Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Opioid compounds, which are widely administered for a variety of medical indications, are associated with a number of side effects, including constipation and troublesome subjective effects (e.g., dysphoria, dizziness, nausea, and pruritus). Clinically, it would be desirable to reduce opioid-induced peripherally mediated side effects, while maintaining centrally mediated analgesic effect. Selective antagonism of opioid-induced side effects by tertiary compounds such as naloxone or nalmephene have been attempted. Success has been limited by the propensity for these compounds to reverse analgesia or to induce opioid withdrawal (Gowan et al., 1988; Sykes, 1991; Culpepper-Morgan et al., 1992; Cheskin et al., 1995).

N-Methylnaltrexone bromide (or methylnaltrexone) is a quaternary derivative of the pure opioid antagonist, naltrexone (Brown and Goldberg, 1985). The addition of the methyl group at the amine in its ring forms a compound with greater polarity and lower lipid solubility. Thus, methylnaltrexone does not cross the blood-brain barrier in humans (Russell et al., 1982; Brown and Goldberg, 1985). These properties provide methylnaltrexone with the potential to block undesired side effects of opioid pain medications predominantly mediated by peripherally located receptors (Tavani et al., 1980; Manara et al., 1986), while sparing centrally mediated analgesic effect.

In previous human volunteer trials, we demonstrated that intravenous methylnaltrexone prevented morphine-induced delay in gastrointestinal motility and transit time without affecting analgesia (Yuan et al., 1996). In another preliminary observation, we observed that the compound reduced morphine-induced troublesome subjective effects, such as nausea, skin itch, stimulation, and flushing (Yuan et al., 1998). The present study was designed to evaluate the efficacy of subcutaneously administered methylnaltrexone, a more convenient route of administration, on morphine-induced changes in gastrointestinal transit time and subjective effects in healthy volunteers. Pharmacokinetic comparisons were also made after intravenous, oral, and subcutaneous methylnaltrexone.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. With approval from the Institutional Review Board at the University of Chicago, eight males (all Caucasians) and four nonpregnant females (two Caucasians and two African Americans) participated and completed this study. Mean age ± S.D. was 24.8 ± 5.9 (range 19-38) years. All subjects were screened with a medical history, physical examination, 12-lead resting ECG, complete blood count with differential and platelet count, blood chemistries (sodium, potassium, chloride, carbon dioxide, creatinine, blood urea nitrogen, total protein, serum glutamic-oxaloacetic transaminase, serum glutamic-pyruvic transaminase, alkaline phosphatase, glucose, calcium, bilirubin, and serum albumin), and urinalysis (specific gravity, pH, protein, blood, glucose, reducing substances, ketones, bilirubin, urobilinogen leukophil esterase, nitrite, and a microscopic examination). Urine toxicology screening for use of other drugs was also performed. Subjects with drug abuse disorders or medical contraindications that would prevent them from participating in the study were excluded.

Protocol. After obtaining written informed consent, subjects were admitted for each experimental day (or session) in the morning to the General Clinical Research Center at the University of Chicago Hospitals after fasting from midnight. There were three sessions, each separated by at least 1 week.

Hydrogen Breath Test. In each session, gastrointestinal transit time was assessed by measuring pulmonary hydrogen (Bond and Levitt, 1975; Basilisco et al., 1985, 1987). This method, which was successfully used in our previous methylnaltrexone studies (Yuan et al., 1996, 1997, 2000), is based on the measurement of hydrogen produced in exhaled air when unabsorbable disaccharide (lactulose) is fermented by colonic bacteria. The time between ingestion of lactulose and the rise of hydrogen in the breath represents the oral-cecal transit time.

Opioid Subjective Effects. A modified opiate adjective checklist reflecting opioid agonist effects was used (Fraser et al., 1961; Zacny et al., 1994; Yuan et al., 1998). This list consisted of 12 items: "flushing", "stimulated", "numb", "drunken", "difficulty in concentrating", "drowsy (sleepy)", "coasting or spaced out", "turning of stomach", "skin itch", "dry mouth", "dizzy", and "nauseous". Subjects were instructed to rate each of these items on a five-point scale from 0 ("not at all") to 4 ("extremely"). The checklist was completed immediately before the onset of the session (baseline or time 0), 5 min after morphine injection (i.e., time 20 min), and 180 min after morphine injection (i.e., time 195 min). After each test, the ratings for the 12 individual items were summed to give a total subjective symptom score.

Blood and Urine Sampling and Analysis. In each session an intravenous catheter was placed for administration of drugs and blood drawing. Vital signs (heart rate and blood pressure) were monitored before and after drug administrations and when venous blood was collected. Venous blood samples were drawn for plasma drug levels at 0, 2, 5, 10, 15, 20, 30, 45, 60, and 90 min and 2, 3, 4, and 6 h. Urine samples during hours 0 to 3 and 3 to 6 were collected to measure the parent compound.

Measurement of Methylnaltrexone Concentrations. Plasma and urine methylnaltrexone levels were determined by high-performance liquid chromatography technique using a previously reported method (Kim et al., 1989; Yuan et al., 1996). The practical limit of detection for plasma samples was approximately 2 ng/ml.

Drugs. Drugs used were N-methylnaltrexone bromide (Mallinckrodt Chemicals, St. Louis, MO), morphine sulfate (Sanofi Winthrop Pharmaceuticals, New York, NY), and lactulose (Duphalac, Solvay Pharmaceuticals, Marietta, GA).

Statistics. Results of oral-cecal transit time and subjective rating before and after administration of different drug combinations were analyzed using the Wilcoxon signed rank test. In all cases, P < 0.05 was considered statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

One Asian male was excluded from the study after session 1 due to a low hydrogen value with no peak (all < 8 ppm) up to 4.0 h after lactulose administration. Hydrogen production requires a colonic bacterial flora capable of fermenting carbohydrate and yielding hydrogen gas. Previous studies showed that 2 to 27% of individuals, like this subject, had no hydrogen production after lactulose ingestion (Bond and Levitt, 1977; Gilat et al., 1978).

Methylnaltrexone Prevents Morphine-Induced Delay in Oral-Cecal Transit Time. Oral-cecal transit time data are presented in Fig. 1. Transit time increased after morphine administration in all 12 subjects. For the group of six subjects who received 0.1 mg/kg subcutaneous methylnaltrexone (Fig. 1A), intravenous morphine significantly increased the transit time from a baseline level of 85 ± 20.5 min (mean ± S.D.) to 155 ± 27.9 min (P < 0.01). After methylnaltrexone plus morphine, the transit time decreased to 110 ± 41.0 min. For the group of six subjects who received 0.3 mg/kg subcutaneous methylnaltrexone (Fig. 1B), intravenous morphine significantly increased the transit time from a baseline level of 98 ± 49.1 min to 140 ± 58.2 min (P < 0.01). After methylnaltrexone plus morphine, the transit time decreased to 108 ± 59.6 min (P < 0.05 compared with placebo plus morphine). No laxation response was reported by the subjects after each session.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Oral-cecal transit time changes according to drug administration (abscissa) in three different sessions. Morphine indicates 0.05 mg/kg intravenous morphine. A, individual oral-cecal transit time (ordinate) of six healthy volunteers. MNTX 0.1 indicates 0.1 mg/kg subcutaneous methylnaltrexone . B, individual oral-cecal transit time of another six healthy volunteers. MNTX 0.3 indicates 0.3 mg/kg subcutaneous methylnaltrexone . In both parts, heavy lines represent the mean.

View larger version (53K): [in this window] [in a new window]   Fig. 2.   Dose-related effects of subcutaneous methylnaltrexone compared with 0.45 mg/kg intravenous methylnaltrexone from a previous study in 12 normal subjects or 0.45 IV group (Yuan et al., 1996). 0.30 SC Group, 0.3 mg/kg subcutaneous methylnaltrexone in six subjects. 0.10 SC Group, 0.1 mg/kg subcutaneous methylnaltrexone in another six subjects. 0.45 IV group, 0.30 SC group, and 0.10 SC group prevented 97%, 76%, and 64% of morphine-induced increase in oral-cecal transit time, respectively.

Methylnaltrexone Reduces Morphine-Induced Subjective Effects. Five minutes after morphine administration (i.e., at time 20 min), there were significant increases in subjective ratings (Fig. 3, A and B) (both P < 0.01 compared with time 0). Five minutes after 0.1 mg/kg (Fig. 3A) and 0.3 mg/kg (Fig. 3B) subcutaneous methylnaltrexone, morphine-induced subjective ratings were significantly reduced (P < 0.05 and P < 0.01 compared with placebo plus morphine, respectively).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Subjective ratings according to drug administration (abscissa) in three different sessions. Morphine indicates 0.05 mg/kg intravenous morphine. A, ratings of six subjects in 0.1 mg/kg subcutaneous methylnaltrexone group. B, ratings of another six subjects in 0.3 mg/kg subcutaneous methylnaltrexone group. In each session, the ratings were obtained at baseline or immediately before the onset of the session (Time 0), 5 min after 0.05 mg/kg intravenous morphine injection (Time 20 min), and 180 min after morphine administration (Time 195 min).

Pharmacokinetics and Safety. Plasma concentrations after two subcutaneous methylnaltrexone doses are provided in Fig. 4. After the administration of 0.1 and 0.3 mg/kg subcutaneous methylnaltrexone, the unchanged compound detected in urine from 0 to 6 h was 51.8% and 47.3%, respectively. This can be compared to 0.45 to 0.64 mg/kg intravenous methylnaltrexone , in which the amount of unchanged drug excreted during the same period of time was approximately 50% (Yuan et al., 1996; Foss et al., 1997). No adverse effects of clinical importance were observed in this study.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Box plot of plasma methylnaltrexone concentrations in healthy volunteers after subcutaneous drug injection. A, after 0.1 mg/kg subcutaneous methylnaltrexone (n = 6). B, after 0.3 mg/kg subcutaneous methylnaltrexone (n = 6). Black squares represent the mean values, short horizontal lines represent median values, vertical lines represent 90 and 10 percentile values, respectively, and open boxes represent 75 and 25 percentile values, respectively. Note the same time scale but different concentration scales in parts A and B.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Constipation is a very common side effect of advanced cancer patients receiving chronic opioid treatment (Walsh, 1984; Glare and Lickiss, 1992). Tertiary opioid receptor antagonists, such as naloxone, naltrexone, and nalmephene, cross the blood-brain barrier and block both the beneficial pain-relieving effect and the side effects of morphine. Although oral naloxone may reverse opioid-induced constipation, the therapeutic index is very narrow, i.e., reversal of the gut effects with naloxone occurred at doses near the reversal of analgesia (Sykes, 1991). Naloxone may also induce opioid withdrawal symptoms (Gowan et al., 1988; Fifield, 1991; Culpepper-Morgan et al., 1992). As a novel quaternary, peripheral opioid receptor antagonist, methylnaltrexone, even at high doses, has not shown reversal of analgesic effect of morphine in humans (Ferritti et al., 1981; Yuan et al., 1996; Foss et al., 1997). No opioid withdrawal symptoms were observed in our recent study in chronic methadone subjects (Yuan et al., 2000), further indicating that methylnaltrexone does not penetrate into the brain in humans.

In a previous healthy volunteer study, a dose of 0.45 mg/kg intravenous methylnaltrexone was able to completely reverse morphine-induced changes in gut transit time (Yuan et al., 1996). In this study, a lower subcutaneous methylnaltrexone dose (0.1 mg/kg) was also chosen, because we planned to test this compound in chronic opioid users who were very sensitive to methylnaltrexone compared with normal opioid-naive subjects (Yuan et al., 1999, 2000). Approximately 18% of the dose of intravenous methylnaltrexone (0.08 mg/kg) was needed to reverse chronic opioid-induced constipation compared with a dose of 0.45 mg/kg used in normal subjects (Yuan et al., 1996, 2000). In this study, we observed that 0.3 mg/kg subcutaneous methylnaltrexone significantly prevented morphine-induced delay of oral-cecal transit time in subjects who received a single acute dose of morphine. Among six subjects in the 0.1 mg/kg group, there was a nonresponder (Fig. 1A). Thus, statistical significance was not achieved due to high variability and small sample size. Since chronic opioid users are very sensitive to opioid antagonist (Yuan et al., 2000), it seems that subcutaneous methylnaltrexone at a dose of 0.1 mg/kg or less will reverse chronic opioid-induced constipation significantly.

In addition to opioid-induced gut side effects, opioid-induced unpleasant feelings, such as dizziness, headache, nausea, dry mouth, warmth, tingling, and itchiness, have long been recognized (Lasagna et al., 1955). In humans, the effects of drugs on subjective responses could be expressed in quantitative terms (Smith and Beecher, 1959, 1962; Fraser et al., 1961; Zacny et al., 1994). In this study, using the modified adjective checklist, which is sensitive to the subjective and somatic effects of µ-class opioid agonists (Preston et al., 1989), we observed that subcutaneous methylnaltrexone significantly reduced the overall subjective effect rating.

Data from previous animal studies showed that heroin-induced "rush" sensation was involved in the opioid receptors located within the central nervous system (Koob et al., 1984) and aversive conditioning effects of morphine were primarily mediated through peripheral opioid receptors (Bechara et al., 1987). Human volunteer study data demonstrated that intravenous methylnaltrexone, a peripheral opioid receptor antagonist, did not reverse morphine-induced centrally mediated analgesic effects (Yuan et al., 1996). While some items in the checklist (e.g., "coasting or spaced out") are believed to occur due to opioid effects on the central nervous system, observations from our previous volunteer studies suggest that a peripheral opioid antagonist reduced some of these subjective effects (Yuan et al., 1998). Several items in the checklist used in this study (e.g., "nauseous", "flushing", "skin itch") may not be centrally mediated symptoms (Levy et al., 1989; Foss et al., 1993; Reisine and Pasternak, 1996). Because of relatively low statistical power (six subjects per group), we summed the 12 individual items in the checklist to obtain a total subjective symptom score, rather than analyzing each individual item.

"Drowsy (sleepy)" perhaps is another centrally mediated opioid effect, and only a slight increase in this rating after morphine was noted in most subjects. However, two subjects in the 0.1 mg/kg methylnaltrexone group were very sleepy before the onset of two sessions (placebo plus placebo session and placebo plus morphine session), and they marked very high "drowsy (sleepy)" ratings. Since there were only six subjects per group, their selection made the average of these two sessions in Fig. 3A much higher with long error bars. In future experiments, it would be desirable to evaluate the subjective effects in a higher number of subjects, and each item in the checklist can be analyzed separately. The site of action (peripheral versus central) of some subjective symptoms has not been determined yet. Methylnaltrexone can be used as a "probe" to differentiate putatively peripherally mediated or centrally mediated subjective symptoms.

Clinically, subjective effects caused by morphine can cause an unpleasant, troublesome experience. Opioid medications, often given to patients during and after surgical procedures, may possibly delay postoperative recovery because of these subjective effects. Our data suggest that some or many of these effects may be separable from centrally mediated opioid-induced analgesia. As a peripheral opioid receptor antagonist, methylnaltrexone may facilitate faster recovery, as it may decrease unpleasant effects but still allow faster mobilization while preserving analgesia. It appears that methylnaltrexone may have a potential therapeutic value in decreasing some uncomfortable subjective effects due to opioid medication.

In this study, we also compared pharmacokinetic data for subcutaneous methylnatrexone to those parameters from our past drug trials using different routes of administration. Table 1 presents the pharmacokinetic parameters for subcutaneous methylnatrexone obtained from this study and compares these data to previous results from intravenous methylnaltrexone (Yuan et al., 1996, 2000) and oral methylnaltrexone (Yuan et al., 1997). Peak free plasma concentration is significantly lower after subcutaneous injection compared with intravenous administration. Whereas T_max can be reached instantaneously after intravenous dosing, T_max is reached at approximately 16 to 20 min after subcutaneous administrations. T_max is significantly faster after subcutaneous injection compared with oral medication. The dose difference between oral and subcutaneous routes was approximately 100 times, but the AUC values were not remarkably different. For eight chronic methadone subjects after approximately 0.08 ± 0.04 mg/kg slow intravenous infusion of methylnaltrexone (Yuan et al., 2000), the AUC value was similar compared with those normal subjects who received 0.1 mg/kg subcutaneous drug in this study.

Bioavailability of oral drugs is often erratic and incomplete (Rowland and Tozer, 1995). Since methylnaltrexone is a charged compound (Brown and Goldberg, 1985), gut absorption is particularly limited (Yuan et al., 1997). Values of AUC after oral dose also showed a greater variability among individual subjects probably due to high impedance of absorption in the gut.

Subcutaneous administration provided rapid onset with a total drug effect comparable with intravenous or oral administration. We observed efficacy of subcutaneous methylnaltrexone in this study in preventing morphine-induced delay in oral-cecal transit time. Our data suggest that methylnaltrexone by subcutaneous route brings on the effect more rapidly and reliably than the oral route, while avoiding the maintenance of an intravenous site.

Compared with intravenous medication, subcutaneous administration is a more convenient and safer method to deliver drugs (Nucci and Cobelli, 2000; Lepore et al., 2000). In addition, patients with chronic opioid-induced constipation would be able to self-medicate at home, like diabetic patients being able to self-inject insulin subcutaneously. Data from this study showed that subcutaneous methylnaltrexone effectively prevented a single acute dose of morphine-induced gut motility change. In future studies, the dose relationship of agonist to antagonist will be evaluated in opioid-tolerant individuals, such as advanced cancer patients receiving chronic opioid pain medications.