Potent analgesics such as the opioids are required to treat moderate to severe acute and chronic pain. The use of opioids is accompanied by serious side effects such as inhibition of gastrointestinal motility 1. As a result, several strategies to prevent opioid bowel dysfunction have been envisaged. In particular, opioid-like compounds with a reduced impact on intestinal function and opioid receptor antagonists with a peripherally restricted site of action have been developed 23. Tramadol is a centrally acting synthetic 4-phenylpiperidine analogue of codeine, which has become available as an analgesic for the treatment of moderate to severe pain 4. Its efficacy has been confirmed in postoperative 56, neuropathic 78 and osteoarthritic pain 910 as well as in patients suffering from chronic pancreatitis 11. The analgesic potency of tramadol appears to depend on the pain state under treatment and has been described as similar 12 or, in the case of pancreatitis, even superior to that of morphine 11. However, there is controversial information as to whether tramadol has an inhibitory action on gut motility in humans, and there is little known as to which mechanisms underlie such an effect 912.

The analgesic effect of tramadol is mediated through two distinct but complementary mechanisms of action. It acts as an opioid agonist with selectivity for the μ-opioid receptor and binds weakly to the κ- and δ-opioid receptor 13. Tramadol is extensively metabolised in the liver, and the O-desmethyl metabolite displays a 200-fold higher affinity for opioid receptors than the parent drug 4. The analgesic and antinociceptive effects of tramadol are only partially antagonised by the opioid antagonist naloxone, which suggests that nonopioid mechanisms are also involved 413. Thus, at the same concentrations at which it binds to opioid receptors, tramadol acts on monoamine systems to inhibit the reuptake of norepinephrine and serotonin 131415. Furthermore, the (+)- and (-)-enantiomers differentially contribute to the analgesic effect of racemic tramadol which is the clinically used form of the drug. The (+)-enantiomer has a higher affinity for the μ-receptor and is a more effective inhibitor of 5-HT reuptake, whereas the (-)-enantiomer is a more effective inhibitor of norepinephrine reuptake and increases norepinephrine release by autoreceptor activation.

In view of these findings the current study pursued three aims. The first goal was to explore whether tramadol is able to inhibit intestinal peristalsis and to characterize this action in quantitative terms. Since tramadol is available in the form of the (+) and (-) enantiomer and the compound is metabolised to the active metabolite O-desmethyltramadol 16, the second aim was to compare (+)- and (-)-tramadol as well as O-desmethyltramadol in their potency and efficacy to depress intestinal peristalsis. The third aim of the study was to explore whether any effect of tramadol on intestinal peristalsis is exclusively due to activation of opioid receptors or whether there is also an opioid receptor-independent component of action. Further experiments were designed to differentiate between an involvement of 5-hydroxytryptamine (5-HT) receptors and adrenoceptors in the opioid receptor-independent inhibition of peristalsis due to tramadol.

Under control conditions regular peristaltic contractions were recorded that stayed constant in all experiments and were not influenced by the addition of vehicle (Tyrode's solution, data not shown). A pilot study showed that racemic tramadol (1–10–30–100 μM) administered to the organ bath in a cumulative manner increased PPT in a concentration-dependent fashion (Figs 1A and 2). Peristalsis was totally abolished by 100 μM racemic tramadol (Fig. 1A) in 5 of 6 segments.

Addition of (+)- and (-)-tramadol to the organ bath led to an increase of PPT and compromised ileal peristaltic contractions in a concentration-related manner. While Tyrode's solution as well as 1 μM (+)-tramadol and (-)-tramadol did not affect peristalsis, PPT was transiently enhanced by 10 μM (+)-tramadol (Fig. 1B, 3A). A more pronounced increase of PPT was caused by 30 μM (+)- and (-)-tramadol (mean Δ PPT 118.8 ± 31.3 Pa and 77.1 ± 20.9 Pa, respectively). An increase in the frequency of peristaltic waves was also observed. Peristalsis completely ceased with a latency of 8.3 ± 4.1 min and a duration of 9.7 ± 0.4 min in two segments exposed to 30 μM (+)-tramadol. Regular peristalsis reoccurred in these two segments, however, with a slightly higher PPT than in the initial control period. Administration of 100 μM (+)- and (-)-tramadol abolished peristalsis after 11.2 ± 4.3 min and 5.3 ± 0.8 min, respectively, in all segments tested (Fig. 1C). In 5 of 6 segments inhibited by 100 μM (-)-tramadol spontaneous ileal contractions recovered 9.2 ± 3.0 min with intermittend periods of disturbed motility. The mean increase of PPT due to 100 μM (+)-tramadol was more prolonged than that due to 100 μM (-)-tramadol (Figs 3A and 3B).

O-Desmethyltramadol had a stronger inhibitory effect on peristalsis than the (+)- and (-)-tramadol enantiomers. The lowest concentrations of 0.1 and 1 μM O-desmethyltramadol did not affect the PPT, but peristalsis was already inhibited by 10 μM O-desmethyltramadol for 18.3 min in one segment. The mean increase of PPT due to 10 μM O-desmethyltramadol during the observation period of 60 min was 91.8 ± 32.9 Pa (Fig. 3C). At 30 μM, O-desmethyltramadol abolished peristalsis transiently in 3 of 6 segments after a latency of 3.6 ± 1 min and on average increased PPT by 239.1 ± 51.9 Pa in all segments tested during the first 20 min post-administration (Fig. 3C). Peristalsis was completely blocked by 100 μM O-desmethyltramadol in all segments, an effect that took place in less than 4 min after addition of the drug. Inhibition of peristalsis manifested itself either as complete absence of any motor activity or as stationary high-frequency oscillations of intraluminal pressure, which did not propel the intraluminal contents.

The inhibitory potency of (-)-tramadol to impair intestinal peristalsis (EC₅₀ = 73 μM) was lower than that of (+)-tramadol (EC₅₀ = 53 μM), and the metabolite O-desmethyltramadol with an EC₅₀ = 13.6 μM was more potent than either enantiomer (Fig. 4).

Pre-treatment with the pan-opioid receptor antagonist naloxone (0.5 μM) prevented the inhibition of intestinal peristalsis caused by tramadol. The reduction of the maximum inhibitory effect of 100 μM (-)-tramadol, 100 μM (+)-tramadol and 100 μM O-desmethyltramadol on PPT was 70 %, 80 % and 90 %, respectively (Fig. 5).

Pre-treatment with a combination of the 5-HT receptor antagonists methysergide plus tropisetron did not affect the increase of PPT caused by the equieffective concentrations of 30 μM (-)-tramadol, 30 μM (+)-tramadol and 10 μM O-desmethyltramadol (Fig. 6). In contrast, the inhibitory effect of 30 μM (+)-tramadol and 30 μM (-)-tramadol was significantly reduced by the combination of the α₁- and α₂- adrenoceptor antagonists prazosin and yohimbine (Fig. 7), while the PPT increase due to 10 μM O-desmethyltramadol remained unchanged.

As summarized in Table 1, the receptor antagonists used here caused only minimal changes of PPT by themselves. PPT slightly decreased due to naloxone and slightly increased due to the 5-HT receptor antagonists, with mean PPT changes lower than 15 Pa.

The (+)- and (-)-enantiomers of tramadol and its major metabolite O-desmethyltramadol impaired intestinal peristalsis in a concentration-dependent manner as deduced from an increase in the peristaltic pressure threshold. The increase in the frequency of the peristaltic waves which was also seen following exposure to submaximal concentrations of tramadol is likely to reflect an impairment of the peristaltic contractions resulting in incomplete emptying of the intestinal segments. O-desmethyltramadol was more potent than the parent compounds (+)- and (-)-tramadol. Overall the inhibitory potency of (+)-tramadol, (-)-tramadol and O-desmethyltramadol was considerably lower than that of the classical opioids such as fentanyl, sufentanil and morphine, which have an EC₅₀ in the nanomolar range as tested in the same experimental in vitro setup 1820. Codeine, the parent molecule of tramadol has an EC₅₀ of 21 μM, which is in the same order of magnitude as O-desmethyltramadol (EC₅₀ = 13.5 μM) 18. The stronger inhibitory effect on gut motility of O-desmethyltramadol is in line with its more potent analgesic capacity compared to that of both enantiomers 4.

The current data are of clinical relevance, given that the ability of tramadol to inhibit gastrointestinal motility in humans is controversial. In one study tramadol was found to slow colonic transit time to a small extent in healthy volunteers, but had no effect on upper gastrointestinal motility 21. Another study found tramadol to have a measurable but smaller inhibitory effect on gastric emptying than morphine and codeine 22. In patients suffering from chronic pancreatitis, orocecal transit time remained unchanged after 5 days of tramadol treatment, but increased with morphine 11. Gastrointestinal motility was also assessed after abdominal hysterectomies in patients receiving 48-h infusions of morphine or tramadol postoperatively in a randomized and double-blind fashion 12. Orocecal and colonic transit times increased after infusion of both morphine and tramadol, but gastric emptying was prolonged only with morphine. In randomized and placebo-controlled studies of patients with diabetic neuropathy and polyneuropathy, who received 210 mg or 200–400 mg tramadol per day with good analgesic benefit, a constipating effect of tramadol was observed 78. Studies in patients suffering from cancer have shown that tramadol has a constipating effect, which, however, is less severe than that of morphine 23. The present results show that tramadol has the potential to inhibit peristalsis in a concentration-dependent manner.

Another aim of this study was to explore the mechanisms underlying the inhibitory action of the tramadol enantiomers and O-desmethyltramadol on intestinal peristalsis. The increase of PPT due to tramadol could to a large extent be prevented by the opioid receptor antagonist naloxone, and the inhibition due to O-desmethyltramadol was in fact attenuated by 90 %. This confirms that most of the inhibitory effect of tramadol on gut motility depends on opioid receptor activation, whereas its analgesic effect could be antagonized with naloxone by 30 % only 24. A major part of the tramadol-induced analgesia is mediated by reuptake inhibition and release of serotonin and norepinephrine 131415. Of these two transmitters, only norepinephrine seems to play an appreciable role in the inhibitory action of (+)- and (-)-tramadol on peristalsis, since the combination of the α₁- and α₂- adrenoceptor antagonists prazosin plus yohimbine diminished the inhibitory effect of both enantiomers to a significant extent. The localization of these adrenoceptors to the various components of the peristaltic nerve-muscle circuitry underlying peristalsis awaits to be addressed. In contrast, peristaltic motor inhibition due to the metabolite O-desmethyltramadol remained unchanged by α₁- and α₂- adrenoceptor antagonism. The 5-HT receptor antagonists methysergide plus tropisetron (antagonists at 5-HT₁, 5-HT₂ and 5-HT₇ as well as 5-HT₃ receptors) did not significantly affect the motor action of (+)-tramadol, (-)-tramadol and O-desmethyltramadol. It was beyond the scope of this study to analyze the possible implication of all 5-HT receptor subtypes present in the gastrointestinal tract. In agreement with its analgesic action, the inhibition of peristalsis due to O-desmethyltramadol is more pronounced than that due to (+)- and (-)-tramadol and the inhibitory motor action of O-desmethyltramadol appears to be solely mediated by opioid receptor mechanisms.

The present data raise the important question as to how the concentrations of (+)-tramadol, (-)-tramadol and O-desmethyltramadol found to inhibit peristalsis in the in vitro setting compared with the concentrations of tramadol, its enantiomers and metabolite that are analgesic under clinical conditions in vivo. In this respect, the evaluation of drug effects on in vitro organ bath preparations has both advantages and limitations. In vitro, environmental conditions such as temperature, oxygenation, composition of the bath solution and drug concentration can be kept constant in a controlled manner. However, protein binding, distribution, metabolism and excretion of drugs cannot be modelled in vitro. As a consequence, the pharmacokinetic and pharmacodynamic conditions of drug action differ considerably between the in vitro and in vivo settings. In this context it is evident that our in vitro preparation has the advantages (i) to reveal effects of low drug concentrations on peristalsis, (ii) to describe these effects in quantitative terms, (iii) to confirm that the observed drug effects arise exclusively from interaction with intestinal structures and mechanisms and (iv) to reveal that the observed drug effects are independent of changes in local, regional and systemic perfusion and other hemodynamic parameters.

In systematic pharmacokinetic studies, plasma concentrations of tramadol, its enantiomers and O-desmethyltramadol have been found to show considerable interindividual variations 25262728. Depending on the study design the minimum effective concentration of tramadol to cause analgesia is in the range of 0.1 – 10 μM 2528. Compared with the concentrations that inhibited intestinal peristalsis in the current study, these figures suggest that clinically relevant plasma concentrations of tramadol, its enantiomers and metabolite will gradually impair intestinal peristalsis in a concentration-dependent manner. However, the concentrations of (+)-tramadol, (-)-tramadol and O-desmethyltramadol (100 μM) that completely block peristalsis are unlikely to be reached under routine clinical conditions.

Racemic tramadol, its (+) and (-) enantiomers as well as their major metabolite O-desmethyltramadol inhibit peristalsis in the guinea pig small intestine at micromolar concentrations. Since the clinically relevant plasma concentrations of tramadol are below those that abolish peristalsis in the guinea pig small bowel, it is tentatively concluded that tramadol is an alternative to classical opiate therapy as it has a more favourable profile of gastrointestinal adverse effects.

MKH and RW contributed equally to this work. RW performed the experiments and drafted the manuscript. MKH and RW designed the studies and analyzed the data. MKH and PH organized all experiments on peristalsis and prepared the manuscript. All authors read and approved the manuscript.