Stereoselective pharmacokinetics of cisapride in healthy volunteers and the effect of repeated administration of grapefruit juice

Abstract

Aims

To determine whether the pharmacokinetics of cisapride and its interaction with grapefruit juice are stereoselective.

Methods

The study was a randomized, two-phase cross over design with a washout period of 2 weeks. Ten healthy volunteers were pretreated with either water or 200 ml double strength grapefruit juice three times a day for 2 days. On the 3rd each subject ingested a single 10 mg dose of rac-cisapride tablet. Double strength grapefruit juice (200 ml) or water was administered during cisapride dosing and 0.5 and 1.5 h thereafter. Blood samples were collected before and for 32 h after cisapride administration. Plasma concentrations of cisapride enantiomers were measured by a chiral h.p.l.c. method. A standard 12-lead ECG was recorded before cisapride administration (baseline) and 2, 5, 8, and 12 h later.

Results

This study showed that cisapride pharmacokinetics are stereoselective. In control (water treated) subjects, the mean C_max (30 ± 13.6 ng ml⁻¹; P = 0.0008) and AUC(0, ∞) (201 ± 161 ng ml⁻¹ h; P = 0.029) of (−)-cisapride were significantly higher than the C_max (10.5 ± 3.4 ng ml⁻¹) and AUC(0, ∞) (70 ± 51.5 ng ml⁻¹ h) of (+)-cisapride. There was no marked difference in elimination half-life between (−)-cisapride (4.7 ± 2.7 h) and (+)-cisapride (4.8 ± 3 h). Compared with the water treated group, grapefruit juice significantly increased the mean C_max of (−)-cisapride from 30 ± 13.6–55.5 ± 18 ng ml⁻¹ (95% CI on mean difference, −33, −17; P = 0.00005) and of (+)-cisapride from 10.5 ± 3.4 to 18.4 ± 6.2 ng ml⁻¹ (95% CI on mean difference, −11.8, −3.9, P = 0.00015). The mean AUC(0, ∞) of (−)-cisapride was increased from 201 ± 161 to 521.6 ± 303 ng ml⁻¹ h (95% CI on mean difference, −439, −202; P = 0.0002) and that of (+)-cisapride from 70 ± 51.5 to 170 ± 91 ng ml⁻¹ h (95% CI on mean difference, −143, −53; P = 0.0005). The t_max was also significantly increased for both enantiomers (from 1.35 to 2.8 h for (−)-cisapride and from 1.75 to 2.9 h for (+)-cisapride in the control and grapefruit juice group, respectively; P < 0.05). The t_½ of (−)-cisapride was significantly increased by grapefruit juice, while this change did not reach significant level for (+)-cisapride. The proportion of pharmacokinetic changes brought about by grapefruit juice was similar for both enantiomers, suggesting non-stereoselective interaction. We found no significant difference in mean QTc intervals between the water and grapefruit juice treated groups.

Conclusions

The pharmacokinetics of cisapride is stereoselective. Grapefruit juice elevates plasma concentrations of both (−)- and (+)-cisapride, probably through inhibition of CYP3A in the intestine. At present, there are no data on whether the enantiomers exhibit stereoselective pharmacodynamic actions. If they do, determination of plasma concentrations of the individual enantiomers as opposed to those of racemic cisapride may better predict the degree of drug interaction, cardiac safety and prokinetic efficacy of cisapride.

Introduction

Cisapride (+/–)cis-4-amino-5-chloro-N-[1[3-(4-fluorophenoxypropyl]-3-methoxy-4-piperidinyl]-2 methoxybenzamide (Figure 1), is a gastrointestinal (GI) prokinetic agent that has been widely used in adults and children for the treatment of upper GI motility disorders such as gastroesophageal reflex diseases, nonulcer dyspepsia and gastroparesis [1, 2]. Cisapride has gained popularity as a prokinetic agent primarily because it is devoid of antidopaminergic effects and, therefore, has less CNS or psychomotor adverse effects when compared with older prokinetic drugs such as metoclopramide [2].

Figure 1.

Chemical structure and specific configurations of cisapride enantiomers.

However, the link of cisapride use with fatal ventricular dysrhythmias in recent years has become a major safety concern for researchers, regulatory authorities and patients. Although epidemiological studies [3, 4] failed to recognize or identify potential cardiac adverse effects of the drug, cisapride-induced tachycardia which was attributed to its hypotensive effect [5] and dizziness that resulted in clinical dropout [6] have been reported as early as 1986. In 1992, Olsson & Edward [7] reported seven cases of tachycardia and palpitation with cisapride administration. However, the true seriousness of cisapride cardiac risk was appreciated in 1996 when the US FDA through its MedWatch reporting program received in the period between September 1993 to April 1996, a total of 57 cases of dysrhythmias associated with cisapride use, some of which were fatal [8]. Subsequent, clinical cases and studies have implicated this drug as the cause of serious cardiac dysrhythmias including torsade de pointes that may precipitate syncope and sudden cardiac death (review [9, 10]). These dysrhythmias are believed to be due to the ability of cisapride to delay cardiac repolarization [11] and prolong the action potential duration as well as QT interval [12, 13], partly by blocking the rapid component of the delayed rectifier potassium current (I_kr) [11]. To date, 386 reports of serious ventricular dysrhythmias and 125 fatalities suspected to be due to cisapride are documented worldwide [14]. In the US, as of December 31, 1999, the FDA has received a total of 341 reports of rhythm abnormalities and 80 fatalities associated with cisapride use [15]. Currently, broad marketing of cisapride is suspended in many countries including the US, UK, Canada and Germany [14], but the drug continues to be available to patients who meet eligibility criteria in the US and it is still widely used in other countries. It follows that knowledge of factors that modulate the cardiac risk of cisapride is important to improve its safe use.

The occurrence of ventricular dysrhythmias with cisapride are generally very rare in the absence of other risk factors; elevated plasma concentrations produced by concurrently administered metabolic inhibitor agents being important in this respect [8, 10, 16]. We [17] and other authors [18] have demonstrated that CYP3A is the major isoform involved in the in vitro metabolism of rac-cisapride. Indeed, cardiac abnormalities have been reported to occur when CYP3A inhibitor drugs such as clarithromycin, erythromycin, ketoconazole, itraconazole and dilitiazem are coadministered with cisapride (review [8, 10, 16]).

Clinical studies demonstrate that inhibition of CYP3A augments plasma cisapride concentrations and/or prolongs QTc interval. van Haarst et al. have recently established a link between increased plasma concentrations of cisapride and QTc interval prolongation [19]. These authors have shown that repeated administration of clarithromycin brings about a three-fold increase in the steady-state plasma concentrations of cisapride in normal volunteers that was associated with an average QTc increase of 25 ms above pretreatment values. Two studies in normal volunteers have shown that grapefruit juice increases the systemic availability of rac-cisapride [20, 21]. Grapefruit juice is known to increase the oral bioavailability of several drugs, particularly those whose presystemic metabolism in the intestine by CYP3A is extensive [22, 23].

However, all studies involving the pharmacology of cisapride have only been performed with the racemate. As shown in Figure 1, cisapride has two asymmetric carbons at the 3 and 4 positions of the piperidinyl ring and is clinically marketed as a racemate of two optical isomers [(−)-cisapride and (+)-cisapride] of a cis (but not a trans) configuration. Despite the use of this drug for more than a decade, little is known about its stereoselective disposition. We have in vitro evidence that cisapride metabolism is enantiospecific [24], and our preliminary data in two healthy volunteers suggest stereoselective pharmacokinetics [25]. In the present study, we tested whether cisapride pharmacokinetics is stereoselective after administration of a single oral dose of 10 mg rac-cisapride to healthy subjects. Previous studies indicate that grapefruit juice interacts with rac-cisapride [20, 21], but it is not known to what extent it modifies the pharmacokinetics of the individual enantiomers. Therefore, we also determined the effect of repeated administration of grapefruit juice on the pharmacokinetics of cisapride enantiomers in healthy volunteers.

Methods

Subjects and study design

Plasma samples collected originally from 10 healthy volunteers to determine the effect of repeated grapefruit juice on the pharmacokinetics of rac-cisapride [20] were reanalysed to measure the plasma concentrations of each cisapride enantiomer. Subjects, study design and blood and ECG sampling protocols are detailed in an earlier publication [20]. In brief, 10 healthy volunteers (age 21–31 years; weight, 57–83 kg) who had fulfilled eligibility criteria and provided written informed consent were studied in a randomized, two periods (2 week interval), crossover design. The volunteers ingested 200 ml double strength grapefruit juice [12 ounces (355 ml) Minute Maid frozen concentrated grapefruit juice, Coca Cola Foods, Houston, Texas] or water (200 ml) three times a day for 2 days. On day 3, after subjects had fasted overnight, each was given a single oral dose of 10 mg rac-cisapride (one 10 mg Propalsid® tablet, Janssen-Cilag, Beerse, Belgium) with 200 ml grapefruit juice or water at 09.00 h. In addition, subjects received 200 ml grapefruit juice or water 0.5 and 1.5 h after cisapride intake. Blood samples were collected before (t = 0 h) and 0.5, 1, 1.5, 2, 3, 4, 5, 6, 9, 12, 24 and 32 h after cisapride administration into siliconized Venoject tubes containing EDTA. Plasma was separated within 30 min and stored at −70 °C until analysis. A standard 12-lead ECG was recorded before and 2, 5, 8 and 12 h after cisapride administration.

Assay of cisapride enantiomers

Plasma cisapride enantiomers were measured using a chiral h.p.l.c. assay with u.v. detection that we have developed [25]. In brief, a 1 ml aliquot of human plasma samples and 50 µl of the internal standard solution (2 µg ml⁻¹ clebopride in ethanol) were made alkaline by adding 0.25 ml of 0.1 m disodium tetraborate buffer (pH 13.3) and extracted using tert-butyl methylether (3 ml). The organic phase was back extracted with 0.75 ml 0.05 m sulphuric acid and the acidic phase obtained after centrifugation at 2000 rev min⁻¹ for 5 min was re-extracted with 3 ml tert-butyl methyl ether after it was made alkaline with disodium tetraborate buffer (0.1 m; 0.75 ml). The organic phase was then evaporated using speed vacuum and 150 µl aliquot was injected onto the h.p.l.c. column after reconstitution of the residues in 200 µl of mobile phase. The separation system consisted of (250 × 4.6 mm) OJ stainless steel ChiralCel column (Chiral Technologies, Inc. Exton, PA, USA) with a Waters CN guard column and a mobile phase consisting of ethanol-hexane-diethylamine (35:64.5:0.5, v/v/v; flow rate, 1.2 ml min⁻¹). The limit of quantification and detection of both cisapride enantiomers in human plasma were 5 and 1 ng ml⁻¹, respectively [25]. This assay method was linear over a range of concentrations (5–125 ng ml⁻¹) of each enantiomer. The intra- and interday CVs determined at 7.5 ng ml⁻¹ in plasma were less than 15% for both enantiomers of cisapride.

Pharmacokinetic and ECG analysis

The peak concentration in plasma of each cisapride enantiomer (C_max) and time to C_max (t_max) were observed values from the original data. The terminal elimination rate constant (k_e) and the t_½ were derived from regression of the natural log of the plasma concentration vs time for concentration in the log-linear elimination phase. AUC(0, 32 h) was calculated by the trapozoidal rule, AUC(0, ∞) was calculated by dividing the last measured concentration by k_e.

The QT interval recorded was corrected for heart rate (QTc interval) by dividing the measured QT interval by the cube root of the RR interval using the formula of Fridericia [26].

Statistical analysis

The data are expressed as mean values (± s.d.), with 95% confidence intervals on mean differences. Data were analysed by Student's t-test (two-tailed) for paired values. P < 0.05 were considered significant.

Results

The mean (± s.d.) plasma concentration vs time curves of (−)-cisapride (Figure 2a, open circles) and (+)-cisapride (Figure 2b, open circles) in subjects who were treated with water as control indicate that the plasma concentrations of (−)-cisapride were markedly higher than those of (+)-cisapride at any given sampling time. The mean (± s.d.) pharmacokinetic parameters of each cisapride enantiomer (with 95% confidence intervals for mean differences) in the water treated group are listed in Table 1. A significantly higher C_max (2.9 ± 4-fold; range: 1.9–6.2; P = 0.0008) and AUC(0, ∞) (2.9 ± 3.1; range: 1.4–4.7; P = 0.029) of (−)-cisapride relative to the C_max and AUC(0, ∞) of (+)-cisapride were observed. No significant difference was noted with regard to the elimination half-life of (−)-cisapride and (+)-cisapride (4.7 ± 2.7 h vs 4.8 ± 3 h, respectively). The t_max of (+)-cisapride (1.75 h) was slightly but significantly higher than the t_max of (−)-cisapride (1.35 h) (Table 1). The mean (−)/(+)-cisapride ratios of plasma concentrations in the water treated group is shown in Figure 2c, essentially confirming stereoselectivity through out the sampling time.

Figure 2.

Plasma concentrations (mean±s.d.) of (−)-cisapride and (+)-cisapride after a single oral dose of 10 mg rac- cisapride in water (W) (control; open circles) or 200 ml double-strength grapefruit juice (GFJ; solid circles) treated healthy volunteers (n = 10). a) (–)-cisapride with and without GFJ, b) (+)-cisapride with and without GFJ; and c), mean (−)/(+)-cisapride ratios of plasma concentrations at post dose times in W (open circles) and GFJ (solid circles) treated groups.

Table 1.

Pharmacokinetic (PK) parameters (mean±s.d.) of (−)-cisapride and (+)-cisapride after a single oral dose of 10 mg rac-cisapride (CIS) in water (control) treated healthy volunteers (n = 10).

PK parameters	(−)-CIS	(+)-CIS	Ratio [(−)-/(+)-CIS]	Mean differences (95% CI)	P value
tmax (h)	1.35 ± 0.4	1.75 ± 0.5	0.8 ± 0.22	−0.4 (−0.78, −0.03)	0.037
Cmax (ng ml−1)	30 ± 13.6	10.5 ± 3.4	2.98 ± 1.27	19.6 (10.6, 28.6)	0.0008
AUC(0, ∞) (ng ml−1 h)	201.2 ± 161	70 ± 51.5	3.7 ± 3.7	131 (17, 245)	0.029
t1/2 (h)	4.7 ± 2.7	4.8 ± 3.1	1.22 ± 0.82	−0.07 (−2.6, 2.5)	0.95

Peak plasma concentration (C_max), time to C_max (t_max), area under the plasma concentration-time curve (AUC(0, ∞)) and elimination half-life (t_½).

The mean (± s.d.) plasma concentration vs time curves of (−)- and (+)-cisapride enantiomers in subjects who were treated with 200 ml double-strength grapefruit juice are shown in Figure 2 (solid circles). The plasma concentrations of each cisapride enantiomer were increased by grapefruit juice (Figure 2a and b, solid circles) in all subjects when compared with those water treated subjects (Figure 2a and b, open circles). The mean pharmacokinetic parameters derived (± s.d.), including 95% confidence interval for mean differences, are shown in Table 2. Grapefruit juice significantly increased the mean C_max (1.8 ± 1.3 fold; range: 1.2–2.4, P = 0.00005) and AUC(0, ∞) (2.6 ± 1.88 fold; range: 1.8–3.5; P = 0.00018) of (−)-cisapride compared to the water treated controls (see Table 1). Similarly, the C_max (1.75 ± 1.83 fold; range: 0.7–3.6; P = 0.00015) and AUC(0, ∞) (2.43 ± 1.76; range: 1.1–4.3; P = 0.00048) of (+)-cisapride were significantly increased by grapefruit juice compared with water treated controls. As shown in Figure 3, the effect of grapefruit juice on C_max and AUC of both enantiomers was highly variable among the subjects. In Figure 2c, the mean (−)-cisapride/(+)-cisapride ratios of plasma concentrations in the water and grapefruit juice treated groups are compared. Despite the marked stereoselectivity in the pharmacokinetics of cisapride enantiomers, the proportion of pharmacokinetic changes brought about by grapefruit juice was comparable for both (−)- and (+)-cisapride. Similarly, the mean (−)-cisapride/(+)-cisapride ratios of C_max and AUC for both enantiomers in the water and grapefruit juice treated groups were comparable (Table 2). Compared with water treated controls, grapefruit juice increased the elimination half-life of (−)-cisapride (from 4.7 ± 2.7 to 6.3 ± 2.3 h; P = 0.026). Although the average half-life of (+)-cisapride was also increased from 4.8 ± 3.1 to 6.5 ± 2.8 h by grapefruit, the difference was not statistically significant (P = 0.21). Interestingly, the time to C_max (t_max) was significantly increased (P < 0.05) for both enantiomers [1.35 vs 2.8 h for (−)-cisapride and 1.75 vs 2.9 h for (+)-cisapride in the control and grapefruit juice group, respectively].

Table 2.

Pharmacokinetic (PK) parameters (mean±s.d.) of (−)-cisapride and (+)-cisapride after a single oral dose of 10 mg rac-cisapride (CIS) in water (W) (control) and 200 ml double-strength grapefruit juice (GFJ) treated healthy volunteers (n = 10).

PK parameters	W phase	GFJ phase	Ratio GFJ/W phase	Mean differences (95% CI)	P values
(–)-Cisapride
tmax (h)	1.35 ± 0.4	2.8 ± 1	2.33 ± 1.19	−1.45 (−2.33, −0.57)	0.0048
Cmax (ng ml−1)	30 ± 13.6	55.5 ± 18.2	1.96 ± 0.52	−25 (−33, −17)	< 0.0001
AUC(0, ∞) (ng ml−1 h)	201.2 ± 161	522 ± 303	2.79 ± 0.54	−320 (−439, −202)	0.0002
t1/2 (h)	4.7 ± 2.7	6.3 ± 2.4	1.48 ± 0.42	−1.5 (−2.8, −0.23)	0.026
(+/–)-Cisapride
tmax (h)	1.75 ± 0.5	2.9 ± 1.6	1.71 ± 0.86	−1.15 (−2.12, −0.18)	0.025
Cmax (ng ml−1)	10.5 ± 3.4	18.4 ± 6.2	1.89 ± 0.75	−7.9 (−11.8, −3.9)	0.0015
AUC(0, ∞) (ng ml−1 h)	70 ± 51.5	170 ± 91	2.75 ± 1.04	−100 (−143, −53)	0.0005
t1/2 (h)	4.8 ± 3.1	6.5 ± 2.8	1.8 ± 1.21	−1.75 (−4.6, −1.14)	0.205

Peak plasma concentration (C_max), time to C_max (t_max), area under the plasma concentration-time curve (AUC(0, ∞)) and elimination half-life (t_½).

Figure 3.

Interindividual variability in C_max, AUC(0, ∞) and t_½ of (−)-cisapride and (+)-cisapride after a single oral dose of 10 mg rac-cisapride in water (W) (control) and 200 ml double-strength grapefruit juice (GFJ) treated healthy volunteers (n = 10).

The QTc intervals determined at baseline and 2, 5, 8 and 12 h following cisapride administration in subjects who were pretreated with water and grapefruit juice have been documented in a previous publication [20]. Compared with baseline, QTc interval was significantly increased 5 h after cisapride administration in both control (water phase; P = 0.0011) and grapefruit juice (P = 0.0067), but no significant difference was observed between the water and grapefruit juice group. In only two subjects (subjects 5 and 9), grapefruit juice markedly increased the QTc interval (ΔQTC_max > 30 ms) compared with the water treatment phase. However, no correlation could be established between these QTc interval changes and plasma concentrations of the enantiomers as the pharmacokinetic changes brought about by grapefruit juice in these two subjects were not different from those subjects whose QTc interval remained unchanged.

Discussion

In the present study, we have demonstrated that cisapride pharmacokinetic is stereoselective. In the water treated subjects, the mean C_max and AUC(0, ∞) of (−)-cisapride were significantly higher than those of (+)-cisapride after administration of a single oral dose of 10 mg rac-cisapride, essentially confirming what was noted in our earlier preliminary data involving two subjects [25]. These data also correlate very well with our recent in vitro microsomal study that both cisapride enantiomers are predominantly oxidized by CYP3A at a different rate, (+)-cisapride being more efficiently metabolized relative to (−)-cisapride [24]. The pharmacokinetic differences we noted here were reflected more in the AUC and C_max where the mean (−)-cisapride/(+)-cisapride ratios were 2.87 and 2.88, respectively, with little difference in other PK parameters, suggesting that differences in the presystemic metabolism of the enantiomers by CYP3A in the intestine and/or the liver is the predominant mechanism of cisapride stereoselective pharmacokinetics.

The ability of grapefruit to alter the disposition of rac-cisapride has been previously reported [20, 21]. Our data show that repeated administration of grapefruit juice markedly increases the plasma concentrations of both cisapride enantiomers compared to the water treatment group. Pronounced increase in the C_max and AUC(0, ∞) of each enantiomer were observed in all subjects. The effect of grapefruit juice on the pharmacokinetics of cisapride (racemic and both enantiomers) is likely to be explained through inhibition of CYP3A. First, grapefruit juice components are known to inhibit CYP3A in vitro [22, 27, 28], and in human volunteers a pronounced increase in the systemic availability of CYP3A substrates drugs such as midazolam, terfenadine and felodipine [22, 23]) has been documented during concurrent administration of grapefruit juice. Second, the major oxidative metabolic pathways of rac-cisapride [29] and its enantiomers [24] have been shown to be primarily catalysed by CYP3A in vitro [17, 18, 24]. The marked increase in the AUC and C_max of cisapride enantiomers with less pronounced alteration in half-lives indicate that the pharmacokinetic changes were caused predominantly by inhibition of pre-systemic metabolism of cisapride, thus increasing the systemic availability of the enantiomers. This is not surprising as grapefruit is known to preferably down regulate CYP3A in the intestine without much effect on the liver [22, 27, 28]. Of note, clinical drug-interaction studies by Kivisto et al. [20] and Gross et al. [21] have documented that grapefruit juice slows the presystemic elimination of rac-cisapride relative to systemic elimination. At the doses of grapefruit juice used, the half-lives of the enantiomers were increased by 34–35%, suggesting some contribution of inhibition of liver CYP3A in this interaction. The wide interindividual variation in the degree of grapefruit juice interaction with cisapride enantiomers is consistent with previous data with rac-cisapride [20, 21] and other CYP3A substrate drugs [22, 23], and is probably attributed to variation of CYP3A expression in the intestine [28]. Grapefruit and grapefruit juice are common food constituents and their intake may contribute to wide interindividual variability in the pharmacokinetics of cisapride and its enantiomers.

Interestingly, we noted a higher t_max of both enantiomers in the grapefruit juice group. This could be simply due to delayed absorption of cisapride by grapefruit juice. Alternatively, it could reflect an effect of grapefruit juice on certain transport proteins. For drugs that are substrates of membrane transporters, delayed t_max might be expected during induction of P-glycoprotein or during inhibition of organic anion transporting polypeptides (OATPs). In vitro, cisapride has been shown not to be a substrate of P-glycoprotein [30] and there is no evidence that grapefruit is an inducer of P-glycoproteins. Recently, grapefruit juice has been shown to inhibit OATPs [31]. Although we do not know whether cisapride enantiomers are substrates of OATPs, the possibility that grapefruit juice might have delayed the t_max of cisapride enantiomers through inhibition of OATPs cannot be excluded.

Studies have shown that the degree of metabolic inhibition is predictably high for drugs whose bioavailability is very low as a result of very high first metabolism (e.g. nimodipine, saquinavir and terfenadine) even after a single dose of the substrate and the inhibitor of CYP3A [32]. On the other hand, maximum interaction is anticipated at a steady-state plasma concentration for drugs whose bioavailability is high [32]. The absolute oral bioavailability of rac-cisapride is 40–50% [2], suggesting that the presystemic metabolism of the drug is intermediate. Since (+)-cisapride appears to be more efficiently eliminated during first-pass relative to (−)-cisapride and probably has lower bioavailability, we anticipated that the effect of grapefruit juice on the pharmacokinetics of (+)-cisapride would be higher than that of (−)-cisapride. If (+)-cisapride happens to be more cardiotoxic, a high degree of inhibition and the relative accumulation of (+)-cisapride may place patients at greater cardiac risk. In the present study, grapefruit juice was shown to have no stereoselective effect on the pharmacokinetics of cisapride. These findings, although contrary to our initial expectation, are in good agreement with our recent in vitro data where other CYP3A inhibitors such as troleandomycin showed comparable degree of inhibitory potency with regard to the metabolism of (−)-cisapride (K_i, ∼12 µm) and (+)-cisapride (K_i, ∼14 µm) [24]. The inhibitory effect of grapefruit juice on the pharmacokinetics of different drug enantiomers that are substrates of CYP3A has been well described. For example, Ho et al. [33] have shown that grapefruit juice administration produces proportional increase of AUC and C_max of R- and S-verapamil, with no evidence of a stereoselective interaction. Similarly, Soon et al. [34] reported that the systemic availability of nitrendipine enantiomers increases at similar proportion when grapefruit is administered with racemic nitrendipine, although other authors [35] have found a slightly higher effect of grapefruit juice on (−)-nitrendipine than (+)-nitrendipine elimination. Thus, the interaction of grapefruit juice with cisapride and other chiral CYP3A substrate drugs does not appear to involve stereoselectivity, provided the metabolic route and the enzymes catalysing the enantiomers are the same. Since the inhibitory effect of grapefruit juice on cisapride seems to be largely limited to the intestine, it would be interesting to test whether other potent inhibitors of CYP3A in both the liver and intestine (e.g. ketoconazole and clarithromycin) when evaluated at steady state of both the inhibitor and cisapride enantiomers would produce maximum interaction and preferential accumulation of one enantiomer over the other and thus modify toxicity. Of note, treatment with clarithromycin (500 mg twice a day for 10 days), a drug that inhibits CYP3A in the intestine and liver, has been shown to significantly increase steady state plasma rac-cisapride concentrations with substantial QTc interval prolongation [19], but plasma concentrations of the individual enantiomers were not measured in this study.

An earlier study has shown that mean QTc interval was significantly lengthened from the baseline value during both the water and grapefruit juice treatment phases 5 h after racemic cisapride administration [20]. The maximum QTc interval appears to occur later than the time (t_max, < 3 h) to mean C_max, probably reflecting delay in distribution of cisapride from plasma to cardiac tissues. On the other hand, the pharmacokinetic changes in cisapride enantiomers brought about by grapefruit did not bring about parallel QTc interval changes [20]. However, there were some limitations to our pharmacodynamic study. The ECG recordings were done at only four different times and the study was done at a single 10 mg dose of cisapride to minimize potential harm to the subjects studied, although we recognize that the effective daily dose of cisapride ranges from 15 to 40 mg [2, 36]. Thus, these data do not exclude the possibility that the interaction of cisapride with grapefruit juice may produce clinically significant ECG changes, particularly in patients who consume large amounts of grapefruit juice, take high doses of cisapride over extended period of time and have other risk factors (e.g. electrolyte imbalance and long QT syndrome). Physicians should instruct patients receiving cisapride on the potential for adverse effects during excessive consumption of grapefruit juice.

The widespread marketing of cisapride in the US and elsewhere has been suspended owing to its cardiac toxicity, but the drug continues to be available for selected patients who meet eligibility criteria for a limited-access protocol. Inhibition of CYP3A by a number of inhibitors/substrates has been shown to slow the elimination of rac-cisapride and increase its cardiac risk [8, 10, 16]. Full understanding of these interactions requires consideration of the marked stereoselective pharmacokinetics of cisapride documented in the present study. It may be possible that one enantiomer is more cardiotoxic while the other is mainly responsible for the prokinetic action, or the enantiomers have similar effects but with different potency. Provided the actions of cisapride are stereoselective and there is concentration-effect relationship, the plasma concentrations of the individual enantiomers as opposed to the rac-cisapride may be a more reliable predictor of therapeutic failure during treatment with CYP3A inducers or for toxicity with inhibitors of CYP3A.