Dr B. Coffin, Service d'Hépato-Gastroentérologie, Hôpital Louis Mourier, 178 rue des Renouillers, 92701 Colombes Cedex, France. E-mail: email@example.com
Background : Tegaserod reduces the symptoms associated with irritable bowel syndrome, and anti-nociceptive effects have been demonstrated in animals. Its effect on the rectal sensitivity in humans has not been delineated clearly.
Aim : To evaluate the action of tegaserod on rectal sensitivity in response to distension by means of a reflexological technique based on electrophysiological recordings of the RIII nociceptive reflex.
Methods : A randomized, double-blind, placebo-controlled study, performed in 20 healthy women, quantified the effects of slow or rapid rectal distensions on the RIII reflex at baseline and on day 8 following treatment with either placebo or tegaserod (6 mg b.d.).
Results : At baseline, slow distensions performed up to the pain threshold induced gradual inhibitions of the RIII reflex. On day 8, these inhibitory effects were significantly reduced in the tegaserod group, but not in the placebo group (P = 0.0001). The effects of rapid distensions were not significantly modified by tegaserod or placebo. The intensity of subjective pain perception and rectal compliance were not altered by either treatment.
Conclusion : These results suggest that tegaserod reduces the sensitivity to rectal distension in healthy subjects and interacts with the processing of sensory visceral information.
Irritable bowel syndrome is one of the most frequent functional gastrointestinal disorders. It is characterized by recurrent abdominal pain and discomfort, bloating and altered bowel function.1 In these patients, several pathophysiological hypotheses have been suggested.2 Visceral hypersensitivity in response to distension has been seen, and may be a biological marker of irritable bowel syndrome.3 This new concept has highlighted the need for standardized, reproducible methods of evaluation of the visceral sensitivity in humans. We have developed a reflexological technique, based on the recording of a somatic, nociceptive, cutaneo-muscular flexion reflex (i.e. the RIII reflex), that might provide an interesting tool to allow the objective assessment of somato-visceral interactions and visceral sensitivity in humans. The RIII reflex is a polysynaptic reflex elicited by electrical stimulation of a cutaneous sensory nerve and recorded from a flexor muscle on the ipsilateral limb.4,5 The threshold and amplitude of this reflex response are closely related to those of concomitant painful cutaneous sensations evoked by electrical stimulation. We have already shown that, in healthy subjects, graded gastric distensions induce inhibitions of the RIII reflex response that can be correlated with both the volume of the distension and the visceral sensation.6 In subsequent studies,7 we have observed that rectal distensions induce more complex effects, depending on the site of recording of the reflex (i.e. upper or lower limb) and the mode of distension (i.e. rapid or slow ramp).8 Rapid rectal distensions enhance the RIII reflex response recorded from the lower limb, but biphasic effects are observed at the highest distension level (i.e. enhancement followed by inhibition). In contrast, slow ramp distensions induce a gradual inhibition of the reflex that can be correlated with both the volume of distension and visceral sensations. We propose that reflex inhibitions are related to the activation of spinal and/or supra-spinal pain modulation systems. One plausible explanation for the facilitatory effects is that they are caused by the convergence of rectal and reflex afferents at the same levels of the spinal cord.9 In any case, the differential effects of rapid and slow ramp rectal distensions on the RIII reflex suggest the activation of two functionally distinct populations of mechano-receptors by these two modes of distension, i.e. mucosal and serosal mechano-receptors.8,10 Thus, these methods could lead to a more objective evaluation of the respective contributions of superficial and deeper receptors in both visceral sensory disorders and pharmacological studies.
Several studies have shown that serotonin (5-HT) has a pivotal role in the control of gastrointestinal motility and visceral sensations.11 The diverse effects of 5-HT result from the stimulation of different receptor subtypes. 5-HT4 receptors are involved in the regulation of motility and pain sensation via visceral afferents. Tegaserod is a new compound, designed as a selective 5-HT4 partial agonist,12 which has been shown to stimulate motility throughout the gastrointestinal tract.13–17 In established animal models, it has been demonstrated that tegaserod inhibits pseudo-affective responses,18 and dose dependently inhibits the firing of rectal afferents after rectal distension.19 The effect of tegaserod on the rectal sensitivity in humans has not been reported to date. Recent phase III clinical studies in female patients with constipation-predominant irritable bowel syndrome have demonstrated the clinical efficacy of tegaserod, as assessed by the subject's global assessment of relief; it also decreases abdominal pain and discomfort.12,20,21
The purpose of this randomized, placebo-controlled study was to evaluate the action of tegaserod on the rectal sensitivity in healthy female subjects. Using a similar electrophysiological approach to that of our previous studies, we compared the effects of rapid and slow ramp rectal distensions on the RIII reflex, before and after 7 days of administration of oral tegaserod or placebo.
Subjects and methods
Twenty-one healthy female subjects gave written informed consent to the protocol, which was approved by the Ethics Committee of Saint-Louis Hospital (Paris, France). All were healthy on physical examination, had no gastrointestinal symptoms, no previous abdominal surgery (except for appendectomy) and were not taking any medication during the study except for oral contraception.
Nociceptive flexion reflex (RIII reflex)
The RIII reflex was elicited and recorded from the lower limb, according to a previously described and validated technique6–8(Figure 1). Briefly, the subjects were placed in the lateral decubitus position and the RIII reflex was elicited and recorded by an entirely computerized system (Physio Labo System, Notocord, Igny, France). The sural nerve was electrically stimulated at a frequency of 0.17 Hz (10 stimulations/min) via a pair of surface electrodes, placed 2 cm apart, on the degreased skin overlying the nerve within its retromalleolar path. Each electrical stimulation consisted of a train of five constant-current pulses of 1-ms duration. Electromyographic responses were recorded from the ipsilateral biceps femoris via a pair of surface electrodes, placed 2 cm apart, on the degreased skin over the muscle. The RIII response was identified as a multi-phasic signal appearing between 90 and 180 ms after each stimulation. After amplification, each reflex response was digitized, full-wave rectified and integrated. The size of this integrated surface area was used to quantify the RIII response. The RIII reflex threshold was defined as the average minimal current that elicited the reflex response. Before each experiment, this threshold was determined by four successive sequences of increasing and decreasing electrical stimuli. The intensity of electrical stimulation of the sural nerve was then adjusted to 20% above the threshold and kept constant during the control, distension and post-distension periods of each experimental sequence.
An oversized spherical polyvinyl bag (10 cm in diameter; infinite compliance until a maximal volume of 600 mL) was mounted on the tip of a double-lumen polyvinyl tube (12 Fr), folded tightly, lubricated and inserted into the rectum. The distal attachment site was 4 cm from the anal verge. The tube was secured in position with tape. The proximal opening of the tube was linked to an electronic barostat (INRA, Toulouse, France) that allowed controlled inflation and deflation of the balloon with air, and continuous monitoring and recording of the volume and pressure inside the balloon. When in place, the balloon was unfolded by slowly injecting air under controlled pressure (< 20 mmHg) and then completely deflated. After a 20-min period of rest, the barostat was used to inflate the balloon in one of two ways: a series of rapid (900 mL/min) inflations with a constant pressure plateau between each inflation (rapid phasic distension), or a slow and continuous inflation at a constant rate of 40 mL/min (slow ramp distension).
After verification of inclusion and exclusion criteria, experiments were performed at baseline after a 12-h fast. The experimental session began with the determination of the RIII reflex threshold, and thereafter slow ramp and rapid phasic distensions were performed in a randomized order (Figure 2). Four rapid distensions were performed (to 10, 20, 30 and 40 mmHg) in a randomized order. Each distension was maintained for 3 min, and 10 min then elapsed before the next distension to avoid sensitization phenomena. The RIII reflex responses were measured during the 3 min before distension (control period), during the 3-min distension period and during the 3 min after distension (post-distension period).
Slow ramp distensions were performed up to the pain threshold or to the maximal volume of the balloon (600 mL), whichever occurred first. The RIII reflex responses were measured during the 3 min before distension (control period), during the continuous rectal distension period and during a 4-min post-distension period. The sensations elicited by rectal distension were scored as described below. At the end of the experimental session, the volunteers were randomized to receive either tegaserod, 6 mg b.d., or placebo for the following 7 days. On day 8 post-treatment, the last tablet of tegaserod or placebo was ingested 2 h before the second experimental session, which was performed as described for day 1.
Perception of rectal distension
Before the experiments, participants were informed of the visceral sensations they might experience. The sensation elicited by the rectal distension was graded in the range 0–6 using a verbal questionnaire: 0, no perception; 1, initial perception; 2, sensation of gas; 3, sensation of stool; 4, urge to defecate or onset of discomfort; 5, moderate pain; 6, intense or unbearable pain. In the case of rapid distension, subjects reported their sensations at the end of each distension period. In the case of slow ramp distension, subjects reported their sensations at fixed intervals (every 50 mL of distension). Whenever an intense or unbearable pain (score 6) was experienced during any level of distension, the experiment was immediately suspended.
For each distension, the RIII responses were averaged at 1-min intervals, and those on day 8 were expressed as a percentage of the mean value obtained during the control period. During slow ramp distensions, when discomfort (score ≥ 5) was reported before 600 mL, the RIII value recorded at the maximal tolerated volume was reported for subsequent volumes in order to standardize the data. The mean volume was calculated for each rapid distension step (10, 20, 30 or 40 mmHg). Pressure was recorded at fixed intervals (every 50 mL of distension) during slow ramp distension. The safety analyses were performed for all randomized participants who received at least one dose of study medication.
Results were expressed as the mean ± standard error. Intra- and inter-group changes from pre-treatment (baseline) to post-treatment (day 8) were tested by means of analysis of variance (anova) or Wilcoxon rank sum test. A significance level of 0.05 (two-sided) was applied in all cases. Wilcoxon rank sum and Fisher's exact tests were used to assess the homogeneity of the treatment groups. The number of subjects with adverse events was summarized by body system and by individual adverse event. Severity and relatedness were listed. An interim analysis was not performed.
Twenty-one healthy female subjects (mean age, 27.5 ± 2.7 years; range, 19–46 years) were included in the study. One volunteer was excluded before the second experimental session due to concomitant chickenpox, and was not considered for analysis. Demographic data were not different between the volunteers receiving tegaserod or placebo.
The sural nerve stimulation at 1.2 times the threshold elicited a fairly stable RIII response during the pre-distension period and evoked a moderate sensation of pain of the pinprick type. This sensation fluctuated minimally and was well tolerated throughout the experiments. The mean threshold for evoking the RIII reflex was not significantly different between the tegaserod and placebo groups at baseline (8.4 ± 0.8 mA vs. 7.0 ± 0.4 mA, respectively), and was not altered by either treatment on day 8 (8.0 ± 0.7 mA vs. 7.5 ± 0.5 mA, respectively).
Slow ramp distensions
At baseline, the effects of slow ramp rectal distensions on the RIII reflex were similar in the two groups of volunteers. The maximal volumes of distension were 406 ± 32 mL (tegaserod) and 368 ± 29 mL (placebo). Gradual inhibitions of the RIII reflex of similar magnitude during slow ramp distensions were observed in both groups. At the maximal volume of distension, the RIII reflex responses were 50.1 ± 7.0% of control values in the tegaserod group and 55.1 ± 4.0% of control values in the placebo group.
Post-treatment, the maximal volumes of distension were similar to those observed at baseline, and were not significantly different between the tegaserod and placebo groups (349 ± 26 mL and 317 ± 32 mL, respectively). However, as shown in Figure 3, the inhibitory effects on the RIII reflex elicited by slow ramp distensions were significantly reduced in the tegaserod group (72.8 ± 11.4% of control values at the maximal volume of distension), but not in the placebo group (i.e. 54.8 ± 5.7% of control values at the maximal volume of distension). Inter-group comparisons revealed a significant difference between the placebo and tegaserod groups (P < 0.0001, anova).
The intensity of visceral sensations, as recorded by the verbal questionnaire during slow ramp distension, was similar in both groups at baseline and was not significantly changed post-treatment (Figure 4).
Rapid phasic distensions
At baseline, all subjects tolerated the 10, 20 and 30 mmHg levels of phasic distension. The 40 mmHg step distension was not performed in five subjects who reported a painful sensation (i.e. score 5) after the 30 mmHg distension. In another subject, the 40 mmHg distension was interrupted because she reported intense pain (score 6) before the end of the 3 min of distension. As shown in Figure 5, at baseline, the effects of rapid distension on the RIII reflex were similar in the tegaserod and placebo groups. No significant change in the RIII reflex was observed during the 10 mmHg distension. During the 20 and 30 mmHg distension steps, the RIII responses increased during the first minute, with a progressive return to baseline values by the third minute of distension. During the 40 mmHg distension step, an increase in the RIII value was observed during the first minute, followed by a decrease in RIII during the second and third minute, and a progressive return to baseline during the post-distension period.
Post-treatment, the effects of rapid distensions on the RIII reflex response were similar to those observed at baseline, and were not significantly different between the tegaserod and placebo groups (results not shown). Seven subjects were not able to reach or complete the 40 mmHg distension: two in the tegaserod group and five in the placebo group.
The visceral sensations elicited by the rapid distensions were similar in the two groups at baseline, and were not significantly modified by either treatment (Figure 6). The mean distending volumes during 10, 20, 30 and 40 mmHg distension at baseline and post-treatment are reported in Figure 7. No significant differences could be seen between baseline and post-treatment results.
The compliance curves during slow ramp and rapid phasic distensions were similar in both groups at baseline and were not affected by either of the treatments on day 8 (results not shown).
Two subjects in the tegaserod group reported two adverse events, and one subject in the placebo group reported one adverse event. One adverse event in the tegaserod group was a case of chickenpox, not related to treatment. The two other adverse effects, one in each group, were non-systematic gastrointestinal disorders; this adverse event was associated with frequent bowel movements in the tegaserod group.
This study in healthy female volunteers showed that tegaserod significantly decreased the inhibition of the RIII nociceptive flexion reflex induced by slow ramp rectal distensions, whereas the enhancing effects on the reflex induced by rapid distensions remained unaltered. The threshold of the RIII reflex and the elastic properties of the rectum were not altered by tegaserod, suggesting that this drug interacts specifically with the processing of sensory visceral information.
The differential effects of rapid phasic and slow ramp distensions on the RIII reflex, observed in the current study, are similar to those previously reported, resulting from similar distending protocols.7,8 Such differential effects are probably the result of the activation of distinct functional populations of rectal mechano-receptors by these two different modes of distension. This was suggested by the fact that intra-rectal administration of lignocaine (lidocaine), acting as a local anaesthetic agent, blocked the inhibition of the RIII reflex induced by slow ramp distensions, but not the enhancements observed during rapid phasic distension.8 These electrophysiological data are in accordance with the results of studies based on measures of subjective sensory thresholds in volunteers or patients.10 In healthy subjects, discomfort thresholds are higher during rapid phasic distension than during slow ramp distension.3,22 Topical application of lignocaine (lidocaine) decreases the sensation elicited by slow ramp distension, but has no effect on the sensation elicited by rapid phasic distension.3,8,10 On the basis of these results in humans, and those of anatomical and electrophysiological studies in animals,23 it has been hypothesized that sacral afferents connected with mucosal mechano-receptors are preferentially stimulated during slow ramp rectal distensions, whereas splanchnic afferents, whose receptive fields are in the serosa and mesentery and project to the lumbar spinal cord, are preferentially activated during rapid phasic distensions.3,24
Thus, the fact that, in the present study, tegaserod specifically reduced the effects of slow ramp distensions suggests that it interferes preferentially with the transmission and integration of sensory visceral information mediated through superficial (mucosal) receptors. Such an action of tegaserod could take place peripherally or centrally. Studies in cats have demonstrated that tegaserod dose dependently inhibits the firing of rectal afferents during distension.19 On the other hand, the direct inhibition of mucosal mechano-receptors appears unlikely as tegaserod is rapidly absorbed, and its action mimics that of endogenous serotonin released from enterochromaffin cells, stimulating the intrinsic sensory neurones in the intestinal mucosa.12 Alternatively, the reduction of the inhibitions induced by slow ramp distensions might be due to a spinal or supra-spinal action of tegaserod. We propose that such viscero-somatic inhibitory interactions, which have also been described in animals,25 probably involve segmental and/or extra-segmental systems that modulate the spinal transmission of nociceptive signals. There is little information concerning these modulatory mechanisms in humans. However, one such system, named diffuse noxious inhibitory controls (DNIC), initially described in the rat,26 has also been demonstrated in humans.27,28 DNIC are supra-spinally mediated inhibitory controls, triggered selectively by somatic or visceral nociceptive stimuli and acting specifically on spinal nociceptive neurones. Little is known about the pharmacological modulation of DNIC. In animals and humans, low doses of morphine block the activation of DNIC, demonstrating the importance of opioidergic systems.29 Interestingly, studies in rats have demonstrated that the descending serotonergic systems are also involved in DNIC.30,31 The effects of tegaserod might be due to an action on the supra-spinal structures involved in the DNIC circuitry.
Previous studies have suggested that the enhancement of the RIII reflex, observed during rapid phasic distension, may be explained by the convergence on to the same spinal segments of the rectal afferents and those of the RIII reflex, which are integrated in the lumbar and sacral levels of the spinal cord.7,8 Tegaserod did not affect this enhancement, and this suggests that the drug does not interfere with these spinal mechanisms.
One possible mechanism of action for a drug acting on visceral sensations could be the modification of the elastic properties of the digestive tract (relaxation or change in compliance), as suggested for other molecules, such as alosetron.32 The effects of tegaserod on rectal elastic properties have not been reported so far in humans. In the present study, we found that the maximal tolerated volume during slow ramp distension and the mean volume during each rapid phasic distension step were not significantly different between the treatment groups. In addition, rectal compliance curves were not significantly affected by treatment. This lack of an effect on the mechanistic properties of the rectal wall is a strong argument for suggesting that tegaserod acts on sensory afferents.
However, although the present data demonstrate a specific action of tegaserod on the processing of visceral information, it remains difficult to draw conclusions regarding its analgesic properties and clinical efficacy. Indeed, we did not observe any modification of the perceived sensations after tegaserod, irrespective of the type of distension. The lack of an effect of tegaserod on visceral sensations might be due to an insufficient dosage of the drug to demonstrate an effect in healthy subjects; however, we used a dose of 6 mg b.d. which has been shown to have an effect on patients during clinical studies.12,20,21 Alternatively, this may reveal the limits of psychophysical evaluation in pharmacological studies. Gut distension tests are now widely used to test visceral sensation in healthy volunteers, with satisfactory reproducibility.33 However, distension tests with measurements of pain thresholds might be insufficient to show the analgesic action of a drug in healthy volunteers. In pharmacological studies on somatic pain in healthy volunteers, it has been shown that experimental pain thresholds were only slightly modified or not modified at all by analgesics, including opiates. In this respect, the analysis of the responses to supra-threshold stimuli appears to be more sensitive for the demonstration of analgesic properties.34,35 Recordings of the RIII reflex, which reflects the spinal transmission of nociceptive signals, have been widely used to analyse the mechanisms of action of various analgesics in healthy subjects.36,37 Analysis of the effects of visceral stimuli on the RIII reflex might in fact be more sensitive than simple psychophysical measurements for the detection of a potential analgesic action.
We tested the effects of tegaserod only in healthy female volunteers. Phase III studies have demonstrated that tegaserod is active in a sub-group of female patients with constipation-predominant irritable bowel syndrome.12,20,21 This might be explained by the predominance of irritable bowel syndrome in women. The lack of an effect of tegaserod in the male sub-group might be the consequence of a type II error, related to an insufficient number of patients. During previous experiments, we did not observe any gender difference in the effects of rectal distensions on the RIII response reflex, irrespective of the site or type of distension.6–8,38 Other data have also demonstrated that the activation of DNIC is not significantly affected by sex.39 However, a potential effect of tegaserod on visceral sensation in healthy male subjects should be confirmed in further studies.
In irritable bowel syndrome patients, we have recently reported that the effect of slow ramp distension on the RIII reflex response is dramatically different from that in healthy subjects: we observed a significant enhancement in the RIII response in most patients.40 These data suggest that a defect of certain pain modulating systems might be implicated in the genesis of irritable bowel syndrome symptoms. Investigation of the effects of tegaserod in such patients warrants further study using the RIII technique.
In conclusion, tegaserod, a 5-HT4 partial agonist, specifically decreased the inhibition of the RIII nociceptive reflex, induced by slow ramp distension, in healthy female subjects. These results strongly suggest that, in healthy humans, tegaserod has an anti-nociceptive effect; this explains the reduction in abdominal discomfort/pain observed during phase III clinical studies in female patients with constipation-predominant irritable bowel syndrome. Further investigations are required in this patient population using the same methods.
This work was performed at Hôpital Louis Mourier, 178 rue des Renouillers, 92700 Colombes Cedex, France.