In mice, following i.v. administration, racecadotril was rapidly metabolized to thiorphan; thus, 30 min after the injection only thiorphan was recovered from the kidney whereas the parent compound racecadotril was not detected (de la Baume et al., 1988). In rats 92% of a single-dose (10 mg/kg) of radioactively labeled racecadotril was eliminated within 24 h (Matheson and Noble, 2000).

Racecadotril is rapidly absorbed following oral administration. For example, in a single-dose, placebo-controlled, double-blind cross-over study racecadotril doses of 30, 100, and 300 mg reached C_max within 60 min after oral administration; the latter two doses were associated with significant inhibition of plasma NEP activity seen as early as 30 min after ingestion, and that inhibition exhibited a t_1/2 of 3 h (Lecomte, 2000; Matheson and Noble, 2000). A single-dose study in healthy elderly people reported similar findings (Matheson and Noble, 2000). After ingestion of a single oral dose of 300 mg racecadotril plasma thiorphan levels peaked after 60 min reaching 805–1055 nM; after 240 min plasma levels were still at 92–204 nM (Hinterleitner et al., 1997). In a more recent study, a well validated liquid chromatography/tandem mass spectrometry method has been used to detect thiorphan in human plasma; 20 volunteers received a single oral dose of 200 mg racecadotril, which resulted in a thiorphan C_max of 520 ng/mL, a t_max of 1.35 h and a t_1/2 of 6.14 h (Xu et al., 2007). The pharmacokinetic properties of racecadotril are similar with chronic dosing as observed in a placebo-controlled, double-blind study with 30, 100, and 300 mg racecadotril being given orally thrice daily for 7 days, where pharmacokinetic parameters were similar on day 1 and day 7 and to the values observed in the single-dose studies (Matheson and Noble, 2000), indicating lack of accumulation upon chronic dosing. Concomitant food intake does not modify the bioavailability of racecadotril but peak NEP inhibition is delayed by about 90 min (data on file).

After oral administration racecadotril is rapidly and effectively metabolized to the active metabolite thiorphan which is the predominant species detected in plasma; the occurrence of thiorphan coincides in time with the inhibition of plasma NEP (Hinterleitner et al., 1997; Xu et al., 2007). Thiorphan has a plasma protein binding of approximately 90% (data on file).

An important question for any drug interfering with the endogenous opioid system is whether it crosses the blood–brain-barrier, i.e., whether upon oral racecadotril administration parent compound or active metabolite reaches relevant levels in the brain to cause NEP inhibition. Animal studies suggest that central nervous effects can occur after parenteral administration of racecadotril but not after oral racecadotril or parenteral thiorphan administration (see Studies on Central Nervous System Function); this is the apparent result of the combination of the rapid conversion of absorbed racecadotril to thiorphan and the lack of thiorphan passage through the blood–brain-barrier due to its less lipophilic chemical structure (Figure 1). In line with these animal data it has been reported that i.v. administration of racecadotril causes quantitatively similar NEP inhibition in plasma and cerebrospinal fluid in five healthy volunteers as compared to saline infusion (Spillantini et al., 1986), whereas a single high oral racecadotril dose (20 mg/kg) to two volunteers caused a marked reduction of plasma NEP activity within 30 min but did not affect liquor NEP activity (Lecomte, 2000). In accordance with the proposed lack of central nervous effects of orally administered racecadotril, a placebo-controlled cross-over study in 12 subjects being treated with 300 mg/kg racecadotril for 3 days did not detect any impairment of vigilance (Lecomte, 2000).

The active racecadotril metabolite thiorphan is converted to inactive metabolites, but the pathways mediating this conversion have not been characterized in great detail. While two studies using different methodological approaches have reported that racecadotril inhibits CYP 3A4-mediated formation of metabolites of the cancer chemotherapeutic drug irinotectan with an IC₅₀ of 46 μM in human liver microsomes (Haaz et al., 1998a,b), it should be noted that this concentration is equivalent to about 10,000 times the potency of thiorphan for NEP inhibition. Thus, in concentrations which are achieved by therapeutic doses racecadotril is neither an inhibitor nor an inducer of cytochrome P450 enzymes and also not a substrate of the P-glycoprotein transporter (data on file). Accordingly, to date no interactions with other medicinal products have been identified and specifically concomitant treatment with loperamide or nifuroxazide does not affect the pharmacokinetics of racecadotril (data on file). Moreover, racecadotril does not modify protein binding of active substances strongly bound to proteins such as tolbutamide, warfarin, niflumic acid, digoxin, or phenytoin (data on file). The elimination of the inactive thiorphan metabolites occurs mainly via the renal route (data on file).

Based on the role of morphine and other opiates in pain control, it was an obvious choice to test racecadotril in various pain models. The results have been rather inconsistent across models but much more consistent within models indicating that NEP inhibition selectively targets certain pain pathways. Thus, racecadotril was analgesic in the hot-plate jump test in unspecified mice with i.c.v. administration (Roques et al., 1980) and with i.v. administration in Swiss albino (Lecomte et al., 1986; Costentin et al., 1998), NMRI (Lambert et al., 1993, 1995) and DBA/2J mice but not C57BL/6J mice (Michael-Titus et al., 1989). Interestingly, in the latter study racecadotril increased locomotion in both strains, indicating that the difference in analgesic effect does not reflect a pharmacokinetic strain difference. In NMRI mice i.v. acetyl-thiorphan and thiorphan were also effective in this model, although thiorphan less than the more lipophilic acetyl-thiorphan and racecadotril (Lambert et al., 1993, 1995). I.v. racecadotril was also analgesic in Swiss albino mice in the tail-withdrawal and the phenylbenzoquinone-induced writhing test (Lecomte et al., 1986). In an arthritis-based pain model in rats (vocalization induced by applying pressure to the left hind paw) racecadotril also exhibited analgesic activity (Kayser and Guilbaud, 1983), but this was weaker than that of the mixed peptidase inhibitor kelatorphan in the same model (Kayser et al., 1989). In the same test racecadotril was not analgesic in non-arthritic rats (Kayser and Guilbaud, 1983). In unspecified mice i.c.v. thiorphan was ineffective in the tail removal test but enhanced the analgesic effects of several enkephalines which are NEP substrates but not of others which are not NEP substrates (Roques et al., 1980). Moreover, racecadotril did not exhibit analgesic properties in the hot-plate licking test in Swiss albino (Lecomte et al., 1986) or NMRI mice (Lambert et al., 1993) or in albino mice in the tail immersion test, but enhanced the anti-nociceptive effect of an exogenously administered enkephalin in the latter model (Livingston et al., 1988). In GB1 mice in the acetic acid-induced abdominal constriction assay racecadotril was effective only at subcutaneous doses of 10 mg/kg and higher (Gray et al., 1998).

Upon chronic stimulation the opioid system can exhibit both desensitization and sensitization. Thus, the analgesic response to i.v. racecadotril was blunted after 14 days of i.c.v. treatment with thiorphan in rats (Bousselmame et al., 1991a). However, no cross-desensitization between racecadotril and morphine was observed in mice (Bousselmame et al., 1991b) or rats, and in the latter also no cross-sensitization for locomotive effects was observed (Khallouk-Bousselmame and Costentin, 1994).

It has also been tested whether racecadotril can enhance analgesic effects of other treatments. Thus, racecadotril or thiorphan enhanced naloxone-sensitive analgesia induced by transcranial electrostimulation in rats (Malin et al., 1989) or by nefopam in mice (Gray et al., 1999), and also enhanced analgesic effects of electroacupunture in rats (Zhou et al., 1990). In the acetic acid-induced abdominal constriction pain model in GB1 mice racecadotril enhanced the analgesic effect of morphine and of anti-depressants such as dothiepine and amitriptyline (Gray et al., 1998). On the other hand, racecadotril did not enhance naloxone–sensitive analgesic effects in four different pain models in mice (Michael-Titus and Costentin, 1987), and neither i.v. racecadotril nor i.c.v. thiorphan enhanced analgesic effects of dopamine receptor agonists in mice (Michael-Titus et al., 1990a). Another type of interaction between pain-related pathways was suggested by findings in mice in which nociceptin attenuated the analgesic response to racecadotril (Costentin et al., 1998).

In conclusion, racecadotril has direct analgesic effects and can enhance analgesic effects of some other types of drugs in some but not all pain models, but such studies were largely restricted to parenteral administration. Similar to direct opioid receptor agonists, racecadotril-induced analgesia can undergo desensitization upon long-term exposure but despite both morphine and racecadotril effects involving opioid receptors, they did not exhibit cross-desensitization. As the anti-nociceptive effects of racecadotril are restricted to some model systems, it can be expected that, if at all, racecadotril would be effective only in some forms of pain in patients and only upon parenteral administration; however, a clinical testing of potential analgesic effects of racecadotril in patients has not been reported to our knowledge.

The administration of natural or synthetic opioid receptor agonists elicits a locomotor response in rodents, which is considered to be an index of the activity of mesolimbic dopaminergic neurons. In both mice and rats i.v. racecadotril was reported to enhance locomotion in mice and rats in a naloxone–sensitive manner, and that response was blocked by a dopamine receptor antagonist and enhanced by a dopamine uptake inhibitor (Michael-Titus et al., 1987, 1990b). Similar to the analgesic racecadotril response (see Analgesia), the locomotor racecadotril response also was desensitized following a 14-day i.c.v. treatment with thiorphan (Bousselmame et al., 1991a). In line with the idea that racecadotril can affect dopaminergic transmission in the brain, it was found that i.v. racecadotril modulates dopaminergic transmission in rat olfactory tubercle but not striatum (Dourmap et al., 1990).

Both the opioid and the dopamine system in the brain are prone to addiction. Therefore, it was important to find that racecadotril did not exhibit abuse potential in rats or monkeys in doses up to 50 mg/kg (Knisely et al., 1989). Nevertheless, i.p. racecadotril prevented some but not all withdrawal symptoms in opioid-habituated mice and rats (Livingston et al., 1988; Dzolic et al., 1992). However, racecadotril alone was ineffective in inhibiting naloxone-induced morphine withdrawal symptoms in mice in another study, but a combination of subthreshold doses of racecadotril, CCK-4, and caerulein was effective (Bourin et al., 1999). In a double-blind, double-dummy, clinical proof-of-concept study in 19 heroin-addicted patients comparing 50 mg i.v. racecadotril and 75 μg oral clonidine; racecadotril appeared more effective than clonidine against objective withdrawal symptoms as quantified by the Himmelsbach scale, whereas both treatments were similarly effective against subjective withdrawal symptoms (Hartmann et al., 1991).

Some studies have explored potential metabolic effects of racecadotril. In sheep oral and i.v. administration of racecadotril increased food intake, whereas i.c.v. thiorphan reduced it; as oral racecadotril treatment does not lead to NEP inhibition in the brain and as i.c.v. thiorphan did not mimick the racecadotril effects, this appears to be a peripheral effect (Riviere and Bueno, 1987). Moreover, naltrexone blocked the former but not the latter effect, indicating an involvement of peripheral opioid receptors. A study with i.v. racecadotril in cats reported increases in sham food intake (Bado et al., 1989). In rats i.v. racecadotril caused naloxone-insensitive dose-dependent lowering of blood glucose which was accompanied by increased plasma insulin and C-peptide levels (Wu et al., 2010). While racecadotril did not affect insulin release from isolated pancreatic islets, the muscarinic receptor antagonist atropine blocked and the cholinesterase inhibitor physostigmine enhanced the racecadotril-induced insulin elevation, indicating that they may occur via modulation of parasympathetic nerve activity. In a follow-up study the same investigator group provided evidence that i.v. racecadotril and i.c.v. thiorphan directly inhibit an insulin-degrading enzyme in the brain, and that brain insulin acts via the vagal nerve on plasma glucose (Lee et al., 2011). The clinical findings of these intriguing observations have not been explored in patients to our knowledge.

In a mouse behavioral dispair test, a model system for anti-depressant effects, immobility time was reduced by 10 mg/kg i.v. or 50 mg/kg i.p. in mice (Lecomte et al., 1986). Moreover, it was reported that thiorphan can provide neuroprotection in newborn mice (Medja et al., 2006).

Racecadotril has been tested in various animal models of hypertension, specifically for its ability to lower blood pressure, improve renal function, and to prevent or reverse organ hypertrophy. These models include spontaneously hypertensive rats (Bralet et al., 1990) and its stroke-prone substrain (Stasch et al., 1995), transgenic rats harboring a mouse renin gene (Stasch et al., 1996), hypertension induced by treatment with the immunosuppressant cyclosporine A in rats (Takeda et al., 2000), and the rat deoxycorticosterone acetate-salt model of mineralocorticoid-induced hypertension (Ito et al., 1999). Blood pressure lowering by oral racecadotril or ecadotril was consistently shown (Stasch et al., 1995, 1996; Ito et al., 1999; Takeda et al., 2000). In a first pilot study in 12 hypertensive patients receiving increasing sinorphan doses (25–200 mg bid) for a total of 6 weeks a dose-dependent blood pressure reduction was also observed (Lefrancois et al., 1990). In a subsequent randomized, double-blind clinical pilot study with a cross-over design in 16 hypertensive patients racecadotril was less effective than captopril in lowering blood pressure, but the combination of both drugs was more effective than either monotherapy (Favrat et al., 1995).

In line with the effects in normotensive animals and healthy human volunteers, racecadotril also increased diuresis and/or natriuresis in spontaneously hypertensive rats (Bralet et al., 1990), in transgenic rats (Stasch et al., 1996) and in deoxycorticosterone acetate-salt-treated rats (Ito et al., 1999) whereas numerical increases of diuresis and natriuresis did not yield statistical significance in stroke-prone spontaneously hypertensive rats (Stasch et al., 1995).

A prognostically relevant complication of hypertension is the development of hypertrophy of the heart and other cardiovascular organs. In this regard, chronic treatment with racecadotril reduced heart hypertrophy in stroke-prone spontaneously hypertensive rats (Stasch et al., 1995), in transgenic rats (Stasch et al., 1996) and in deoxycorticosterone acetate-salt-treated rats (Ito et al., 1999). A reduction of renal enlargements was observed less consistently in these studies.

Taken together these studies demonstrate beneficial effects of treatment with racecadotril on blood pressure, renal function, and cardiac hypertrophy in various animal models of hypertension, which is in line with the elevated ANP levels in such animals. While clinical pilot studies have confirmed blood pressure lowering effects in hypertensive patients, these were too modest in comparison to established anti-hypertensive treatments to warrant further clinical investigation of racecadotril in this indication; potential enhancement of blood pressure lowering by other drugs was also deemed insufficient to be of clinical relevance.

Congestive heart failure leads to atrial dilatation which is the most important physiological stimulus for ANP secretion. This enhanced ANP secretion is generally seen as a counter-measure to increase diuresis and natriuresis and thereby lower cardiac after-load. Accordingly, racecadotril has been evaluated in animal models and in patients with heart failure. Animal models of heart failure in which racecadotril or ecadotril have been tested include rats with volume overload due to aortic valve insufficiency (Kimura et al., 1998) or to an atrio-ventricular fistula (Wegner et al., 1996), rats after a myocardial infarction (Duncan et al., 1999), dogs with coronary microembolization (Olivier et al., 2000; Mishima et al., 2002), and dogs with heart failure due to sino-atrial pacing (Solter et al., 2000).

In line with the fluid retention, an activation of the renin–angiotensin system is a hallmark of congestive heart failure. Such activation was mitigated by treatment with racecadotril in animal models (Wegner et al., 1996; Kimura et al., 1998; Duncan et al., 1999). A similarly reduced activity of the renin–angiotensin system was observed in an early and short-term pilot study in heart failure patients (Kahn et al., 1990) but not in a larger chronic study in such patients (Cleland and Swedberg, 1998). Whether indirectly by reducing activity of the renin–angiotensin system or more directly by increasing ANP exposure, diuretic and/or natriuretic effects of racecadotril were observed in the volume overload rats (Wegner et al., 1996), coronary microembolization dogs (Olivier et al., 2000), and in dogs with pacing-induced heart failure (Solter et al., 2000) but not in post-myocardial infarction rats (Duncan et al., 1999). Accordingly, improvements of the cardiac pump function were observed in most of these models (Kimura et al., 1998; Olivier et al., 2000; Mishima et al., 2002), whereas blood pressure did not change (Mishima et al., 2002).

Some of these heart failure models, particularly those involving volume or pressure overload typically lead to cardiac hypertrophy. In line with the other findings it has been found that chronic racecadotril treatment ameliorates such cardiac hypertrophy in the volume overload rat models (Wegner et al., 1996; Kimura et al., 1998). Similarly, racecadotril treatment also reduced cardiac hypertrophy in the dog microembolization model (Mishima et al., 2002). However, in the post-myocardial infarction rat model neither racecadotril nor perindopril alone reduced development of cardiac hypertrophy whereas their combination did (Duncan et al., 1999).

In accordance with the observed increase in plasma ANF and/or urinary cGMP, three studies have explored whether racecadotril or ecadotril may have therapeutic benefit in heart failure patients. In a series of small pilot studies in patients with severe heart failure (left ventricular ejection fraction 20%), 2 days of ecadotril treatment doubled plasma ANP levels despite starting from a markedly elevated baseline (Kahn et al., 1990). This was accompanied by a reduced renin activity and pulmonary capillary wedge pressure. Based on those encouraging findings a clinical, placebo-controlled dose-ranging study (50–400 mg ecadotril twice daily for 6 months) was performed in 259 patients with moderate heart failure (left ventricular ejection fraction of <35%; Cleland and Swedberg, 1998). This confirmed a dose-dependent increase in plasma and urinary cGMP but patients did not show reduced activation of the renin–angiotensin system or, more importantly, clinical improvement. Another study in a similar population of 50 moderate heart failure patients with 10 weeks of treatment with increasing racecadotril doses (up-titration from 50 to 400 mg twice daily) also failed to demonstrate clinical improvement (O’Connor et al., 1999).

Taken together racecadotril has shown promising findings in animal models of congestive heart failure but patient studies have not confirmed a sufficient clinical potential to warrant further development in this indication.

Racecadotril has also been tested in animal models and/or clinical pilot studies for various other indications related to cardiovascular function. Studies in rats have reported that i.v. racecadotril may protect the heart against adrenaline-induced arrhythmia (Lishmanov et al., 2001) or against arrhythmia induced by short ischemia–reperfusion episodes (Naryzhnaia et al., 2001), the latter effect being blocked by a δ-opioid receptor antagonist. In a mouse model of pulmonary hypertension ecadotril was found to have synergistic beneficial effects with sildenafil (Baliga et al., 2008).

In a single-dose pilot study in liver cirrhosis patients racecadotril 30 and 100 mg increased plasma ANP and cGMP and caused a transient diuresis and natriuresis response relative to placebo; the activity of the renin–angiotensin system apparently was not affected in these patients (Dussaule et al., 1991). In a single-dose cross-over pilot study in chronic renal failure patients 100 mg ecadotril inhibited enkephalinase, increased plasma cGMP and natriuresis; aldosterone, glomerular filtration rate, or blood pressure were not affected (Dussaule et al., 1993). Finally, a single dose of 200 mg ecadotril increased plasma endothelin-1, ANP, and cGMP and diuresis and natriuresis in a randomized placebo-controlled study in heart transplant recipients (Piquard et al., 2002).

It is well established that enkephalines have potent anti-secretory properties in the gut but do not affect gut motility (Turvill and Farthing, 1997). In an initial study in rats, i.v. racecadotril inhibited castor oil-induced diarrhea, an effect which was abolished by the opioid receptor antagonist naloxone (Lecomte et al., 1986). Such findings in the castor oil model of diarrhea were confirmed in rats with oral racecadotril and with i.v. thiorphan; they were blocked by subcutaneous but not by i.c.v. naloxone, indicating that the opioid receptors mediating this effects are located peripherally (Marcais-Collado et al., 1987). Racecadotril also reduced castor oil-induced diarrhea in human volunteers in a placebo-controlled study (Baumer et al., 1992).

To explore the underlying mechanism of anti-diarrhea effects of racecadotril, several studies have been performed. Thus, excessive fluid secretion from the gut is a pathophysiological hallmark of diarrhea. Racecadotril inhibited cholera toxin-induced but not basal secretion in canine jejunum (Primi et al., 1999). This was confirmed in a parallel group study in human volunteers, in which cholera toxin was administered by segmental perfusion directly into the proximal jejunum (Hinterleitner et al., 1997). In another study, racecadotril inhibited secretion induced by rotavirus infection in an in vitro model of intestinal secretion, Caco-2 cells (Guarino et al., 2009), a model which may be of value because rotavirus infection is a very frequent cause of childhood diarrhea.

A potential complication of diarrhea treatment is inhibition of intestinal motility as it can lead to secondary constipation and, perhaps even more important, intestinal retention of harmful infectious organisms. In rats oral 40 mg/kg racecadotril was reported not to affect gastro-intestinal transit time, whereas 2 mg/kg loperamide did (Marcais-Collado et al., 1987). Using the same approach in mice, 20 mg/kg i.v. of racecadotril or thiorphan or 0.5 mg/kg oral loperamide also did not significantly affect transit time, whereas 10 mg/kg oral or 0.5 mg/kg i.v. loperamide significantly prolonged it (Marcais-Collado et al., 1987). A potential consequence of effects on gastro-intestinal transit time was explored in newborn piglets, in which a 4-day oral treatment with 20 mg/kg racecadotril twice daily did not significantly affect E. coli content of the proximal jejunum, whereas 1 mg/kg oral loperamide twice daily markedly increased it; accordingly, the E. coli content of the stool was significantly reduced by loperamide but not by racecadotril (Duval-Ilfah et al., 1999). In placebo-controlled studies in human volunteers racecadotril treatment for up to 1 week also did not modify oro-coecal, colonic or overall gastro-intestinal transit times (Baumer et al., 1989; Bergmann et al., 1992).

In conclusion, both racecadotril and direct μ-opioid receptor agonists have effects on the gut which lead to limitation of pathological fluid loss. While the receptor agonists do so primarily by prolonging transit time and hence providing more opportunity for fluid reabsorption, racecadotril does so by inhibiting fluid secretion; the latter may be preferable as it directly targets the primary pathophysiological mechanism underlying acute diarrhea and also reduces the chance of retention of infectious agents in the gut.

I.v. administration of racecadotril was shown to inhibit gastric secretion in cats induced by pentagastrin, histamine or 2-deoxy-d-glucose in a naloxone-sensitive manner, whereas the meal-induced secretion was not affected (Bado et al., 1987). In rats i.v. racecadotril and i.c.v. thiorphan, but not i.v. thiorphan, inhibited gastric acid secretion; this was no longer observed following vagotomy, indicating a central nervous system-mediated effect (Chicau-Chovet et al., 1988). In mice i.p. racecadotril and thiorphan given prior to a fatty meal enhanced gastric emptying in a naloxone-sensitive manner; gastric emptying in response to a non-fat meal was enhanced by low and inhibited by a high thiorphan dose in a naloxone–resistant manner, and racecadotril was without significant effect under these conditions (Liberge et al., 1988).

Effects of racecadotril have also been investigated in the feline gall bladder. In a feline cholecystitis model racecadotril inhibited fluid secretion in a naloxone–sensitive manner, but did not block fluid transport in the normal gall bladder (Jivegard et al., 1989). In that study racecadotril also caused transient gall bladder contraction and increased bile outflow from the liver. In another study in the same species i.v. racecadotril but not i.v. thiorphan caused naloxone–sensitive contraction of the spincter Oddi, indicating a central nervous system-mediated effect (Thune et al., 1992).

In a controlled cross-over study in 10 healthy volunteers, 2.5 mg/kg i.v. racecadotril attenuated relaxation of the lower esophagus sphincter but did not affect contraction (Chaussade et al., 1988). Racecadotril enhanced the propagation of electrical signals in the distal colon in rats in the fasted and fed state in a naloxone–sensitive manner (Benouali et al., 1993). In a study in healthy volunteers racecadotril produced an atropine-resistant promotion of electrical activity in the rectum, which was absent in patients with Hirschsprung’s disease (Grimaud et al., 1989).

The data of all controlled studies with racecadotril in the treatment of acute diarrhea in adults are summarized for efficacy in Table 1 and for adverse events in Figure 2. The original registration of racecadotril as a treatment for acute diarrhea in adults in France in 1992 was based on three placebo-controlled studies. In a double-blind, placebo-controlled, randomized dose-ranging study 49–55 patients per group with acute diarrhea presumed to be due to food poisening or infection-related received 30, 100, or 300 mg racecadotril or placebo three times per day until recovery for up to 10 days (data on file). The primary outcome parameter was time to cure, which was 68.4, 69.6, 65.0, and 72.0 h with 30, 100, and 300 mg racecadotril and placebo, respectively (not significant). Secondary outcome parameters included number of diarrhoic stools in the first 10 h (2.0–2.2 with the three racecadotril doses as compared to 2.7 days with placebo; p = 0.06) and in first 3 days (6.8–7.7 with the three racecatodril doses as compared to 8.6 days with placebo; p = 0.03). The incidence of the adverse events anal burning, painful anal contractions, spontaneous abdominal pain, nausea, vomiting, loss of appetite, asthenia, and insomnia did not differ significantly between groups.

In a double-blind, randomized, placebo-controlled study adult out-patients with acute diarrhea of presumed infectious origin having started less than 5 days before (95 on racecadotril, 98 on placebo) were treated with an initial dose of 200 mg racecadotril followed by an additional 100 mg dose after each unformed bowel movement until recovery or for a maximum of 10 days (Baumer et al., 1992). Resolution of diarrhea occurred significantly faster in a Kaplan–Meier type analysis with racecadotril than with placebo, e.g., on day 4 the cumulative probability of recovery was 75% with racecadotril vs. 37% with placebo. Accordingly, mean duration of treatment was 3.0 ± 0.2 days with active treatment vs. 4.4 ± 0.3 days with placebo. Several secondary endpoints including anal burning, spontaneous abdominal pain, nausea, anorexia, pain on abdominal palpation, and abdominal distension were also significantly improved by racecadotril as compared to placebo. The percentage of patients reporting adverse effects with racecadotril and placebo was 16.8 vs. 18.4%, respectively.

In another double-blind, parallel group, placebo-controlled study 70 adult patients with acute diarrhea of presumed infectious origin were included and randomized to receive 100 mg racecadotril or placebo three times daily until recovery for a maximum of 6 days (Hamza et al., 1999). The primary outcome parameter was mean stool weight, which was significantly smaller in the racecadotril as compared to the placebo group (355 ± 35 vs. 499 ± 46 g, respectively). Secondary outcome parameters for which racecadotril was significantly superior to placebo included number of diarrhoic stools after 1 day of treatment 4.3 ± 0.4 vs. 5.4 ± 0.4, respectively) and percentage of patients passing at least one formed stool on the second day of treatment (15.6 vs. 5.3%, respectively). The incidence of reported adverse events was 3.1% with racecadotril vs. 5.3% with placebo; abdominal distension, not classified as an adverse event in this study, was 5.6% with racecadotril vs. 18.2% with placebo.

Following registration in 1992, one additional double-blind, randomized, placebo-controlled trial was performed in 110 men with cholera, in which 100 mg racecadotril was administered every 4 h until recovery for a maximum of 72 h as an adjunct to standard treatment (Alam et al., 2003). Both treatments did not differ significantly with regard to total stool output, duration of diarrhea or patients with resolution of diarrhea within 72 h. Adverse events noted as per-protocol such as vomiting, reappearance of dehydration, abdominal pain, headache, or anorexia were not different between the treatment groups. Taken together these studies consistently demonstrate efficacy of racecadotril as compared to placebo with similar adverse event incidences with both treatments. However, it should be noted that treatment regimens and outcome parameter varied considerably between studies.

Some studies have been performed with racecadotril in the context of diarrhea associated with cancer chemotherapy. In a pilot study in 15 cancer patients treated with 5-fluoro-uracil were given a daily dose of 300 mg/d racecadotril for 7 days for the treatment of acute diarrhea (Dorval et al., 1995). As compared to the control period, i.e., earlier cycle of chemotherapy, the number of stools per day was reduced in each patient with a statistically significant reduction of mean number from 6.3 to 4.9 and the number of days with liquid stools significantly dropped from 4.7 to 2.4.

Delayed onset diarrhea is a dose-limiting side-effect of the second-line anti-cancer drug irinotecan, which is often used in the treatment of colon cancer. Among patients exhibiting delayed diarrhea upon irinotecan treatment, 4 out of 11 patients responded to 100 mg racecadotril thrice daily in one cohort, whereas 9 of 10 patients responded to a combination of racecadotril and loperamide (p < 0.02 vs. racecadotril alone) (Saliba et al., 1998). In a randomized open-label study 136 patients receiving a total of 714 irinotecan chemotherapy cycles received 300 mg/d racecadotril for 15 days as a prophylactic treatment or no prophylactic treatment but the two groups did not differ significantly in the incidence or severity of delayed diarrhea (Ychou et al., 2000).

In an open, randomized cross-over study the efficacy of 100–300 mg racecadotril thrice daily and 50–150 μg octreotide thrice daily was compared in 13 treatment-resistant diarrhea in AIDS patients (Beaugerie et al., 1996). From a baseline of 7.0 ± 1.2 stools/day racecadotril caused a significant reduction to 4.6 ± 1.1 stools/day, whereas octreotide caused a non-significant reduction to only 5.6 ± 1.2 stools/day. Daily lipid output was increased non-significantly by racecadotril, but was nearly doubled with octreotide.

Based upon the results of racecadotril in the treatment of acute diarrhea in adults, several studies have explored its use in the treatment of acute diarrhea in children. The data of all controlled studies with racecadotril in the treatment of acute diarrhea in children are summarized for efficacy in Table 2 and for adverse events in Figure 3. In the first of such studies 135 boys aged 3–35 months including 73 with a verified rotavirus infection with 1.5 mg/kg racecadotril every 8 h or placebo in a randomized, double-blind study with both treatments being administered on top of oral rehydration solution (Salazar-Lindo et al., 2000). Total stool output in the first 48 h of treatment, the primary study endpoint, was 157 ± 27 g/kg with racecadotril as compared to 331 ± 39 g/kg with placebo (p < 0.001). The duration of diarrhea was significantly shorter with racecadotril (28 h regardless of rotavirus status) than with placebo treatment (72 and 52 h in rotavirus-positive and –negative boys, respectively). Moreover, the amount of required oral rehydration solution was also significantly less with racecadotril treatment. Adverse events were reported in 7 of 68 and 5 of 67 children receiving racecadotril and placebo, respectively; 51 and 52% reported vomiting at at least some point during treatment, and study withdrawal occurred in 9 and 14 patients, respectively.

A second double-blind, placebo-controlled randomized study of very similar design with racecadotril on top of oral rehydration solution was performed in 173 infants aged 3 months to 4 years, except that this study included children of both genders (Cezard et al., 2001). Total stool output was significantly lower by 60% (95% confidence interval 43–88%) with racecadotril as compared to placebo, and this was again independent of rotavirus status. Stool output in the first 24 h of treatment, a secondary endpoint, was also significantly less with racecadotril. The time to recovery was also significantly shorter with racecadotril in a Kaplan–Meier analysis. Moreover, the need for oral rehydration solution on the second day of treatment was also significantly lower with racecadotril treatment. Adverse events were reported by nine patients of each group, but abdominal distension was not noted in either treatment group.

In a third study the effect of racecadotril given as adjunct to oral rehydration solution was compared to rehydration alone in 166 children aged 3–34 months in a randomized open-label study (Cojocaru et al., 2002). The primary endpoint was the number of medical exams within a week after start of treatment which was significantly lower with racecadotril than without (14 vs. 27). Secondary endpoints the number of stools within the first 48 h (6.8 ± 3.8 vs. 9.5 ± 4.5) and the duration of diarrhea (97.2 ± 35.6 vs. 137.7 ± 42.4 h) were also significantly less in the racecadotril group.

In an open-label parallel group study 189 children aged 3–36 months were treated with oral rehydration solution or racecadotril (10 or 30 mg thrice daily for children with 9–13 or >13 kg body weight, respectively) on top of such solution until two normal stools were observed or no bowel movement occurred within 12 h for up to 7 days (Santos et al., 2009). The primary endpoint, number of bowel movements in the first 48 h after initiation of treatment, did not differ significantly between the two treatments (4.1 ± 2.7 vs. 3.8 ± 2.4 for control and racecadotril, respectively), and the duration of gastroenteritis, a secondary endpoint, also differed numerically but not significantly (4.7 ± 2.2 vs. 4.0 ± 2.1 days, respectively). The incidence of adverse events was similar in both treatment groups (20.2 vs. 19.1%, respectively).

In an open-label study 3873 children aged 3 months to 12 years were treated with 1.5 mg/kg thrice daily (Chacon, 2010). The primary endpoint of that study was time to relief, time from start of treatment to last watery bowel movement, which was reached after 18.5 ± 12.5 h (95% confidence interval 17.9–19.0 h). Using the very large number of children in this study, the authors have performed a multiple regression analysis to explore factors affecting drug performance. Among a range of potential explanatory variables diarrhea severity was the only with a significant and independent weight on racecadotril effectiveness, explaining 23% of time to relief variance, but even in severe cases mean time to relief was less than 24 h. Specific adverse event incidence was not reported, but overall tolerability was rated as excellent or good in 95.9% of cases.

While some of the above mentioned randomized pediatric studies have been systematically reviewed in the past (Szajewska et al., 2007; Tormo et al., 2008), those analyses covered only part of the existing trial databases. More importantly, an individual patient meta-analysis of nine pediatric studies with raw data available for analysis from 1384 children has been reported more recently which also included some studies that had not been reported before as full papers (Lehert et al., 2011). The proportion of children with recovery was higher with racecadotril treatment with a hazard ratio of 2.04 (95% confidence interval 1.85–2.32). For in-patient studies, the ratio of mean stool output racecadotril/placebo was 0.59 (0.51–0.74, p < 0.001), for out-patient studies, the ratio of the mean number of diarrhoic stools racecadotril/placebo was 0.63 (0.51–0.74, p < 0.001).

Accordingly, a survey of treatment patterns among all office-based pediatricians in France found racecadotril to be prescribed by 62% of pediatricians as compared to only 28% prescribing loperamide (Uhlen et al., 2004). In line with the above studies, as an addition to oral rehydration treatment, racecadotril is being recommended for the treatment of acute diarrhea in children by recent guidelines, e.g., from the World Gastroenterology Organisation (World Gastroenterology Association, 2008), the European Society of Pediatric Gastroenterology, Hepatology and Nutrition/European Society for Pediatric Infectious Diseases (Guarino et al., 2008), a guideline panel from Spain and Latin America (Gutierrez Castrelion et al., 2010), and the German Society for Pediatric Gastroenterology and Nutrition (Koletzko and Lentze, 2008). Similarly, a very recent international panel of experts from France, Ireland, Italy, Malaysia, Peru, Spain, USA, and Vietnam emphasized that oral rehydration solution is the basis of the treatment of acute diarrhea in children; the use of loperamide was discouraged, whereas racecadotril was recognized as an option for additional active treatment on top of rehydration solution (Guarino et al., 2012).

As loperamide has been the primary medical treatment of acute diarrhea for a long time, particularly in adults, six studies have reported direct comparisons of the efficacy and tolerability of racecadotril and loperamide including one study in children.

Roge et al. (1993) reported a double-blind controlled study in which 100 mg racecadotril was compared to 1.33 mg loperamide (two doses at start of treatment, followed by one dose every 8 h) in 37 vs. 32 patients, respectively. The study did not report a primary endpoint but rather several outcome parameters in parallel including physician evaluation of efficacy, time to diarrhea resolution, abdominal pain for more than 1 day, abdominal distension for more than 1 day, duration of abdominal distension and secondary constipation. Racecadotril was numerically superior to loperamide for all of these endpoints, and the difference reached statistical significance for the latter three. The authors did not specifically report incidence of adverse events but secondary constipation can be considered as such and was found significantly less frequent in racecadotril as compared to loperamide-treated patients (8.1 vs. 31.3%).

A second randomized, double-blind, double-placebo-controlled study compared the efficacy and safety of 100 mg racecadotril thrice daily with that of 2 mg loperamide after each diarrhoic stool in 82 and 75 patients, respectively, administered until recovery for a maximum of 7 days (Vetel et al., 1999). Both groups passed a similar number of stools (3.5 ± 0.5 vs. 2.9 ± 0.4) and had a similar duration of diarrhea (14.9 ± 2.0 vs. 13.7 ± 2.2 h). Adverse events were reported in 7.4% of racecadotril and 12% of loperamide patients, and rebound constipation was experienced by 9.8% of racecadotril vs. 18.7% of loperamide patients.

The third and largest direct head-to-head study compared 473 patients receiving 100 mg racecadotril thrice daily with 472 patient receiving 2 mg loperamide thrice daily in a single-blind design (Prado, 2002). The primary efficacy criterion was duration of diarrhea, defined as time between start of treatment and appearance of first formed stool; this was 55.0 h in both groups (95% confidence interval 50.0–65.0 and 48.0–66.0 h in the racecadotril and loperamide group, respectively). The median duration of abdominal pain was similar in both groups, but difference in pain intensity between start and end of study was significantly in favor of racecadotril; moreover, a significantly greater percentage of patients reported residual pain at study end with loperamide as compared to racecadotril treatment (7 vs. 3%). Rebound constipation (objectively defined as 36 h without passing stool) was significantly more frequent with loperamide than with racecadotril treatment (25 vs. 16%). The incidence of reported adverse events was also significantly greater with loperamide than with racecadotril (23.9 vs. 14.2%).

A fourth randomized study compared 31 patients receiving 100 mg racecadotril thrice daily to 31 patients receiving 2 mg loperamide twice daily in a single-blind manner (Wang et al., 2005). The primary endpoint was duration of diarrhea, which did not differ significantly between treatments (median 19.5 vs. 13.0 h for racecadotril and loperamide, respectively). Duration of abdominal pain, abdominal distension, anal burning, and nausea also did not differ significantly between treatments. Adverse events were reported in 25.0% of racecadotril and 22.0% of loperamide patients. Among these rebound constipation was reported in four and nine racecadotril and loperamide patients, respectively, whereas itching was found in two racecadotril but no loperamide patients.

A fifth double-blind, randomized study compared 100 mg racecadotril thrice daily to loperamide (4 mg starting dose, followed by 2 mg after each unformed stool for a maximum of 8 mg/d) in 30 and 31 patients, respectively, being administered until recovery which was defined by two consecutive normal stools or no stools in a 12-h period (Gallelli et al., 2010). In contrast to the other studies, this one recruited only elderly nursing home residents with acute diarrhea (mean age 82 years). The primary endpoint was time to recovery which was met significantly earlier with racecadotril than with loperamide treatment (36 ± 4 vs. 63 ± 6 h). Secondary endpoints included duration of abdominal pain, number of diarrhea episodes, and total stool output in the intention-to-treat and in the per-protocol populations; racecadotril was numerically superior to loperamide all of these, and this reached statistical significance in several cases. In 50% of patients loperamide was ineffective within 4 days; these were switched to racecadotril resulting in rapid normalization of all symptoms. Adverse events were reported in 12% of racecadotril and 60% of loperamide patients; the latter percentage is much higher than in other loperamide studies, probably due to the elderly population being studied. Specifically, nausea and rebound constipation were noted more frequently with loperamide than with racecadotril, whereas abdominal pain, headache, and anorexia were seen similarly with both treatments. Based on genotyping for cytochrome P450 3A4 and 2C8 it was excluded that the group differences were not attributable to the presence of ultra-rapid or poor metabolizers. In a pharmacoeconomic analysis the average cost in the loperamide group was twice as high as in the racecadotril group (€ 91.99 vs. € 44.85).

A sixth randomized study directly comparing racecadotril and loperamide in the treatment of acute diarrhea was performed in a pediatric population with a mean age of 4.7 years (range 2–10 years) in a double-blind, double-placebo design (Turck et al., 1999). Fifty-two children received 1.5 mg/kg racecadotril thrice daily and 50 received 0.03 mg/kg loperamide thrice daily. The primary endpoint was number of passed stools until recovery which did not differ significantly between the two groups (2.7 ± 0.4 with racecadotril and 2.1 ± 0.4 with loperamide). The mean duration of diarrhea also did not differ significantly between groups (10.7 ± 1.7 h with racecadotril and 8.8 ± 2.3 h with loperamide). Adverse events were noted in 11.5% of racecadotril and 22% of loperamide patients. There were only statistically significant differences between the two treatments: rebound constipation (36.5% of racecadotril and 58% of loperamide patients) and need for concomitant medication (anti-emetics 5 vs. 8, analgesics 0 vs. 3 patients, oral rehydration 0 vs. 2, and laxatives 0 vs. 1 patient on racecadotril and loperamide, respectively.

Taken together these six studies demonstrate that the efficacy of racecadotril and loperamide in the treatment of acute diarrhea did not differ significantly for some endpoints in some studies, but that racecadotril was significantly more effective for at least some endpoints in some studies. The incidence of adverse events also was similar in some studies but significantly less frequent with racecadotril in some other studies. A much lower frequency of rebound constipation was seen with racecadotril in almost all studies. Thus, in comparison to loperamide racecadotril appeared to be superior in efficacy and tolerability when all studies are taken into consideration.

While the mechanisms underlying differential effects of racecadotril and loperamide in acute diarrhea have not been fully established, two candidates have emerged: Firstly, loperamide preferentially acts on μ-opioid receptors (Dehaven-Hudkins et al., 1999), whereas endogenous enkephalines active both μ- and δ-receptors (Huighebaert et al., 2003). Secondly, NEP inhibition will not only increase exposure to endogenous enkephalines but also to endogenous neuropeptide Y and, possibly, peptide YY, both of which have strong anti-secretory effects in the gut (Playford and Cox, 1996).