Metabolizers Of Dextromethorphan Research Articles

The effect of perazine on the CYP2D6 and CYP2C19 phenotypes as measured by the dextromethorphan and mephenytoin tests in psychiatric patients.

There is evidence that the antipsychotic drug perazine is an inhibitor of CYP2D6. This study aimed at evaluating its effect on CYP2D6 and CYP2C19 activities in submitting psychiatric patients to phenotyping with dextromethorphan and mephenytoin, respectively, substrates of these enzymes, before and during a treatment with perazine. A total of 31 patients were phenotyped with dextromethorphan (CYP2D6) and mephenytoin (CYP2C19) before and after a 2-week treatment with 450±51mg/day (mean±sd) perazine. At baseline, five patients appeared to be poor metabolizers (PM) of dextromethorphan and two patients of mephenytoin. The metabolic ratio (MR) of dextromethorphan/dextrorphan as determined in collected urine increased significantly (Wilcoxon; P<.0001) from baseline (0.39±1.38 [mean±sd]) till day 14 (1.46±2.22). In 19 out of 26 extensive metabolizers (EM) of dextromethorphan, the phenotype changed from EM to PM. This suggests an almost complete inhibition of CYP2D6 by perazine and/or its metabolites. On the other hand, perazine (or some of its metabolites) did seemingly not inhibit CYP2C19. In conclusion, this study suggests that in patients treated with perazine and co-medicated with CYP2D6 substrates, there could be an increased risk of adverse effects as a consequence of a pharmacokinetic interaction.

Randomized open-label trial of dextromethorphan in Rett syndrome.

To determine safety and perform a preliminary assessment of dose-dependent efficacy of dextromethorphan in normalizing electrographic spikes, clinical seizures, and behavioral and cognitive functions in girls with Rett syndrome. We used a prospective randomized, open-label trial in fast metabolizers of dextromethorphan to examine the effect of dextromethorphan on core clinical features of Rett syndrome. Interictal spike activity and clinical seizures were determined using EEG and parent reporting. Cognitive data were obtained using the Mullen Scales of Early Learning and Vineland Adaptive Behavior Scales, while behavioral data were obtained from parent-completed checklists, the Aberrant Behavior Checklist-Community Version, and the Screen for Social Interaction. Anthropometric data were obtained according to the National Health and Nutrition Examination Survey. The Rett Syndrome Severity Scale provided a clinical global impression of the effect of dextromethorphan on clinical severity. Dextromethorphan is safe for use in 3- to 15-year-old girls with Rett syndrome. Thirty-five girls were treated with 1 of 3 doses of dextromethorphan over a period of 6 months. Statistically significant dose-dependent improvements were seen in clinical seizures, receptive language, and behavioral hyperactivity. There was no significant improvement in global clinical severity as measured by the Rett Syndrome Severity Scale. Dextromethorphan is a potent noncompetitive antagonist of the NMDA receptor channel that is safe for use in young girls with Rett syndrome. Preliminary evidence suggests that dextromethorphan may improve some core features of Rett syndrome. This study provides Class IV evidence that dextromethorphan at various doses does not change EEG spike counts over 6 months, though precision was limited to exclude an important effect.

A case of acute psychosis secondary to coricidin overdose.

A case of acute psychosis secondary to coricidin overdose.

CYP2D6 phenotyping in North Indian subjects in Delhi

The recognition of the importance of inter-subject variation in drug metabolism has increased markedly over the past 20 years. Inter-subject variation may be genetically determined, environmentally induced, or most commonly, a combination of both, resulting in genetic polymorphism. The objective of the present study is to examine the CYP2D6 phenotypes in north Indian population. The study was carried out on seventy-five unrelated healthy north Indian subjects. They were phenotyped with respect to their ability to metabolize dextromethorphan to dextorphan. Oral dextromethorphan (30 mg) was administered to each subject. Urine was collected during 0 to 8 h period after dosing and was analyzed for dextromethorphan and dextorphan by liquid chromatography mass spectrometry (LC-MS/MS). The log 10 (metabolic ratio), calculated as the ratio of dextromethorphan to dextorphan, was bimodally distributed. An antimode value of 0.3 was obtained by plotting a frequency histogram of the log metabolic ratios. The frequency of occurrence of poor metabolizers of dextromethorphan in north Indian subjects from Delhi was found to be 2.6%. Ethiopian Pharmaceutical Journal Vol. 24, 2006: 65-70

Evaluation of CYP2D6 oxidation of dextromethorphan and propafenone in a Chinese population with atrial fibrillation.

The objective of this study was to determine the percentage of patients with paroxysmal atrial fibrillation who were poor metabolizers of CYP2D6 in a Chinese population from Hong Kong and to assess the relationship between the dextromethorphan/dextrorphan ratio and the propafenone/5-hydroxypropafenone ratio or the steady-state propafenone concentration. Patients (n = 60) were recruited from the Arrhythmia Clinic at the University of Hong Kong and given dextromethorphan 30 mg. The dextromethorphan and dextrorphan concentrations in urine over the next 8 hours were used to determine metabolizer status. If the metabolic ratio was greater than 0.3, the patient was determined to be a poor metabolizer. In phase 2, patients (n = 38) were given propafenone 150 mg twice daily, and at steady state, the propafenone and 5-OH propafenone plasma concentrations were determined. It was found that 15% of the patients were poor metabolizers of dextromethorphan. There was a significant correlation between the metabolic ratios of dextromethorphan/dextrorphan and propafenone/5-OH propafenone (r = 0.49, p = 0.0019) and between the dextromethorphan/dextrorphan ratio and the concentration of propafenone (r = 0.32, p = 0.05). No correlations were found in the extensive or poor metabolizer subgroups. It was concluded that the percentage of poor metabolizers in atrial fibrillation patients from Hong Kong was much larger than in previous studies of Chinese patients who were not from Hong Kong. The ability to metabolize dextromethorphan to dextrorphan is related to the ability to metabolize propafenone to 5-hydroxypropafenone.

Fluoxetine augmentation in citalopram non-responders: pharmacokinetic and clinical consequences.

In an open trial 11 in-patients with a major depressive episode (ICD-10), extensive metabolizers of mephenytoin (CYP2C19) and dextromethorphan (CYP2D6) and who were non-responders to a 3-wk pretreatment with 40 mg/d citalopram (Cit), were co-medicated for 7 wk (days 0-49) with fluoxetine (Fluox) (10 mg/d). Plasma concentrations of S-Cit and R-Cit significantly increased from day 0 (means+/-S.D.: 28+/-9 and 47+/-11 &mgr;g/l, respectively) to day 49 (58+/-12 and 72+/-21 &mgr;g/l, respectively) (p & 0.01 for each comparison), and the S-Cit/R-Cit ratio increased from 0.61+/-0.16 to 0.82+/-0.12 (p & 0.01). Therefore, Fluox increases the pharmacologically more active S-Cit (in comparison with R-Cit) with some stereoselectivity, most probably by inhibition of CYP2D6 and CYP3A4. Eight of the 11 patients showed clinical improvement (reduction > 50% of the MADRS score) and the combined treatment was generally well tolerated.

Polymorphism of dextromethorphan oxidation in South Indian subjects.

One hundred fifty-six unrelated healthy South Indian subjects were phenotyped according to their ability to metabolize dextromethorphan to its O-demethylated metabolite dextrorphan. Each volunteer was administered 25 mg oral dextromethorphan hydrobromide (19.3 mg dextromethorphan). Urine was collected during an 8-hour period after drug administration and was analyzed for dextromethorphan and dextrorphan by HPLC with fluorescence detection. This analysis was performed with and without previous deconjugation. The log10 (metabolic ratio), calculated as the ratio of dextromethorphan to dextrorphan, was bimodally distributed, and it was inferred that the frequency of occurrence of poor metabolizers of dextromethorphan in South Indian subjects is 3.2%. Phenotype assignment remained the same with both methods of analysis. Furthermore, a fairly good correlation (Spearman rank order correlation coefficient [r(s)] = 0.61; P < .0001) was observed between the log-transformed metabolic ratio derived from both methods.

P.1.021 Fluoxetine augmentation in citalopram non-responders: Pharmacokinetic and clinical consequences

A genetic deficiency in the metabolism of dextromethorphan and mephenytoin may be revealed by the excretion pattern of dextromethorphan and its metabolite dextrorphan, and mephenytoin, 4-OH-mephenytoin, respectively, after a single dose of the test drugs. Existing methods were modified for determining the compounds in 0.1–0.5 ml urine samples. No prior derivatization of the compounds was necessary before their gaschromatographic or mass-spectrometric analysis by using crosslinked 5% phenylmethyl silicone fused silica columns.Seven healthy volunteers were phenotyped at weekly intervals with either 25 mg dextromethorphan or 100 mg mephenytoin, or both drugs. One subject was a poor metabolizer of mephenytoin, while all subjects were extensive metabolizers of dextromethorphan. Neither a pharmacokinetic nor an analytical interference was observed when the results of the single test were compared with those of the combined test. The results of the mephenytoin test were also tentatively given in form of metabolic ratios. The GC-MS assay was designed for clinical studies so that patients treated with other drugs could be phenotyped.

Relationship between fluvoxamine pharmacokinetics and CYP2D6/CYP2C19 phenotype polymorphisms.

The purpose of this study was to investigate whether the disposition of fluvoxamine is associated with the CYP2D6 and CYP2C19 phenotype polymorphisms. The serum concentration of fluvoxamine was followed for 48 h after oral administration of a single dose of 50 mg fluvoxamine to five poor metabolizers of the CYP2D6 test drug dextromethorphan, five poor metabolizers of the CYP2C19 test drug mephenytoin, and five extensive metabolizers of both test drugs. Poor metabolizers of dextromethorphan had significantly higher areas under the serum concentration-time curve than extensive metabolizers of dextromethorphan (mean 1.31 vs 1.00 mumol.h.l-1). There were no differences between poor and extensive metabolizers of mephenytoin (mean, 1.00 vs 1.15 mumol.h.l-1). The results are consistent with a possible minor to moderate role of CYP2D6, but not CYP2C19, in fluvoxamine metabolism.

Single and multiple dose pharmacokinetics of nefazodone in subjects classified as extensive and poor metabolizers of dextromethorphan

Single and multiple dose pharmacokinetics of nefazodone in subjects classified as extensive and poor metabolizers of dextromethorphan

Racial differences in propranolol enantiomer kinetics following simultaneous i.v. and oral administration.

1. Racial differences in propranolol enantiomer kinetics following oral dosing were previously documented in our laboratory. The purpose of this study was to more completely describe propranolol kinetics in black and white subjects with the goal of gaining a better understanding of the mechanism(s) responsible for racial differences in oral propranolol kinetics. 2. Twelve white and 13 black healthy males were included in the study. Poor metabolizers of dextromethorphan and mephenytoin were excluded. Subjects took oral propranolol 80 mg every 8 h for 16 doses and received an intravenous dose of radiolabelled propranolol with the 16th dose. Serum and urine samples were collected for 24 h after the 16th dose. Serum concentrations of R- and S-propranolol and urine concentrations of its three primary metabolites were determined by h.p.l.c. 3. Apparent oral clearances of R- and S-propranolol were higher (P < 0.05) in blacks than whites (R-propranolol: 5036 +/- 4175 ml min-1 vs 2854 +/- 879 ml min-1; S-propranolol 3255 +/- 1723 ml min-1 vs 2125 +/- 510 ml min-1; blacks vs whites). 4. R- and S-propranolol clearances were higher in blacks than whites (R-propranolol 1069 +/- 316 ml min-1 vs 841 +/- 161 ml min-1; S-propranolol 947 +/- 271 ml min-1 vs 771 +/- 142 ml min-1; blacks vs whites, P < 0.05). 5. There were trends (P > 0.05 < 0.10) toward higher side chain oxidation, 4-hydroxylation and R-propranolol glucuronidation in blacks compared with whites. Ethnic differences in the enantiomeric ratios of partial metabolic clearance values were not observed. 6. We conclude the higher propranolol oral clearances in black subjects are explained by blacks having slightly higher hepatic metabolism via all three of its major metabolic pathways. Higher propranolol clearances among black subjects were also observed and we conclude this finding is explained largely by the higher hepatic metabolism, but also by slightly higher liver blood flow among black subjects.

CYP2D6 phenotype and the metabolism of nicotine and cotinine.

That CYP2D6 activity is an important determinant of nicotine metabolism and that people who have abnormal CYP2D6 genes are poor metabolizers of nicotine has been reported in the medical literature. The aim of this study was to assess the role of CYP2D6 in nicotine metabolism. Specifically, we compared the clearance of stable isotope-labelled nicotine and cotinine in 11 poor and 33 extensive metabolizers of dextromethorphan, a probe for CYP2D6 activity. The groups were matched by gender, age and ethnicity. Nicotine and cotinine kinetics were quite similar for cases and controls. These data suggest that CYP2D6 is not a major isozyme for the metabolism of nicotine or cotinine.

Assessment of individual CYP2D6 activity in extensive metabolizers with renal failure: Comparison of sparteine and dextromethorphan*

To examine whether the variability of CYP2D6 activity in patients with chronic renal failure can be assessed, particularly among subjects with the extensive metabolizer phenotype, by use of standard in vivo indexes of CYP2D6 activity derived from oral administration of dextromethorphan and sparteine. A single 100 mg oral dose of sparteine and a single 40 mg oral dose of dextromethorphan were administered on two occasions to 12 patients with chronic renal failure (creatinine clearance ranging from 20 to 70 ml/min) and 12 age- and sex-matched healthy subjects. Sparteine clearances, sparteine metabolic ratio, and urinary recovery of dextrorphan were calculated. Patients and healthy control subjects were not selected on the basis of their CYP2D6 phenotypes. Chronic renal failure was associated with a decrease in sparteine partial metabolic clearance to dehydrosparteine (median of 322 ml/min and range of 62 to 670 ml/min in patients with renal failure versus median of 635 ml/min and range of 77 to 1276 ml/min in normal subjects; p < 0.02). Sparteine apparent oral clearance (p < 0.03) and renal clearance (p < 0.001) decreased in patients with renal failure. However, sparteine metabolic ratio was not significantly altered in patients with renal failure and showed that all patients were extensive metabolizers of sparteine. Although fractional urinary excretion of dextrorphan decreased in patients with renal failure (median, 24.4%; range, 9.7% to 55.9%) compared with control (median, 47.5%; range, 24.1% to 72.1%) (p = 0.02), it also showed that all subjects were extensive metabolizers of dextromethorphan. The amount of dextromethorphan excreted in urine correlated with creatinine clearance independently from CYP2D6 activity measured as sparteine partial metabolic clearance. However, it did not correlate with sparteine metabolic ratio or with fractional urinary excretion of dehydrosparteine. Assessment of CYP2D6 activity by use of dextromethorphan and sparteine is possible in extensive metabolizer patients with chronic renal failure. However, in these subjects, dextromethorphan and sparteine do not reflect CYP2D6 activity in the same way.

Citalopram‐lithium combination treatment of elderly depressed patients: A pilot study

AbstractFourteen elderly depressive patients (age 67‐88 yr), phenotyped with dextromethorphan and mephenytoin before and during the trial, were treated for 4 weeks with citalopram (final dose 20‐30 mg/day, except one patient 60 mg/day). The clinical state of the patients was recorded weekly using the Hamilton Depression Rating Scale, the CGI (psychopathology) scale, the VAS and the UKU scale for side‐effects. As assessed by the Hamilton Depression Rating Scale, nine patients improved by more than 50% and continued with their citalopram treatment. The treatment of the five non‐responders was then continued for another 2 weeks by addition of lithium (target plasma levels 0.4‐0.8 mmol/l) to the ongoing citalopram medication. After 1 week, one patient had to be withdrawn for non‐response, three were responders, while the fifth patient was a responder only after 2 weeks of lithium addition. Due to side‐effects, the lithium dose had to be decreased in one patient who had responded to the combination therapy. Plasma levels of citalopram were within 145‐459 nmol/l after 4 weeks of citalopram treatment. All patients were extensive metabolizers of dextromethorphan, and all but possibly one also of mephenytoin. After 4 weeks of citalopram, there was a highly significant correlation between the ratios of S/R‐mephenytoin in urine and citalopram/desmethylcitalopram in plasma, which suggests a common mechanism in the metabolism of these drugs. The preliminary finding that a citalopram‐Li combination therapy may be useful in elderly depressive patients resistant to citalopram alone needs to be replicated by a controlled double‐blind study.

Pharmacokinetics of Desipramine Coadministered With Sertraline or Fluoxetine

The pharmacokinetic interactions of sertraline and fluoxetine with the tricyclic antidepressant desipramine were studied in 18 healthy male volunteers phenotyped as extensive metabolizers of dextromethorphan. Concentrations in plasma were determined after 7 days of desipramine (50 mg/day) dosing alone, during the 21 days of desipramine and selective serotonin reuptake inhibitor (SSRI) coadministration (fluoxetine, 20 mg/day; sertraline, 50 mg/day), and for 21 days of continued desipramine administration after SSRI discontinuation. Desipramine Cmax was increased 4.0-fold versus 31% and AUC0-24 was increased 4.8-fold versus 23% for fluoxetine versus sertraline, respectively, relative to baseline after 3 weeks of coadministration. Desipramine trough concentrations approached baseline within 1 week of sertraline discontinuation but remained elevated for the 3-week follow-up period after fluoxetine discontinuation. Concentrations of SSRIs and their metabolites correlated significantly with desipramine concentration changes (for fluoxetine/norfluoxetine, r = 0.94 to 0.96; p < 0.001; for sertraline/desmethylsertraline, r = 0.63; p < 0.01). Thus, sertraline had less pharmacokinetic interaction with desipramine than did fluoxetine at their respective, minimum, usually effective doses.

Variable disposition kinetics and electrocardiographic effects of flecainide during repeated dosing in humans: Contribution of genetic factors, dose-dependent clearance, and interaction with amiodarone

We studied the influence of cytochrome P450 2D6 (CYP2D6) on the steady-state disposition kinetics and the electrocardiographic effects of flecainide at two doses and during combination with amiodarone. Seven extensive and five poor metabolizers of dextromethorphan were studied during a three-period crossover study. All subjects received 50 mg flecainide every 12 hours, alone or together with 200 mg amiodarone every 12 hours, and 100 mg flecainide every 12 hours for 5 days. Mean steady-state plasma concentration of flecainide and QRS change from predrug value did not differ significantly among extensive and poor metabolizer subjects during each study period. Except for a shortened elimination half-life and nonlinear kinetics in extensive metabolizer subjects, phenotype had no significant influence on flecainide pharmacokinetics. Combination with amiodarone resulted in an increase in mean flecainide plasma concentration and effect in subjects with both phenotypes. Our findings indicate that CYP2D6 phenotype predicts flecainide nonlinear kinetics and flecainide half-life but has no influence on electrocardiographic effects during repeated administration of flecainide or on the extent of the amiodarone-flecainide interaction.

CYP2D6 phenotype determines the metabolic conversion of hydrocodone to hydromorphone

The contribution of cytochrome P450 2D6 (CYP2D6) to the formation of hydrocodone's active metabolite, hydromorphone, was examined in vitro and in vivo. Human liver microsomes prepared from an individual homozygous for the D6-B mutation of the CYP2D6 gene catalyzed this reaction at a negligible rate. Urinary metabolic ratios of hydrocodone/hydromorphone were highly correlated with O-demethylation ratios for dextromethorphan, an established marker drug of CYP2D6 activity (rs = 0.85; n = 18). The kinetics of hydrocodone after a single oral dose and its partial metabolic clearance to hydromorphone were investigated in five extensive metabolizers of dextromethorphan, six poor metabolizers, and four extensive metabolizers after pretreatment with quinidine, a selective inhibitor of CYP2D6 activity. The mean values for partial metabolic clearance by O-demethylation in the three groups were 28.1 +/- 10.3, 3.4 +/- 2.4, and 5.0 +/- 3.6 ml/hr/kg, respectively. No statistically significant phenotypic differences in physiologic measures were observed. However, over the first hour after dosing, the extensive metabolizers reported more "good opiate effects" and fewer "bad opiate effects" than poor metabolizers and extensive metabolizers in whom CYP2D6 was inhibited by quinidine. These data establish the importance of CYP2D6 in the formation of hydromorphone from hydrocodone and suggest that the activity of this enzyme may limit the abuse liability of hydrocodone.

Ethylmorphine O-deethylation cosegregates with the debrisoquin genetic metabolic polymorphism.

The single oral dose kinetics of ethylmorphine and its fractional metabolic clearance by O-dealkylation and N-dealkylation was investigated in five extensive and four poor metabolizers of dextromethorphan. In addition, the urinary metabolic ratios for these pathways (MRO and MRN, respectively) were investigated in a larger group of 27 extensive metabolizers and six poor metabolizers. The mean values for the fractional metabolic clearance by O-dealkylation and the MRO differed significantly between the poor metabolizers and extensive metabolizers without overlap between the values of either of these parameters in the two groups of subjects. In contrast, the corresponding parameters for the N-demethylation did not differ between poor metabolizers and extensive metabolizers. The area under the plasma concentration versus time curve was significantly higher (about three times higher) in the poor metabolizers compared with the extensive metabolizers (p = 0.004). Our data suggest that ethylmorphine is O-deethylated by the cytochrome P4502D6 isozyme inasmuch as both the fractional metabolic clearance by O-dealkylation and the MRO were found to cosegregate with the phenotype for the O-demethylation of dextromethorphan in our group of subjects.

Dextromethorphan and mephenytoin phenotyping of patients treated with thioridazine or amitriptyline.

The metabolism of most tricyclic antidepressants and some phenothiazine neuroleptics is under the genetic control of hepatic cytochrome P-450IID6, which also regulates the metabolism of dextromethorphan. This study investigated the effect of treatment with amitriptyline or thioridazine on testing for genetically regulated efficiency of the metabolism of dextromethorphan and mephenytoin. One group of 33 patients was treated with 150 mg amitriptyline a day (the AMI group); 25 other patients received a daily dose of thioridazine, either 200 mg (200-THD group; n = 7) or 400 mg (400-THD group; n = 18). Before and after 10 days of this treatment, all patients were tested with 25 mg dextromethorphan and 100 mg mephenytoin to determine their pharmacogenetic status with respect to their hepatic drug oxidizing systems (cytochrome P-450IID6 and P-450 MP). Two patients were poor metabolizers (PMs) of dextromethorphan and three of mephenytoin. Treatment with either psychotropic drug was without significant effect on the metabolism of mephenytoin, but both amitriptyline and thioridazine increased significantly the metabolic ratio of dextromethorphan/dextrorphan. Thioridazine had the effect of changing the pharmacogenetic status of 15 efficient metabolizers of dextromethorphan to poor metabolizers; amitriptyline did not have such an effect. There was no significant correlation between day-11 plasma levels of thioridazine, mesoridazine, or sulforidazine and the metabolism of dextromethorphan, but there was a correlation between the metabolism of dextromethorphan and plasma levels of amitriptyline and nortriptyline. Amitriptyline (p less than 0.05), but not thioridazine, decreases the ratio of conjugated/total dextrorphan in urine.(ABSTRACT TRUNCATED AT 250 WORDS)

Influence of amiodarone on genetically determined drug metabolism in humans

Amiodarone has been shown to interact with the nongenetically determined hepatic elimination of several drugs, including phenytoin and digoxin. Its influence on genetically determined metabolic pathways has not been studied in humans. We examined the effects of oral amiodarone therapy on the genetically determined metabolism of isoniazid (N-acetyltransferase), mephenytoin (cytochrome P450MEPH), and dextromethorphan (CYP2D6). Eight patients with arrhythmias were studied before and 76 +/- 16 days after amiodarone (loading dose of 1000 mg/day for 10 days followed by a maintenance dose of 200 to 400 mg/day). Genetically determined enzyme activity was assessed indirectly by calculating the metabolic ratio (parent drug/metabolite in 8-hour urine for CYP2D6 and P450MEPH and N-acetylisoniazid/isoniazid in plasma for N-acetyltransferase) after oral administration of the parent compounds. At the time of phenotyping, plasma concentrations of amiodarone and N-desethylamiodarone were 0.66 +/- 0.35 micrograms/ml and 0.65 +/- 0.26 micrograms/ml, respectively. Amiodarone increased the log(metabolic ratio) of dextromethorphan from a median of -2.5 (range, -2.9 to -2.0) to a median of -1.9 (range, -2.5 to -1.5; p less than 0.02) but did not alter the metabolic ratio of mephenytoin or isoniazid. The amount of dextromethorphan excreted in urine increased from a median of 0.084 mumol/8 hours (range, 0.041 to 0.161 mumol/8 hours) to a median of 0.205 mumol/8 hours (range, 0.064 to 0.288 mumol/8 hours; p less than 0.02) and the amount of its metabolite (dextrorphan) tended to decrease from a median of 26 mumol/8 hours (range, 15 to 37 mumol/8 hours) to a median of 20 mumol/8 hours (range, 7 to 27 mumol/8 hours; p less than 0.09).(ABSTRACT TRUNCATED AT 250 WORDS)