VoL 267, No. 3 Printed in U.SA.
ABBREVIATiONS: EHC, enterohepatic circulation; neo/chol, neomycin and cholestyramine; CIU, Clinical Investigation Unit; AUC, area under the curve of lorazepam; AUCG, area under the curve of lorazepam glucuronide.
1034
THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS
Copyright 0 1993 byThe American Society for Pharmacology and Experimental Therapeutics
Multiple-Dose Lorazepam Kinetics: Shuttling of Lorazepam Glucuronide between the Circulation and the Gut during Day- and Night-Time Dosing Intervals in Response to Feeding
A. CHAUDHARY, R. A. LANE, D. WOO and R. J. HERMAN
Departments of Medicine and Pharmacology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada Accepted for publication August 2, 1993
ABSTRACT
Lorazepam kinetics were examined in seven healthy males age 1 8 to 30 years after single- and multiple-dose lorazepam admin- istration and in the presence and absence of neomycin and cholestyramine to block the enterohepatic circulation of the drug.
Methods used a simultaneous i.v./p.o. dosing regimen with provision to measure lorazepam clearance dunng day- and night- time dosing intervals. The day-time steady-state clearance of free lorazepam measured 7.55 ± 1 .95 mI/mm/kg (mean ± S.D.) and was identical to that observed after single-dose administra- tion (7.68 ± 3.1 9 mI/mm/kg). Neomycin and cholestyramine increased lorazepam clearances 5 to 45% (P .05) as would be expected for interruption of an enterohepatic circulation and in keeping with previous observations under nonsteady-state con-
ditions. Lorazepam clearances were the same during the day as during the night, except in the presence of neomycin and choles- tyramine, where night-time clearances were significantly greater
(10.16
± 3.52 vs. 8.77 ± 2.43 mI/mm/kg, P .05). Urinary recoveries of lorazepam glucuronide, on the other hand, were greater during the day than during the night (1 1 4 ± 1 1 vs. 77 ± 1 5%, P .05) and in all cases were greater than 1 00% of the administered dose for that interval. Thus, there is a diurnal variation in lorazepam elimination consistent with a fasting-in- duced increase in hepatic glucuronidation during the night. This, combined with the relative inactivity of the gut dunng this period, serves to trap the glucuronide and delay its transfer back to the systemic circulation and urine.Lorazepam is used widely as a sedative and antianxiety agent.
It is also of pharmacokinetic interest, because its elimination
occurs predominantly through hepatic conjugation to the 3-0-
phenolic glucuronide (Greenblatt et al., 1976; Ruelius, 1978) and, thus, its clearance is considered to be an indirect measure
of hepatic glucuronidation (Wilkinson and Shand, 1975; Crome et al., 1987). Recently we have shown that lorazepam also
participates in an EHC in humans (Herman et at., 1989), and
that this cycling must be interrupted in order to estimate the true hepatic clearance of the drug. Treatment with neo/chol has been successful in interrupting lorazepam EHC (Herman and Chaudhary, 1991) and may restore the ability of lorazepam
clearance to profile intrinsic glucuronidation activity.
There are no studies examining the effects of multiple dosing on EHC. Indeed, steady state may be an optimal situation in
which to study EHC, because the changes in clearance and
metabolite profiles that accompany cycling would be com-
pressed within a single, easily measured, dosing interval. More-
over, enteric activity is phasic in humans due to the cyclic
nature of wakefulness and food intake and, accordingly, there
Received for publication November 16, 1992.
may be differences in EHC between day- and night-time pe-
nods. Thus, we were interested in examining the pharmacoki-
netics of lorazepam after single- and multiple-dose administra-
tion, and examining consecutive day/night dosing intervals for possible differences in glucuronidation and EHC.
Methods
Subjects. Seven subjects participated in the study. All were non- smoking, drug-free, healthy males between the ages 18 and 30 years.
The average age was 21.4 ± 2.3 years and their mean weight 76.7 ± 14.0 kg. A complete medical evaluation including a history, physical examination and routine screening of hematologic and biochemical markers of general metabolic, renal and hepatic function was con-
ducted, and each volunteer was asked to sign a consent as approved by
the Ethics Committee at our institution. Subjects were required to
abstain from drugs and alcohol at least 1 week before and throughout the study period, which was conducted in three phases.
Study Design
Single-dose i.v./p.o. lorazepam study. Subjects presented them-
selves to the CIU at 7:30 A.M. after a 12-hr overnight fast. An i.v.
infusion consisting of 0.9% saline was established in each arm; one to administer the test drug and the other to draw samples. After collection
at ASPET Journals on May 8, 2016 jpet.aspetjournals.org Downloaded from
of appropriate blanks and at precisely 8:00 A.M., 50 tCi of [“C]
lorazepam (specific activity, 52 mCi/mmol, Amersham Corp., Arlington
Heights, IL) was given by rapid i.v. injection together with a 1-mg tablet of nonlabelled lorazepam by the p.o. route (Wyeth Laboratories, Philadelphia, PA). Blood and urine samples were collected at frequent intervals over the ensuing 60 hr. Subjects remained fasting until the 4- hr sample (12:00 P.M.), at which time they were given a low fat meal.
A second low fat meal was provided at approximately 4:00 P.M. After
collection of the 12-hr sample, the i.v. cannulae were removed and
subjects were allowed to go home returning periodically thereafter for
collection of further blood and urine samples.
Steady-state lorazepam cycling-intact study. Upon completion of the first part of the study, subjects were started on a multiple-dose regimen consisting of lorazepam, 1 mg p.o. every 12 hr (timed to 8:00 A.M. and 8:00 P.M. by using individual alarm watches). Predose A.M.
blood samples were taken over the next 10 days to monitor the approach to steady state. On the evening of the 11th treatment day, subjects
were again admitted to the CIU. An overnight 8:00 P.M. to 8:00 AM.
dosing-interval was defined for the p.o., nonlabelled, lorazepam dose
by multiple blood and urine sampling. The following morning the i.v./
P.O. lorazepam study was repeated, except that the sampling was limited to the 8:00 AM. to 8:00 P.M. dosing interval for the p.o. dose, but continued up to 60 hr for the i.v. dose.
Steady-state lorazepam cycling-interrupted study. After corn-
pletion of the above study and while continuing lorazepam p.o. dosing, patients were started on neornycin, 1 g every 6 hr, and cholestyramine, 4 g every 4 hr, both by mouth, to interrupt the EHC of lorazepam.
Predose AM. blood was sampled for a further 5 days. Subjects were
then readinitted for a second steady-state, cycling-interrupted braze- pam study (similar to the previous study), with blood and urine samples
collected over sequential night- and day-time dosing intervals. Neo/
chol were continued together throughout the entire 60-hr sampling interval.
Measurement. Lorazepam was measured in plasma and urine by
solvent extraction and high-performance liquid chromatography ac- cording to previously published methods (Herman et aL, 1989). Briefly, 100 ng of desmethyldiazepam (internal standard) was added to 0.1 to 1.0 ml ofplasrna or urine, and this was alkalinized with an equal volume of saturated sodium borate solution. Samples were twice extracted in glass with freshly distilled diethyl ether, dried with N2 at 37’C and reconstituted in mobile phase before high-performance liquid chroma- tography injection. The mobile phase consisted of a mixture of 0.1%
sodium phosphate buffer, pH 3.0, and acetonitrile, 70:30 by volume, delivered at a flow of 2.2 mb/mm. The stationary phase was a 3.9 x 150 mm Waters NovaPak C15 column. Detection was by UV absorption
spectroscopy with the instrument set to 230 nm, the absorption maxi-
mum for borazepam. Standard curves were linear over the concentration range of 5 to 100 ng/ml, with a limit of detection of 3 ng/ml and an interassay coefficient of variation of 2% at 75 and 8% at 5 ng/ml, respectively. Quality controls were included with each assay run.
Lorazepam glucuronide was measured as borazepam equivalents ob-
tamed after overnight hydrolysis at 37’C with 3-glucuronidase/sulfat- ase (from Helix pomatia, Sigma Chemical Co., St. Louis, MO). Loraze-
pam unbound fraction was determined in a prestudy blank plasma
sample and in plasma obtained 0.5, 1, 2, 3 and 4 hr after each of the two meals. Measurement was by equilibrium dialysis of 1 ml of plasma spiked with 5 x 10 MCi of [‘4C]borazepam against an equal volume of isotonic phosphate buffer (pH 7.4) for 4 hr (the time to equilibrium) at 37#{176}C.Radioactivity was measured by liquid scintillation counting with automatic quench compensation (Beckman Instruments Inc., Missis- sauga, Ontario, Canada).
Analysis. Bi- and triexponential equations were fit to the plasma
concentration-time profiles by using iterative, weighted, nonlinear least-squares regression (SAAM 23, Resource Faculty for Kinetic
Analysis, Seattle, WA). The Akaike information criterion (Yamaoka et
aL, 1978) and F test (Boxenbaum et at., 1974) were then applied to
establish the best statistical fit. The AUC from zero to infinity was
determined by the trapezoidal rule to peak concentrations and log
trapezoidal rule after peak concentrations (Yeh and Kwan, 1978) and
using the estimated last concentration-time point and fractional elim-
ination rate constant from the fitting analyses. Steady-state AUCs
were truncated at the end-of-dosing interval concentration-time point.
Clearance was calculated as dose/AUC and apparent volume of distri-
bution at steady state by analogous statistical moment methods (Gi-
baldi and Perrier, 1982). Pharmacokinetic terms, where appropriate,
were corrected for weight and unbound borazepam fraction.
Statistics. Statistical comparisons were made by analysis of vari- ance by using the BMDP Statistical Package (University of California Press, Los Angeles, CA) with P < .05 for acceptance of significance.
Where significant differences were found, individual means were corn- pared by the least significant difference method (Daniel, 1987).
Results
The mean pharmacokinetic parameters for single-dose i.v.
borazepam were: free systemic clearance, 7.68 ± 3.19 ml/min/
kg; free volume of distribution, 15.43 ± 2.20 liters/kg; half-life, 27.7 ± 6.5 hr; and free fraction, 0.09 ± 0.01. The kinetics of i.v.
borazepam were the same in the single- and multiple-dose
cycling-intact (clearance, 6.78 ± 1.84 mi/mm/kg; volume of
distribution, 14.74 ± 1.54 liters/kg; and half-life, 27.3 ± 6.5 hr) and cycling-interrupted (clearance, 6.29 ± 1.42 mb/mm/kg;
volume of distribution, 13.85 ± 1.77 liters/kg; and half-life,
27.8 ± 5.3 hr) parts of the study. Free-fraction of lorazepam was also identical before and after each of the two meals in all subjects and in all parts of the study (table 1). This points to
the high degree of reproducibility of these parameters and
suggests there was no period effect between the different phases of the study.
Average predose concentrations of borazepam on days 10, 11
and 12 (before the start of the cycling-intact part of the study) and days 17, 18 and 19 (before the start of the cycling-inter- rupted part of the study) were 22.06 ± 6.77 and 19.17 ± 6.44 ng/ml, respectively. Concentrations did not differ within each
of the consecutive 3-day intervals (P .45), indicating that
steady state had been established. However, concentrations before the onset of the cycling-interrupted study were 6 to 27%
lower than those obtained during the cycling-intact phase of
the study (P = .01) suggesting a new, lower steady state had
been achieved. Figure 1 shows the concentration-time profiles
of borazepam over a night- and day-time dosing interval in a
single individual in the cycling-intact study. As can be seen, there were no large postprandial peaks and no differences
between consecutive day/night concentration profiles.
The mean free steady-state clearances of p.o. borazepam before and after treatment with neo/chol during day- and night-
time dosing intervals are shown in table 2. There were no
significant differences between the day- and night-time clear-
TABLE 1
Free fraction of lorazepam (%) before and after meals during single- and multiple-dose cycling-intact and cycling-interrupted studies
Data are mean ± S.D. No signfficant differences were observed between any measurements inany of the studies.
Multiple-Dose Studies Single-Dose Study
Cycling-intact Cycling-interrupted
Premeal 9.83 ± 0.69 9.68 ± 0.61 10.43 ± 0.77
0.5 hr Postmeal 9.71 ± 0.42 9.85 ± 0.39 9.58 ± 0.83
1 hr Postmeal 10.26 ± 0.54 10.31 ± 0.30 10.76 ± 0.82
2 hr Postmeal 10.03 ± 0.69 9.96 ± 0.51 10.81 ± 0.80
3 hr Postmeal 9.90 ± 0.64 10.06 ± 0.58 10.64 ± 0.76
at ASPET Journals on May 8, 2016 jpet.aspetjournals.org Downloaded from
Meal
E
S
:!
‘5 S
Sa
U
S 20
‘5
15
‘5
0
.
PM8 PM
*5
-=-
at P s .05, cycling-interrupted vs. cycling-intact; 5gnif)t at P .05, 8:00 P.M. to 8:00 AM. VS. 8:00AM. to 8:00 P.M.
+48% +30%
. S S
120’
S
100’
50 0
Sb SO
0 .5
E.E
:
20
!
Cycling-intact--..- _....1 8PM
Cycling-interrupted
+24% +19%
S S
E
‘5
I
0I
0,5
.55
0
‘5 500
600
400
200
0
5-’- 8PM
Cycling-intact
8 - ---1
Cycling-interrupted fl\\t\Meall
8AM Time (hours)
Fig. 1. Plasma concentration-time profile of a single patient at steady state during the night- and day-time dosing intervals. Timing of meals is indated by the arrows.
TABLE 2
Free steady-state clearance (milliliters per minute per kilogram) of P.O. Ioraz.pam during day- and night-time dosing Intervals
Data are mean ± S.D.
8:00 P.M. tO8:00 AM. 8:00 AM. to 8:00 P.M.
Cycling-intact 8.1 0 ± 1 .99 7.55 ± 1.95
Cycling-interrupted 10.16 ± 3.52 8.77 ± 2.43*
Fig. 2. Percentage of the dose recovered in urine as lorazepam gluc- uronide during the cyding.intact and cycling-interrupted phases of the study. 0, day-time dosing interval; , the corresponding night-time dosing interval. S.E. is denoted above, along with the mean difference day-time vs. night-time. eSgnif at P .05, 8:00 P.M. to 8:00 A.M.
vs. 8:00 AM. to 8:00 p..; e*5jgnif. at P .05, cycling-interrupted vs. cycling-intact.
ances during the cycling-intact part of the study. Neo/chol
produced a 29 ± 15% increase in the apparent clearance of
lorazepam (P .05) consistent with interruption of an EHC.
However, night-time lorazepam clearances in the presence of neo/chol were 15 ± 22% greater than day-time clearances (P .05), so that the effect of interruption of EHC on lorazepam AUCs was significantly greater during the night than during the day.
Virtually ail of the drug administered over the two consecu- tive dosing intervals in the cycling-intact study was recovered in the urine as lorazepam glucuronide (1916 ± 216 g of a total
of 2000 gig). However, day-time urinary recoveries were con-
sistently greater than night-time urinary recoveries (1142 ± 112 g us. 774 ± 149 g, P .05) and, in all cases, were greater than 100% of the administered dose for that interval (fig. 2).
After administration of neo/chol, 80% of lorazepam (1620 ±
172 g of a total 2000 zg) was recovered in the urine. Day-time urinary recoveries were, again, consistently greater than night-
time urinary recoveries (913 ± 103 ig vs. 706 ± 118 ig, P
.05) despite administration of neo/chol. However, only day-
time urinary recoveries showed a significant reduction with the
neo/chol, as night-time recoveries were the same, cycling-intact Us. cycling-interrupted. Renal clearances of lorazepam gluc- uronide were also the same day and night and cycling-intact and cycling-interrupted, but AUCGs corrected for dose were 10 to 40% greater during the day than during the night for both cycling-intact (73 ± 15 vs. 59 ± 9%, P .05) and cycling- interrupted parts of the study (56 ± 10 vs. 47 ± 9%, P .05), and only day-time measurements were significantly reduced by neo/chol (fig. 3).
Discussion
These results show that the pharmacokinetics of lorazepam
are the same after multiple-dosing as after single-dose admin-
istration of the drug. This is consistent with former studies on
the disposition oflorazepam (Greenblatt et at., 1977, 1979), and the known linear-type pharmacokinetics possessed by the drug.
We hypothesized that if lorazepam enterocycled there may be an opportunity for accumulation within the enteric loop on multiple-dosing, and that this may produce postabsorptive
peaks and a larger effect of interruption of EHC. Large post-
prandial peaks have been described for other benzodiazepines
(Morselli, 1977), and these attributed to EHC or feeding-
induced changes in binding and distribution (Korttila and Kangas, 1977; Naranjo et at., 1980). However, the fact that no
such peaks were observed and that lorazepam did not accu-
mulate on multiple-dosing casts doubt on the mechanism of
EHC-induced postprandial peaking. Also, there is no evidence that lorazepam free fraction is altered after ingestion of a standard meal.
The apparent free oral clearance, free volume of distribution and half-life of lorazepam were virtually the same during the night as during the day. This is also consistent with former
observations with lorazepam (Bruguerolle et at., 1985). How-
ever, lorazepam clearances in the presence of neo/chol were significantly greater during the night, and plasma concentra-
tions and urinary recoveries of lorazepam glucuronide were
significantly greater during the day than during the night-time
Fig. 3. AUCG during the cycling-intact and cycling-interrupted phases of the study. 0, day-time dosing interval; B, the corresponding night-time dosing interval. S.E. is denoted above, along with the mean difference day-time vs. night-time. *gnjf at P .05, 8:00 P.M. to 8:00 AM.
vs. 8:00 A.M. to 8:00 p..; at P .05, cycling-interrupted vs. cycling-intact.
at ASPET Journals on May 8, 2016 jpet.aspetjournals.org Downloaded from
dosing interval. Thus, there is a diurnal variation in lorazepam elimination and in the disposition of its glucuronide. Given that cycling-interrupted lorazepam clearances are likely to pro- file intrinsic rates of hepatic glucuronidation, the latter would appear to be greater during the night. Indeed, this is consistent
with our former observations on the effects of fasting on
lorazepam elimination (Chaudhary et at., 1989) and the effects of diet on other drugs known to undergo glucuronidation by the liver (Daneshmend and Roberts, 1982; Liedholm and Me- lander, 1986). However, the fact that the AUCGs and amount of glucuronide recovered in urine decreased during the night rather than increased as would be expected with an increased metabolite formation, raises the question of where the excess glucuronide formed during the night may have gone.
One of the possibilities is that clearances are not increased, but that neo/chol interfered with the absorption of lorazepam.
This would produce smaller AUCs and AUCGs as was observed in this study, and if the effect was somehow greater during the night than during the day, a reciprocal efficiency in lorazepam absorption could result (Bruguerolle et at., 1985). Neo/chol is known to slow the rate but not the extent of lorazepam absorp- tion (Herman et al, 1989) and, indeed, peak concentrations were lower (22.8 ± 8.7 vs. 30.8 ± 7.2 ng/ml during the night and 34.1 ± 12.9 us. 37.2 ± 3.2 ng/ml during the day, P = not
significant for both) in the presence of neo/chol and time to
peak concentrations were about 1 hr delayed (1.9 ± 1.1 vs. 1.0
± 0.4 hr, P .05), night vs. day and a further hour delayed (3.1
± 0.7 vs. 1.9 ± 1.1 hr, P .05), neo/chol us. control. However, lorazepam is generally well absorbed and it is doubtful whether such minor differences could account for the large changes in glucuronides. Secondly, discrepancies between night- and day- time AUCGs and glucuronide recoveries were actually greater in the control part of the study than in the presence of the interrupting agents suggesting that neo/chol was not a causa- tive factor and, indeed, diminished the extent of the underlying abnormality. Finally, the incomplete urinary recoveries of br-
azepam glucuronide in the presence of neo/chol (80%) mirrors
exactly the extent of change observed in night-time clearances and, therefore, is accounted for through interruption of EHC.
Thus, we do not consider this to be a valid explanation for the differences in day/night borazepam disposition.
The other possibility is that there is a complex exchange of
borazepam and borazepam glucuronide between the gut, the
urine and the systemic circulation. After metabolism, part of
the borazepam conjugate is secreted into bile and from there passes into the intestines, where it is eventually hydrolyzed back to borazepam and reabsorbed through the portal circula-
tion. Another part is returned directly to the systemic circula-
tion and excreted into urine (Gessner and Hamada, 1974). How much is diverted into bile and how much into blood and urine
is not known, but judging from the high urinary recoveries of
lorazepam glucuronide observed in humans, biliary secretion is generally considered to be low (Greenblatt, 1981; Greenblatt and Engelking, 1988). It should be noted, however, that the portion of the drug or metabolite which enters the enteric circuit is lost to detection by standard methodology. Because the discrepancies in AUCGs and urinary recoveries are both
unmeasured and amenable to alteration by neo/chol, and the
site of this batter interaction is known to be the glucuronide in the gut (Herman and Chaudhary, 1991), the excess borazepam glucuronide formed during the night-time dosing interval must be sequestered in the gut. Lorazepam and its glucuronide re-
mains in the gall bladder and intestines due to the low intrinsic
activity of this organ during the night, and then must be
returned to the systemic circulation and urine resulting in
greater than expected recoveries during the subsequent day-
time dosing interval. Therefore, we propose that there is a
shuttling of drug and metabolite between the gut and systemic circulation and urine in response to phasic changes in feeding and enteric activity.
To our knowledge, this is the first description of such a mechanism. Its presence is important, because it confirms that lorazepam is enterocycing in humans and that the nature of
EHC is to trap the drug and/or metabolite and delay its
eventual elimination from the body (Herman, 1990). Thus,
bibiary secretion of borazepam glucuronide is significant (ac-
cording to interruption experiments, at least 25-50% of the
dose passes via this route), and EHC is highly efficient in
recovering most of the glucuronide transferred to the bowel.
Indeed, renal glucuronide clearance is probably the only irre- versible pathway of borazepam elimination. EHC, of itself, is
not responsible for the shuttling mechanism. Differences in
lorazepam clearance are maximized when EHC is blocked, and interrupting EHC does not obliterate the day/night differences
in glucuronide AUCGs and urinary recoveries. Moreover, in-
creases in the rate of hepatic glucuronidation, as occurs during
the night, would not be expected to change the relative propor-
tions of drug diverted into the gut or into the circulation and
urine. Thus, the fact that we see a day/night shuttle at all is
evidence that there is a concomitant increase in the movement
of borazepam glucuronide from the hepatocyte into the bile
relative to its movement into the circulation. Therefore, our
study shows both an increase in borazepam glucuronidation and
an increase in borazepam glucuronide shuttling into the gut
with resultant EHC during the night. We do not know what
factors regulate these changes, although preliminary evidence
suggests it is related to feeding (Chaudhary et aL, 1989) and
that insulin or glucagon is involved.
Many drugs are metabolized or their metabolites secondarily
conjugated to a glucuronide. All would be expected to partici-
pate in the shuttling mechanism and their effects, therapeutic or toxic, to be potentially subject to diurnal variations. The gut
is also an organ in which important interactions may occur
between an enterocyding drug and the diet or other enterically
administered agents. Finally, it should be emphasized that the
effects of EHC are to offset the primary influences of hepatic glucuronidation (i.e., an increased glucuronidation associated
with an increased EHC may result in no change in overall
elimination). Thus, many drugs may be cycling to a greater
extent than is presently appreciated, and the absence of an effect of disease or state on the apparent disposition of a drug
does not preclude the presence of complimentary changes in
hepatic glucuronidation and EHC.
References
BOXENBAUM, H. G., RIEGELMAN, S. AND ELA5HOFF, R. M.: Statistical estima- tions in pharmacokinetics. J. Pharmacokinet. Biopharm. 2: 123-148, 1974.
BRUGUEROLLE, B., BOUVENOT, G., BARTOLIN, R. AND DEscoms, C.: Temporal
variations of lorazepam pharmacokinetics. mt. J. Clin. Pharmacol. Ther.
Toxicol. 23: 352-354, 1985.
CHAUDHARY, A., SzAiics, C. B. N. AND HERMAN, R. J.: Lorazepam disposition in Gilbert’s syndrome: Effect of feeding and fasting. Chin. Pharmacol. Ther.
45: 166, 1989.
CROME, W. R., WEBsTER, S. L., BoBo, L., TERE5I, M. E., RELLING, M. V. AND EVANS, W. E.: Simultaneous administration of multiple model substrates to assess hepatic drug clearance. Cliii. Pharmacol. Ther. 41: 645-50, 1987.
DANESHMEND, T. K. AND ROBERTS, C. J. C.: The influence of food on the oral
at ASPET Journals on May 8, 2016 jpet.aspetjournals.org Downloaded from
and intravenous pharmacokinetics of a high clearance drug A study with labetalol. Br. J. Clin. Pharmacol. 14: 73-78, 1982.
DANIEL, W. W.: Biostatistics, a Foundation for Analysis in the Health Sciences, 4th at, John Wiley and Sons, New York, 1987.
GES8NER, T. AND HAMADA, N.: High biliary and systemic excretion of a gluc-
uronide during its hepatic synthesis. Life Sci. 15: 83-94, 1974.
GIwDI, M. AND PERRIER, D.: Noncompartinental analysis based on statistical moment theory. In Pharmacokinetics, 2nd ad., pp. 409-416, Marcel Dekker, Inc., New York, 1982.
GREENBi.a’rr, D. J.: Clinical pharmacokinetics ofoxazepam and lorazepam. Clin.
Pharmacokinet. 6: 89-105, 1981.
GREENBLATT, D. J., ALLEN, M. D., MAcLaUGHLIN, D. S., HUFFMAN, D. H.,
HARMATZ, J. S. AND SHADER, R. I.: Single- and multiple-dose kinetics of oral lorazepani in humans: The predictability of accumulation. J. Pharmacokinet.
Biophe.rin. 7: 159-179, 1979.
GREENBLATF, D. J. AND ENGELKING, L. R.: Enterohepatic circulation of braze- pam and acetaminophen conjugates in ponies. J. Pharmacob. Exp. Ther. 244:
674-679, 1988.
GREENBLATF, D. J., KNOWLES, J. A., COMER, W. H., SHADER, R I., HARMATZ,
J. S. AND RuEuus, H. W.: Clinical pharinacokinetics ofborazepam. IV. Long- term oral administration. J. Clin. Pharmacob. 17: 495-500, 1977.
GREENBLAVI, D. J., SCHILLINGS, R. T., KyiuAicopouLous, A. A., SH.Dn, R.
I., SISENWINE, S. F., KNOWLES, J. A., RuEuus, H. W.: Clinical pharmacoki- netics ofborazepam. I. Absorption and disposition oforal 14C-borazepam. Clin.
PharinacoL Ther. 20: 329-341, 1976.
HExtMN, R. J.: The enterohepatic cycle. Clin. Pharmacol. Ther. 47: 160, 1990.
HERMAN, R. J. AND CHAUDHARY, A.:In vitro binding ofborazepam and borazepam glucuronide to cholestyraniine, cobestipob, and activated charcoal. Pharmacob.
Rag. 8: 538-540, 1991.
Vol. 267
HERMAN, R. J., VAN PHAM, J. D. AND SzKcs, C. B. N.: Disposition of
borazepam in human beings: Enterohepatic recirculation and first-pass effect.
Cliii. Pharmacob. Ther. 46: 18-25, 1989.
KORTTILA, K. AND KANGAS, L.: Unchanged protain binding and the increase of
serum diazepain bevels after food intake. Acta Pharmacol. Toxicob. 40: 241- 246, 1977.
LIEDHOLM, H. AND MELANDER, A.: Concomitant food intake can increase the bioavailability of propranobob by transient inhibition of its presystemic primary conjugation. Clin. Pharmacol. Ther. 40: 29-36, 1986.
NARANJO, C. A., SELLERS, E. M., GILE5, H. G. AND ABEL, J. G.: Diurnal
variations in plasma diazepam concentrations associated with reciprocal changes in free fraction. Br. J. Clin. Pharmacob. 9: 265-272, 1980.
MORSELLI, P. L.: Psychotropic drugs. In Drug Disposition during Development, pp. 431-474, Spectrum Publishers, Inc., New York, 1977.
RuEuus, H. W.: Comparative metabolism of borazepam in man and four animal species. J. Clin. Psychiatry 39: 11-15, 1978.
WILKINsoN, G. R. AND SHAND D. G.: A physiological approach to hepatic drug clearance. Clin. Pharmacol. Ther. 18: 377-390, 1975.
YAMAOKA, K., NAKAGAWA, T. AND UNO, T.: Application of Akaike’s information criterion (AbC) in the evaluation of linear pharmacokinetic equations. J.
Pharmacokinet. Biopharin. 6: 165-175, 1978.
YEH, K. C.AND KWAN, K. C.: Acomparison of numerical integrating algorithms by trapezoidal, Lagrange, and spline approximation. J. Pharmacokinet. Bio- pharin. 6: 79-98, 1978.
Send reprint requests to: Dr. Robert J. Herman, FRCPC, Department of
Pharmacology, College of Medicine, University of Saskatchewan, Saskatoon,
Saskathewan, Canada S7N OWO.
