loperamide is likely to have some local anesthetic-like action. The dose-response curve of the antinociceptive effect of loperamide plateaued at 5%, however. Had this effect been the main effect of loperamide, the dose-response curve of the antihyperalgesic effect of loperamide should also show a plateauing effect. Because this was not the result, the local anesthetic effect of loperamide is at least not the main action for the antihyperalgesic effect. Loperamide also is known to possess calcium channel antagonist action at antidiarrheal doses. [17]Calcium channel antagonist applied to the site of nerve injury is reported to suppress neuropathic pain in rats. [18]Although the naloxone reversibility of the effect demonstrated in the current study reveals that the main action of the antihyperalgesic effect of loperamide is opioid receptor-specific, these components also may play some role in its effect, which results in partial cross-tolerance with systemic morphine. Second, the peripheral site may reflect a unique receptor coupling, which lacks complete tolerance formation. Previous work, however, has shown that acute tolerance to repetitive peripheral opioid application with prostaglandin E^2-inducedhyperalgesia does occur. [19]In contrast, the presence of inflammation may represent a biochemistry that impedes development of tolerance. A lack of tolerance to peripheral opioid analgesia in the presence of inflammation has been reported. [20]Enhancement of axonal transport of opioid receptors in nerves and inflamed tissue [21,22]and the opioid-binding sites demonstrated on immune cells [23]may contribute to the lack of tolerance. This applies only during the condition of chronic inflammation, however. Alternatively, the degree of downregulation of the opioid receptor activity with the treatment used in this study may have been insufficient to show significant cross-tolerance to the highest concentration of loperamide used. This might be especially true with opioids with high intrinsic activities. [24]Therefore, in previous studies, animals rendered tolerant to intrathecal morphine showed a smaller rightward shift in the intrathecal dose-effect curve for sufentanil than for morphine. [25] 

Several explanations have been tendered for the peripheral antihyperalgesic action of opioids in the face of inflammation. First, primary afferents, which are normally silent, develop spontaneous activity during inflammation. Repetitive exposure of the skin to ultraviolet light produces this spontaneous activity of the primary afferent nerves. [26]In this state, local [micro sign] and [small kappa, Greek] but not [small delta, Greek] agonists have been reported to depress the spontaneous activity. This suggests a direct reduction of terminal excitability by locally applied opioids. Recently, Gold and Levine [27]have shown that during prostaglandin E^2-inducedhyperalgesia, tetrodotoxin-resistant voltage-gated Na⁺current (TTX-R I^Na) decreases its threshold and increases the rate and magnitude of activation. Peripherally applied [micro sign] opioid agonist was shown to inhibit this prostaglandin E^2-inducedmodulation of TTX-R I^Na. Inhibition of voltage-gated Ca²⁺currents in sensory neurons, [28]the well-known mechanism of opioid analgesia, and this inhibition of hyperalgesic agent-induced modulation of TTX-R I^Naare both likely to be involved in the mechanism of central and peripheral opioid antinociception. Alternatively, some have suggested that local injury may open the blood-nerve barrier, allowing easy access of drug to nerves. [29]This latter explanation appears unlikely, as lipid-soluble drugs such as loperamide would have no barrier even without inflammation.

We have shown the characteristics of a new peripherally acting [micro sign] opioid agonist, loperamide. It has a significant antihyperalgesic action in a hyperalgesic state induced by injury, without systemic side effects.