BIOPHARMACEUTICS & DRUG DISPOSITION, VOL. 13, 693-701 (1992)
HUMAN DOLASETRON PHARMACOKINETICS: I. DISPOSITION FOLLOWING SINGLE-DOSE INTRAVENOUS ADMINISTRATION TO NORMAL MALE SUBJECTS HAROLD BOXENBAUM*, TODD GILLESPIE', KATHLEEN HECK*and WILLIAM HAHNEg Marion Merrell Dow Inc., P.O. Box 9627, Kansas City, MO 64134, U.S.A.
ABSTRACT Dolasetron is a 5-hydroxytryptamine antagonist active at type 111 receptors; it is presently undergoing clinical evaluation for the reduction/prevention of cancer chemotherapyinduced nausea and vomiting. Following intravenous administration to healthy male subjects of doses ranging from 0.6 to 5 mg kg-I, dolasetron disappeared extremely rapidly from plasma; concentrations were generally measurable for only 2-4 h. Less than 1 per cent of the dose was excreted intact in urine. A major plasma metabolite, reduced dolasetron, peaked rapidly at approximately 0.625 h (median). Its median terminal disposition half-life was 7 - 5 6 h; median values for fraction of dose excreted in urine and renal clearance were 3 1 -0per cent and 2.68 ml min-' kg-I, respectively. Over the dose-range covered, pharmacokinetics of both dolasetron and reduced metabolite appeared to be independent of dose. The median ratio of the areas under the plasma concentration-time curves for metabolite relative to dolasetron was 11 -9. As a result of its activity and significant plasma concentrations, reduced dolasetron may play a significant role in pharmacodynamic activity. KEY WORDS
Dolasetron Pharmacokinetics Intravenous dosing
INTRODUCTION Employing an isolated rabbit heart preparation,' dolasetron was found to be a specific and selective 5-hydroxytryptamine antagonist active at type I11 receptors. It is presently undergoing clinical investigation for the reduction/ prevention of cancer chemotherapy-induced nausea and emesis. The intravenous formulation is prepared using the monohydrated mesylate (methane sulfonate) salt (see Figure 1).
Present addresses: 'Clinical Pharmacokinetics, Wyeth-Ayerst Research, P.O. Box 8299, Phila., PA 19101, U.S.A. 'Lilly Clinic, Wishard Memorial Hospital, 1001 W. 10th St., Indianapolis, IN 46202, U.S.A. *F.A.C.T., 85 N.E.Loop 410, Suite 612, San Antonio, TX 78216, U.S.A. §Addressee for correspondence.
0142-2782/92/09O693-09$09.50 0 1992 by John Wiley & Sons, Ltd.
Received 6 January 1992 Accepted 20 April 1992
694
H. BOXENBAUM ET AL.
*
CH~SOSH* H20
0
OH
0
0
Figure 1 . Chemical structures of: (a) dolasetron mesylate monohydrate, used in preparation of the injectable dosage form; (b) dolasetronfree base; and (c) reduced dolasetron, a major plasma metabolite in humans
A major metabolite of dolasetron identified in human plasma is the reduced compound (Figure 1). Whereas dolasetron is achiral, keto-reduction produces a chiral carbon center. Both intact drug and a racemic mixture of the reduced metabolite have been evaluated for their affinities to 5-HT3 receptors expressed by a neuroblastoma cell line.2.3Reduced dolasetron is approximately 100 times more potent than dolasetron. The purpose of this initial pharmacokinetic investigation in humans was to characterize the pharmacokinetics of dolasetron following single-dose, intravenous administration of escalating doses. Assays were available to measure intact drug in plasma and urine, as well as the combined sum of ( +) and ( - ) reduced dolasetron in these same biological fluids. It is anticipated that future pharmacokinetic and related studies will focus on such things as plasma protein binding, multiple intravenous dosing, single and multiple oral dosing, influence of disease states, pharmacodynamics, metabolite identification/kinetics, chirality metabolite issues, etc.
HUMAN DOLASETRON PHARMACOKINETICS
695
METHODS
Study procedures Thirty-two, healthy male volunteers of mixed race and ethnicity participated in this study; all had provided informed consent. Subjects were between the ages of 19 and 38 years, were in good general health (as determined by physical examination and laboratory tests), were within 10 per cent of their ideal body weight, did not smoke tobacco, and were drug-free. Following a minimum 12 h food fast and 2 h water fast, subjects received dolasetron intravenously at approximately 8 am. After dosing, water was permitted after 2 h and food after 4 h. Doses were administered as dolasetron mesylate monohydrate, contained within approximately 100 ml of sterile solution. The drug was administered intravenously at a constant rate over a period of 1Omin. Four subjects each received nominal doses of 0-6, 1.25,2.0, 2.5, 4-0, and 5.Omgkg-l; eight subjects received doses of 3.Omgkg-l (expressed as dolasetron mesylate monohydrate). Fifteen millilitre blood samples (containing EDTA as anticoagulant) were collected at approximately 0, 0.25, 0.50, 0.75, 1, 2, 4, 6, 8, 12, 24, 36, and 48 h after the start of the infusion. The blood was centrifuged, plasma harvested, and this was subsequently stored frozen (- 15 "C) in glass containers until analysis. Urine samples were pooled at the approximate following intervals from the start of the infusion: - 12 to 0 h; 0 to 24 h; and 24-48 h. Following each voiding, intermediate urines were either kept on ice or refrigerated. Cumulative collections at the end of the specified intervals were measured for volume and aliquots were frozen ( - 15 "C) until analysis. Drug and reduced metabolite stability in plasma and urine Only those stability aspects relevant to the present study will be discussed. Using a sensitive and specific mass spectrometric method (vide infra), plasma stability was investigated at - 15 "C over a period of 24 weeks. No evidence of concentration change was noted for the reduced metabolite. However, with respect to dolasetron, continuous loss was noted over time. Within 2 weeks, an approximate 10 per cent decrease in intact drug was observed. Consequently, plasma assays, which simultaneously measured both drug and reduced metabolite, were always conducted within 3 weeks. Urine analyses for intact drug and reduced metabolite were conducted within 3 months; in separate studies designed to investigate stability (- 15 "C), no evidence of concentration changes for either compound was observed over this time period. The reason for dolasetron plasma instability is unclear. However, studies have indicated it does not degrade to reduced dolasetron. It is speculated that constraints of the ring structure (involving the keto group) cause a non-enzymatic, non-specific ring opening. Reduced dolasetron is presumably stable, because it lacks the keto group. The reason for dolasetron instability in plasma, given its stability in urine, remains unresolved.
696
H. BOXENBAUM ET ,415.
Bioanalytical methodologies Bioanalytical methodologies (all of which are specific) will be published at a future data. Therefore, only salient features of the methods will be discussed here. For studies conducted with doses between 0-6-4-0mg kg-l, the plasma assay procedure utilized a liquid-liquid extraction, derivatization with trifluoroacetic anhydride, gas chromatographic separation, and mass spectrometric determination of both drug and reduced metabolite. The lower limit of quantitation was 5 ng ml- for dolasetron and 2 ng ml- for reduced dolasetron. For studies at a dose of 5 mg kg-l, an HPLC method with UV detection was employed; it utilized essentially the same sample preparation procedure as was used for the GC-MS assay. The lower limit of quantitation was 10 ng ml- for dolasetron and 5 ng ml- for reduced metabolite. Both dolasetron and reduced metabolite were also quantitated from urine using an HPLC method. The lower limits of quantitation were 0.05 pg mland 0.10 pg ml- for dolasetron and reduced metabolite, respectively.
Pharmacokinetic analyses Throughout this paper, assayed concentrations and amounts of drug and metabolite are expressed in terms of free base. This is in contrast to dose, which is expressed in terms of dolasetron mesylate monohydrate. Appropriate molecular weight corrections were made for pharmacokinetic calculations. Visual inspection of plasma concentration-time profiles of intact dolasetron indicated disposition was either mono- or bi-exponential (19 and 13 subjects, respectively; see Figure 2). Due to the extremely rapid decay of dolasetron plasma levels, there were few data points. Consequently, good curve-fits were always obtained. Curves which decayed monoexponentially were analyzed as follows. Linear functions were fitted to semi-logarithmically transformed plasma concentrationtime data employing linear least squares analysis. The zero-time plasma concentration obtained in this fashion was converted to that which would have been obtained had the dose been administered as a bolus. For this purpose, the method of Wagner4 was employed. From bolus-derived, zero-time plasma concentration and monoexponential disposition rate constant, the following parameters were readily calculated9 clearance, volume of distribution, mean residence time, and disposition half-life. The following procedure was used for biexponential curve fitting. Initial estimates of coefficients and exponents were obtained for subject 20 (considered representative) using the ESTRIP program.6 The coefficients were converted to those which would have been obtained had the dose been administered as a bolus.4 Bolus-equivalent coefficients and exponents were then used to calculate microconstants of a two-compartment model, with elimination solely from the central c~mpartment;~ this had the advantage of converting to
697
HUMAN DOLASETRON PHARMACOKINETICS
,---. 1
SUBJECT 26
SUBJECT 28
SUBJECT 29
4.0 MG/KG
4.0 MG/KG
5.0 MG/KG
3 \ c7
5
1000
1000
1000
ci Z
0
I
0 6
I
0
I
I m
4
a
I
100
Z
I
0
I
K
+ W m
4 0
n
O b 0
10
10
5
I
- 1 0
\
\\
0
I
I
1
I
1
2
3
4
(a)
5
I
l
- 1 0 1
10
l
1
1
2
3
4
TIME (HOURS) (b)
5
I
- 1 0
0
I
I
I
I
1
2
3
4
(c)
Figure 2. Representative dolasetron plasma concentration-time profiles following intravenous administration. Disposition for subject 26 is monoexponential (a), whereas that of subject 29 is biexponential (c). Disposition for subject 28 (b) is also biexponential, but due to the last data point. Note that as one moves from left to right, and data are extended in time, curvature becomes more pronounced. See Figure 3 for both the dolasetronand reduced metabolite plasma concentration-time profiles for subject 28
parameters which were all dose-independent (in contrast to coefficients). These were used as initial estimates in all the biexponential curve-fits. PCNONLIN’ was used to obtain final parameter estimates. A subroutine drawn from the PCNONLIN library was used to iterate microconstants. Initial estimates were obtained as described in the previous paragraph; it was assumed, and it was subsequently confirmed, that initial estimates for subject 20 would suffice for all subjects. Initially, a weighting factor of the reciprocal of the observed plasma concentration was employed. This provided satisfactory fits (see beyond for criteria of goodness-of-fit) for 9 of the 13 subjects with biexponential disposition. For those subjects in whom a good fit was not obtained, the reciprocal of the observed plasma concentration squared proved to be satisfactory (4 of the 13 subjects). Goodness-of-fit was judged on the basis of randomness of scatter of data points about the fitted curves.* Clearance, volume of distribution during the terminal phase, volume of distribution at steady-state, mean residence time, and disposition half-life were calculated from parameter estimates by conventional r n e t h o d ~ . ~ * ~ Dolasetron was excreted in the urine to the extent of less than 1 per cent; this was considered quantitatively negligible and was not used in pharmacokinetic analysis.
698
H. BOXENBAUM ET AL.
Based on visual inspection of the reduced dolasetron plasma concentrationtime profiles, maximum plasma concentrations, and their corresponding times were determined. Areas under the plasma concentration-time curves were determined by the linear trapezoidal rule from zero time until the last quantifiable data point. Area extrapolation from the last quantifiable data point to infinity was obtained by dividing the concentration at the last quantifiable data point by the terminal disposition rate constant. The terminal disposition rate constant was obtained from linear least squares analysis of the logarithmically transformed plasma concentrations during the terminal phase. From this rate constant, the terminal disposition half-life was readily determined. Due to insufficient data, a terminal phase was not observed for subject 32. In all but four subjects, the terminal phase was judged (based on visual inspection) to begin with the 12-h data point. In subjects 9-12, the terminal phase was judged to begin with the 24-h data point. Outliers were not used in data analyses. Calculated per cent of the dose appearing as reduced dolasetron in the 0 to infinity urine was calculated as follows. The amounts in the nominal 0-24 and 24-48 h urines were added together. The amount excreted in the urine thereafter (to infinity) was calculated by the method of Beckett and Rowland.lo For purposes of this calculation, it was assumed that drug excreted during the nominal 24-48 h collection interval was excreted during the terminal disposition phase. The extrapolated amount never exceeded more than 6 per cent of the total amount excreted, and usually was less than 1 per cent. Renal clearance was calculated as the amount excreted from time 0 to infinity divided by the area under the plasma concentration-time curve from time 0 to infinity. In a few cases where 24-48 h urinary excretion data were unavailable, the 0-24 h data were used. RESULTS AND DISCUSSION
Dolasetron disposition Visual examination of the plasma concentration-time profiles indicated monoexponential disposition (n = 19) in some subjects and biexponential disposition (n= 13) in others (Figure 2). Generally, biexponential disposition tended to be observed in those subjects receiving the higher doses, in which plasma concentrations could be quantitated for a longer period of time. This suggests biexponential or multiexponential disposition would have been observed in all subjects had the assay sensitivity allowed the quantitation of lower concentrations at later times. Even when biexponential disposition was observed, plasma concentrations generally could not be measured out beyond 2-4 h. Thus, one cannot place a great deal of confidence in the exact nature of disposition. Nonetheless, over the time and concentration ranges observed, inspection of the plasma concentration-time profiles indicates that dolasetron disappears very rapidly from plasma. Concomitantly, plasma concentrations of reduced metabolite appear rapidly (vide infra).
699
HUMAN DOLASETRON PHARMACOKINETICS
Table 1. Median and mean pharmacokinetic parameters characterizing dolasetron disposition. CL is clearance, v d is volume of distribution (monoexponential disposition), v d m is volume of distribution at steady-state (biexponential disposition), t , is monoexponential disposition half-life, teBis biexponential terminal disposition half-life and MRT is mean residence time Parameter
Median
Mean
CL (ml min-' kg-*: all subjects v d (1 kg- l) monoexponential disposition Vd,, (Ikg - l) biexponential disposition t , (min) monoexponential disposition t E B(rnin) biexponential disposition MRT (min) monoexponential disposition MRT (min) biexponential disposition
107.7 1.63 2-76 8.67 94.6 12.5 22.8
116.5 1 -69 3.42 9.36. 106-4t 13.5 32.4
070
cv
31.6 34.6 74.2 22*6* 57.5' 22.1 64.2
*The harmonic mean half-life" is 9.02 min and the pseudo 070 CV is 17.6%. 'The harmonic mean half-life1'is 71.1 min and the pseudo To CV is 70.5%.
Although disposition may not be well characterized due to the inability to quantitate plasma concentrations beyond a few hours, parameters were calculated for heuristic purposes. Table 1 collates mean and median parameters. Correlation analyses of clearance and volume of distribution parameters vs dose (mono- and bi-exponential disposition data treated separately) indicated a lack of statistical significance at the 0.05 level. Thus, there is no evidence of dosedependent disposition.
Reduced dolasetron pharmacokinetics Figure 3 illustrates plasma concentration-time profiles for both dolasetron and reduced dolasetron in a representative subject. Table 2 collates mean and median parameters. Peak plasma concentrations occur at a median time of 0.625 h, suggesting rapid formation from dolasetron; median terminal disposition half-life is 7.56 h. A median 30.96 per cent of the dose is excreted in the urine as this metabolite. Median renal clearance is 2.68 ml min- kg- l . This latter value suggests some active tubular transport in the urinary excretion process. Most important, perhaps, is the observation that the reduced dolasetron area under the plasma concentration-time curve exceeds that of intact drug by a factor of approximately 1 1 - 9 fold (median value, see Table 2) and is doseindependent (vide infra). Since the racemate of this metabolite possesses considerable in vivo anti-emetic activity in an animal model (data on file), it is probable that considerable therapeutic activity resides with this moiety. Areas under the reduced dolasetron plasma concentration-time curves are proportional to dose (Figure 4), as are maximum plasma concentrations (the correlation coefficient is significant with p
HUMAN DOLASETRON PHARMACOKINETICS: I. DISPOSITION FOLLOWING SINGLE-DOSE INTRAVENOUS ADMINISTRATION TO NORMAL MALE SUBJECTS HAROLD BOXENBAUM*, TODD GILLESPIE', KATHLEEN HECK*and WILLIAM HAHNEg Marion Merrell Dow Inc., P.O. Box 9627, Kansas City, MO 64134, U.S.A.
ABSTRACT Dolasetron is a 5-hydroxytryptamine antagonist active at type 111 receptors; it is presently undergoing clinical evaluation for the reduction/prevention of cancer chemotherapyinduced nausea and vomiting. Following intravenous administration to healthy male subjects of doses ranging from 0.6 to 5 mg kg-I, dolasetron disappeared extremely rapidly from plasma; concentrations were generally measurable for only 2-4 h. Less than 1 per cent of the dose was excreted intact in urine. A major plasma metabolite, reduced dolasetron, peaked rapidly at approximately 0.625 h (median). Its median terminal disposition half-life was 7 - 5 6 h; median values for fraction of dose excreted in urine and renal clearance were 3 1 -0per cent and 2.68 ml min-' kg-I, respectively. Over the dose-range covered, pharmacokinetics of both dolasetron and reduced metabolite appeared to be independent of dose. The median ratio of the areas under the plasma concentration-time curves for metabolite relative to dolasetron was 11 -9. As a result of its activity and significant plasma concentrations, reduced dolasetron may play a significant role in pharmacodynamic activity. KEY WORDS
Dolasetron Pharmacokinetics Intravenous dosing
INTRODUCTION Employing an isolated rabbit heart preparation,' dolasetron was found to be a specific and selective 5-hydroxytryptamine antagonist active at type I11 receptors. It is presently undergoing clinical investigation for the reduction/ prevention of cancer chemotherapy-induced nausea and emesis. The intravenous formulation is prepared using the monohydrated mesylate (methane sulfonate) salt (see Figure 1).
Present addresses: 'Clinical Pharmacokinetics, Wyeth-Ayerst Research, P.O. Box 8299, Phila., PA 19101, U.S.A. 'Lilly Clinic, Wishard Memorial Hospital, 1001 W. 10th St., Indianapolis, IN 46202, U.S.A. *F.A.C.T., 85 N.E.Loop 410, Suite 612, San Antonio, TX 78216, U.S.A. §Addressee for correspondence.
0142-2782/92/09O693-09$09.50 0 1992 by John Wiley & Sons, Ltd.
Received 6 January 1992 Accepted 20 April 1992
694
H. BOXENBAUM ET AL.
*
CH~SOSH* H20
0
OH
0
0
Figure 1 . Chemical structures of: (a) dolasetron mesylate monohydrate, used in preparation of the injectable dosage form; (b) dolasetronfree base; and (c) reduced dolasetron, a major plasma metabolite in humans
A major metabolite of dolasetron identified in human plasma is the reduced compound (Figure 1). Whereas dolasetron is achiral, keto-reduction produces a chiral carbon center. Both intact drug and a racemic mixture of the reduced metabolite have been evaluated for their affinities to 5-HT3 receptors expressed by a neuroblastoma cell line.2.3Reduced dolasetron is approximately 100 times more potent than dolasetron. The purpose of this initial pharmacokinetic investigation in humans was to characterize the pharmacokinetics of dolasetron following single-dose, intravenous administration of escalating doses. Assays were available to measure intact drug in plasma and urine, as well as the combined sum of ( +) and ( - ) reduced dolasetron in these same biological fluids. It is anticipated that future pharmacokinetic and related studies will focus on such things as plasma protein binding, multiple intravenous dosing, single and multiple oral dosing, influence of disease states, pharmacodynamics, metabolite identification/kinetics, chirality metabolite issues, etc.
HUMAN DOLASETRON PHARMACOKINETICS
695
METHODS
Study procedures Thirty-two, healthy male volunteers of mixed race and ethnicity participated in this study; all had provided informed consent. Subjects were between the ages of 19 and 38 years, were in good general health (as determined by physical examination and laboratory tests), were within 10 per cent of their ideal body weight, did not smoke tobacco, and were drug-free. Following a minimum 12 h food fast and 2 h water fast, subjects received dolasetron intravenously at approximately 8 am. After dosing, water was permitted after 2 h and food after 4 h. Doses were administered as dolasetron mesylate monohydrate, contained within approximately 100 ml of sterile solution. The drug was administered intravenously at a constant rate over a period of 1Omin. Four subjects each received nominal doses of 0-6, 1.25,2.0, 2.5, 4-0, and 5.Omgkg-l; eight subjects received doses of 3.Omgkg-l (expressed as dolasetron mesylate monohydrate). Fifteen millilitre blood samples (containing EDTA as anticoagulant) were collected at approximately 0, 0.25, 0.50, 0.75, 1, 2, 4, 6, 8, 12, 24, 36, and 48 h after the start of the infusion. The blood was centrifuged, plasma harvested, and this was subsequently stored frozen (- 15 "C) in glass containers until analysis. Urine samples were pooled at the approximate following intervals from the start of the infusion: - 12 to 0 h; 0 to 24 h; and 24-48 h. Following each voiding, intermediate urines were either kept on ice or refrigerated. Cumulative collections at the end of the specified intervals were measured for volume and aliquots were frozen ( - 15 "C) until analysis. Drug and reduced metabolite stability in plasma and urine Only those stability aspects relevant to the present study will be discussed. Using a sensitive and specific mass spectrometric method (vide infra), plasma stability was investigated at - 15 "C over a period of 24 weeks. No evidence of concentration change was noted for the reduced metabolite. However, with respect to dolasetron, continuous loss was noted over time. Within 2 weeks, an approximate 10 per cent decrease in intact drug was observed. Consequently, plasma assays, which simultaneously measured both drug and reduced metabolite, were always conducted within 3 weeks. Urine analyses for intact drug and reduced metabolite were conducted within 3 months; in separate studies designed to investigate stability (- 15 "C), no evidence of concentration changes for either compound was observed over this time period. The reason for dolasetron plasma instability is unclear. However, studies have indicated it does not degrade to reduced dolasetron. It is speculated that constraints of the ring structure (involving the keto group) cause a non-enzymatic, non-specific ring opening. Reduced dolasetron is presumably stable, because it lacks the keto group. The reason for dolasetron instability in plasma, given its stability in urine, remains unresolved.
696
H. BOXENBAUM ET ,415.
Bioanalytical methodologies Bioanalytical methodologies (all of which are specific) will be published at a future data. Therefore, only salient features of the methods will be discussed here. For studies conducted with doses between 0-6-4-0mg kg-l, the plasma assay procedure utilized a liquid-liquid extraction, derivatization with trifluoroacetic anhydride, gas chromatographic separation, and mass spectrometric determination of both drug and reduced metabolite. The lower limit of quantitation was 5 ng ml- for dolasetron and 2 ng ml- for reduced dolasetron. For studies at a dose of 5 mg kg-l, an HPLC method with UV detection was employed; it utilized essentially the same sample preparation procedure as was used for the GC-MS assay. The lower limit of quantitation was 10 ng ml- for dolasetron and 5 ng ml- for reduced metabolite. Both dolasetron and reduced metabolite were also quantitated from urine using an HPLC method. The lower limits of quantitation were 0.05 pg mland 0.10 pg ml- for dolasetron and reduced metabolite, respectively.
Pharmacokinetic analyses Throughout this paper, assayed concentrations and amounts of drug and metabolite are expressed in terms of free base. This is in contrast to dose, which is expressed in terms of dolasetron mesylate monohydrate. Appropriate molecular weight corrections were made for pharmacokinetic calculations. Visual inspection of plasma concentration-time profiles of intact dolasetron indicated disposition was either mono- or bi-exponential (19 and 13 subjects, respectively; see Figure 2). Due to the extremely rapid decay of dolasetron plasma levels, there were few data points. Consequently, good curve-fits were always obtained. Curves which decayed monoexponentially were analyzed as follows. Linear functions were fitted to semi-logarithmically transformed plasma concentrationtime data employing linear least squares analysis. The zero-time plasma concentration obtained in this fashion was converted to that which would have been obtained had the dose been administered as a bolus. For this purpose, the method of Wagner4 was employed. From bolus-derived, zero-time plasma concentration and monoexponential disposition rate constant, the following parameters were readily calculated9 clearance, volume of distribution, mean residence time, and disposition half-life. The following procedure was used for biexponential curve fitting. Initial estimates of coefficients and exponents were obtained for subject 20 (considered representative) using the ESTRIP program.6 The coefficients were converted to those which would have been obtained had the dose been administered as a bolus.4 Bolus-equivalent coefficients and exponents were then used to calculate microconstants of a two-compartment model, with elimination solely from the central c~mpartment;~ this had the advantage of converting to
697
HUMAN DOLASETRON PHARMACOKINETICS
,---. 1
SUBJECT 26
SUBJECT 28
SUBJECT 29
4.0 MG/KG
4.0 MG/KG
5.0 MG/KG
3 \ c7
5
1000
1000
1000
ci Z
0
I
0 6
I
0
I
I m
4
a
I
100
Z
I
0
I
K
+ W m
4 0
n
O b 0
10
10
5
I
- 1 0
\
\\
0
I
I
1
I
1
2
3
4
(a)
5
I
l
- 1 0 1
10
l
1
1
2
3
4
TIME (HOURS) (b)
5
I
- 1 0
0
I
I
I
I
1
2
3
4
(c)
Figure 2. Representative dolasetron plasma concentration-time profiles following intravenous administration. Disposition for subject 26 is monoexponential (a), whereas that of subject 29 is biexponential (c). Disposition for subject 28 (b) is also biexponential, but due to the last data point. Note that as one moves from left to right, and data are extended in time, curvature becomes more pronounced. See Figure 3 for both the dolasetronand reduced metabolite plasma concentration-time profiles for subject 28
parameters which were all dose-independent (in contrast to coefficients). These were used as initial estimates in all the biexponential curve-fits. PCNONLIN’ was used to obtain final parameter estimates. A subroutine drawn from the PCNONLIN library was used to iterate microconstants. Initial estimates were obtained as described in the previous paragraph; it was assumed, and it was subsequently confirmed, that initial estimates for subject 20 would suffice for all subjects. Initially, a weighting factor of the reciprocal of the observed plasma concentration was employed. This provided satisfactory fits (see beyond for criteria of goodness-of-fit) for 9 of the 13 subjects with biexponential disposition. For those subjects in whom a good fit was not obtained, the reciprocal of the observed plasma concentration squared proved to be satisfactory (4 of the 13 subjects). Goodness-of-fit was judged on the basis of randomness of scatter of data points about the fitted curves.* Clearance, volume of distribution during the terminal phase, volume of distribution at steady-state, mean residence time, and disposition half-life were calculated from parameter estimates by conventional r n e t h o d ~ . ~ * ~ Dolasetron was excreted in the urine to the extent of less than 1 per cent; this was considered quantitatively negligible and was not used in pharmacokinetic analysis.
698
H. BOXENBAUM ET AL.
Based on visual inspection of the reduced dolasetron plasma concentrationtime profiles, maximum plasma concentrations, and their corresponding times were determined. Areas under the plasma concentration-time curves were determined by the linear trapezoidal rule from zero time until the last quantifiable data point. Area extrapolation from the last quantifiable data point to infinity was obtained by dividing the concentration at the last quantifiable data point by the terminal disposition rate constant. The terminal disposition rate constant was obtained from linear least squares analysis of the logarithmically transformed plasma concentrations during the terminal phase. From this rate constant, the terminal disposition half-life was readily determined. Due to insufficient data, a terminal phase was not observed for subject 32. In all but four subjects, the terminal phase was judged (based on visual inspection) to begin with the 12-h data point. In subjects 9-12, the terminal phase was judged to begin with the 24-h data point. Outliers were not used in data analyses. Calculated per cent of the dose appearing as reduced dolasetron in the 0 to infinity urine was calculated as follows. The amounts in the nominal 0-24 and 24-48 h urines were added together. The amount excreted in the urine thereafter (to infinity) was calculated by the method of Beckett and Rowland.lo For purposes of this calculation, it was assumed that drug excreted during the nominal 24-48 h collection interval was excreted during the terminal disposition phase. The extrapolated amount never exceeded more than 6 per cent of the total amount excreted, and usually was less than 1 per cent. Renal clearance was calculated as the amount excreted from time 0 to infinity divided by the area under the plasma concentration-time curve from time 0 to infinity. In a few cases where 24-48 h urinary excretion data were unavailable, the 0-24 h data were used. RESULTS AND DISCUSSION
Dolasetron disposition Visual examination of the plasma concentration-time profiles indicated monoexponential disposition (n = 19) in some subjects and biexponential disposition (n= 13) in others (Figure 2). Generally, biexponential disposition tended to be observed in those subjects receiving the higher doses, in which plasma concentrations could be quantitated for a longer period of time. This suggests biexponential or multiexponential disposition would have been observed in all subjects had the assay sensitivity allowed the quantitation of lower concentrations at later times. Even when biexponential disposition was observed, plasma concentrations generally could not be measured out beyond 2-4 h. Thus, one cannot place a great deal of confidence in the exact nature of disposition. Nonetheless, over the time and concentration ranges observed, inspection of the plasma concentration-time profiles indicates that dolasetron disappears very rapidly from plasma. Concomitantly, plasma concentrations of reduced metabolite appear rapidly (vide infra).
699
HUMAN DOLASETRON PHARMACOKINETICS
Table 1. Median and mean pharmacokinetic parameters characterizing dolasetron disposition. CL is clearance, v d is volume of distribution (monoexponential disposition), v d m is volume of distribution at steady-state (biexponential disposition), t , is monoexponential disposition half-life, teBis biexponential terminal disposition half-life and MRT is mean residence time Parameter
Median
Mean
CL (ml min-' kg-*: all subjects v d (1 kg- l) monoexponential disposition Vd,, (Ikg - l) biexponential disposition t , (min) monoexponential disposition t E B(rnin) biexponential disposition MRT (min) monoexponential disposition MRT (min) biexponential disposition
107.7 1.63 2-76 8.67 94.6 12.5 22.8
116.5 1 -69 3.42 9.36. 106-4t 13.5 32.4
070
cv
31.6 34.6 74.2 22*6* 57.5' 22.1 64.2
*The harmonic mean half-life" is 9.02 min and the pseudo 070 CV is 17.6%. 'The harmonic mean half-life1'is 71.1 min and the pseudo To CV is 70.5%.
Although disposition may not be well characterized due to the inability to quantitate plasma concentrations beyond a few hours, parameters were calculated for heuristic purposes. Table 1 collates mean and median parameters. Correlation analyses of clearance and volume of distribution parameters vs dose (mono- and bi-exponential disposition data treated separately) indicated a lack of statistical significance at the 0.05 level. Thus, there is no evidence of dosedependent disposition.
Reduced dolasetron pharmacokinetics Figure 3 illustrates plasma concentration-time profiles for both dolasetron and reduced dolasetron in a representative subject. Table 2 collates mean and median parameters. Peak plasma concentrations occur at a median time of 0.625 h, suggesting rapid formation from dolasetron; median terminal disposition half-life is 7.56 h. A median 30.96 per cent of the dose is excreted in the urine as this metabolite. Median renal clearance is 2.68 ml min- kg- l . This latter value suggests some active tubular transport in the urinary excretion process. Most important, perhaps, is the observation that the reduced dolasetron area under the plasma concentration-time curve exceeds that of intact drug by a factor of approximately 1 1 - 9 fold (median value, see Table 2) and is doseindependent (vide infra). Since the racemate of this metabolite possesses considerable in vivo anti-emetic activity in an animal model (data on file), it is probable that considerable therapeutic activity resides with this moiety. Areas under the reduced dolasetron plasma concentration-time curves are proportional to dose (Figure 4), as are maximum plasma concentrations (the correlation coefficient is significant with p
