The influence of caffeine on the steady-state pharmacokinetics of theophylline During this open, two-period crossover study in eight healthy volunteers, 1200 mg anhydrous theophylline was administered as a two-stage infusion during 24 hours on day 6. During one of the 8-day periods, 300 mg caffeine, t.i.d., was administered orally. After the start of the theophylline infusion, plasma concentrations of theophylline and caffeine and urinary excretion of theophylline and four metabolites were determined frequently during 60 hours. With caffeine administration theophylline steady-state concentration and area under the curve increased by 23% and 40%, respectively, whereas the volume of distribution at steady state seemed unchanged. The cumulative urinary excretion of 1-methyluric acid and 1-methylxanthine did not reach a plateau, suggesting a capacity-limiting factor in their formation. Notwithstanding the mutual interference of theophylline and caffeine metabolism, the reduction in apparent total body clearance and elimination rate constant of theophylline by 29% and 31%, respectively, indicated a pronounced influence of concomitant administration of realistic amounts of caffeine. (Cul.; PHARMACOL THER 1991;49:248-55.)
Jan H. G. Jonkman, PhD, Frans A. E. Sollie, MSc, Rita Sauter, MRA, and Volker W. Steinijans, PhD Assen, The Netherlands, and Konstanz, Germany Theophylline (1,3-dimethylxanthine) and caffeine (1,3,7-trimethylxanthine) are metabolized extensively by hepatic cytochrome P-450-dependent enzymes. Only about 8% to 15% of a theophylline dose is excreted unchanged in urine. About 15% to 30% is metabolized by demethylation to 3-methylxanthine, approximately 16% to 40% is excreted as the C8 oxidation product 1,3-dimethyluric acid, and about 20% to 40% is excreted as 1-methyluric acid." about After an oral dose of 5 mg kg' 0.8% to 1.8% is excreted unchanged.2 The main metabolite, paraxanthine (1,7-dimethylxanthine), is found in urine up to about 80%,9 whereas paraxanthine, theobromine (3,7-dimethylxanthine), and 1,3dimethylxanthine time-averaged plasma concentrations after multiple oral caffeine-containing beverages are found to be 67%, 24%, and 8%, respectively. Many factors affecting hepatic uptake and biotransformation in the liver can influence theophylline and caffeine elimination. 10 Various interactions of theo1 1
From Pharma Bio-Research International B. V., Assen, and Byk Gulden Pharmaceuticals, Konstanz. Received for publication June 6, 1990; accepted Oct. 14, 1990. Reprint requests: Jan H. G. Jonkman, Pharma Bio-Research International B.V., P.O. Box 147, 9400 AC Assen, The Netherlands. 13/1/26092
248
phylline with various co-medications have been described. 12.13 The interaction between theophylline and caffeine in healthy humans has been described previously.14.15 Monks et al.14 found no effect on theophylline metabolism in a study in three healthy male subjects after supplemental caffeine intake (270 mg daily) to a usual caffeine-containing diet (300 to 700 mg daily) during 9 days. Theophylline was administered as an intravenous infusion of 100 mg theophylline on day 8 and urine samples were collected for 48 hours and analyzed for theophylline and major metabolites. Zilly et al.15 detected considerable change in elimination half-life (t1,2) and total clearance (CL) in a study in six healthy subjects. Here theophylline was administered as an intravenous infusion of 193.2 mg theophylline with or without coadministration of a single oral 400 mg dose of caffeine. The objective of our study was to investigate the pharmacokinetics of theophylline with and without concomitant multiple-dose administration of caffeine under intravenous steady-state conditions of theophylline.
SUBJECTS AND METHODS Subjects Eight healthy, nonsmoking adult male volunteers took part in this study. Ages ranged from 20 to 26
VOLUME 49 NUMBER 3
Effect of caffeine on theophylline
249
Table I. Demographic and anthropometric data and pharmacokinetic characteristics as derived from theophylline plasma concentration data obtained without and with concomitant caffeine administration in eight male subjects Subject
Sequence of administration Age (yr) Weight (kg)
1
2
3
4
5
6
7
8
Mean ± SD (range)
RT 25 86
TR 26 64
TR
RT 23 88
TR 26 76
RT 20 69
TR
21
21
RT 22
23(20-26)
77
71
77 (64-88)
87
k,,z (hr-') 0.1450 0.1088 0.1112 0.1054 0.1349 0.1202 0.0788 0.0999 0.0999 0.0753 0.0857 0.0692 0.1009 0.0771 0.0501 0.0675 100 T/R t12 (hr)
100 T/R
69
69
77
66
75
64
64
4.8 6.9
6.4 9.2
6.2
6.6 10.0
5.1 6.9
5.8
8.1
9.0
8.8 13.8
144
144
131
152
135
155
157
10.3 149
6.83 7.84
10.87 13.54
8.47 9.73
8.62 10.49
7.72 9.20
9.13 12.30
11.46 14.92
9.35 11.68
115
125
115
122
119
135
130
125
68
69 (65, 72)*
6.9
6.3 ± 1.2 9.3 ± 2.2 144 (139, 154)*
L-')
Cav (mg
100 T/R AUC (mg L-1
100 T/R
hr-1
kg')
123
364.61 285.33 295.40 237.34 296.49 445.92 331.41 528.55 353.66 434.18 314.77 468.46 713.26 488.38 145
124
147
133
158
160
147
0.0650 0.0514 0.0483 0.0462 0.0665 0.0587 0.0349 0.0510 0.0527 0.0355 0.0390 0.0314 0,0502 0.0371 0.0218 0.0346 100 T/R MRTaaJusted (hr)
81
7.8 10.0 100 T/R (L kg- `)
128
0.5071 0.5285 100 T/R
9.06 ± 1.53 11.21 ± 2.35 123 (117, 130)*
hr)
214.64 264.57 CL (L
0.1130 ± 0.021 0.0782 ± 0.017
104
308.89 ± 72.98 445.73 ± 141.09 140 (128, 154)*
0.0527 ± 0.0104 0.0378 ± 0.0099
81
68
75
63
62
68
71(65, 78)*
9.9
9.7
14.1 142
12.2
10.4 15.7
9.3 13.4
13.2 20.3
11.2 15.7
125
152
7.3 10.8 148
144
154
141
9.8 ± 1.9 14.0 ± 3.3 142 (133, 150)*
69
0.5091 0.4694 0.4778 0.4830 0.5455 0.4606 0.5691 0.4984 0.4739 0.4928 0.5407 0.4960 0.4425 0.5430 91 95 101 103 112 96 98
R, Reference; T, test; k,, elimination rate constant; t,,,, elimination half-life; residence time; Vss, steady-state volume of distribution. *Distribution-free point estimate and 90% confidence interval.
years (median, 23 years) and weights ranged from 64 to 88 kg (median, 77 kg). They signed a written informed consent form in accordance with the signed informed consent recommendations of the U.S. Food and Drug Administration. All participants were judged to be healthy by a physician on the basis of prestudy medical history, normal physical prestudy and poststudy examinations, electrocardiogram, and laboratory tests, including a test on creatinine clearance. No other medication was taken for 2 weeks before and during the study. No alcohol or xanthine-containing food or drinks were taken for 72 and 48 hours, respectively, before and during the study. On entry to the study, a urine sample was collected for a drug screening. All test results were negative.
0.5027 ± 0.0382 0.5020 ± 0.0345 100 (94, 108)*
C. average concentration; AUC, area under the curve: CL. clearance: MAT, mean
Study design The study was an open, two-period crossover trial with two study periods of 8 days and no washout. In each study period a total of 1200 mg anhydrous theophylline was administered as a rapid-loading infusion of 370 mg during 30 minutes, starting at 8 AM on day 6, followed by a constant-rate infusion of 830 mg during 231/2 hours. In one of the study periods 300 mg caffeine was administered orally as immediate-release 50 mg tablets at 8 AM, 2 PM, and 8 PM from day 1 until 2 PM on day 8 (test). The caffeine-free period served as reference. Blood samples (6 ml) were obtained (before administration) at 8 AM on days I to 6 and at 10-, 20-, and 30-minute and 1-hour intervals until 12 hours, 2-hour intervals until 48 hours, and 4-hour intervals until 60
CLIN PHARMACOL THER MARCH 1991
250 Jonkman et al. hours after the start of the infusion (at 8 AM on day 6). Plasma was collected after centrifugation at 1500 g and stored at -20° C until analysis. Plasma samples were analyzed for theophylline and caffeine. Urine samples were obtained from the 24-hour interval starting at 8 AM on day 5 and from 2- to 4-hour intervals starting at 8 AM on day 6 until 8 PM on day 8. Samples were stored at 4° to 6° C until the end of a urine-sampling interval and then stored at -20° C until analysis. Urine samples were analyzed for theophylline and major metabolites. To facilitate collection of urine samples, the volunteers were given 150 ml water every 2 hours starting at 8 AM on day 6. All participants were given a (no xanthine-containing) standard diet throughout the trial. The study protocol was reviewed and approved by the Medical Ethical Committee 'Stichting Beoordeling Ethiek Bio-Medisch Onderzoek' before the beginning of the study.
Study medications Intravenous aminophylline solutions prepared from Euphyllin ampules (240 mg aminophylline/10 ml), batch number 85K14, were from Byk Nederland B.V., Zwanenburg, The Netherlands. Caffeine immediaterelease 50 mg tablets, batch number 85126-F5192, were from Pharmachemie B. V., Haarlem, The Netherlands. Bioanalytic methods Theophylline in plasma. A selective and sensitive HPLC method was used for the determination of theophylline in plasma.16 The method results in good separation of theophylline from other potentially interfering xanthine compounds such as paraxanthine and caffeine. Typical retention times for caffeine, theophylline, theobromine, paraxanthine, and P-hydroxyethyltheophylline were 3.0, 3.6, 4.7, 5.2, and 6.0 minutes, respectively. Characteristic features of the assay method were: detection limit, 0.05 mg L-'; lower limit of quantitation, 0.25 mg L-1; recovery, 88.3% ± 1.8% (mean ± SD); intraday reproducibility (coefficient of variation), 0.8%; interday reproducibility (coefficient of variation), 1.0%; and accuracy, 0.8%. Caffeine in plasma. A selective and sensitive straight-phase HPLC method was used for the determination of caffeine in plasma. To a plasma sample of 200 1.11, 200 ill internal standard solution (D4030, 5 mg L-1, A.B. Draco, Lund, Sweden), 200 pi calibration solution or water, and 200 ill saturated ammonium sulfate was added. The mixed sample preparation was added to 1.6 ml chloroform/isopropanol
(9.5:0.5 vol/vol) and mechanically agitated and centrifuged (500 rpm). The organic layer was evaporated under nitrogen with a Techne Sample Concentrator (Techne, Inc., Princeton, N.J.) at 35° C. After adding 1 ml mobile phase and mixing, 50 ill was injected into the column with a Waters Intelligent Sample Processor, model M710B (Waters Associates, Milford, Mass.). A 3 pt.m microsphere silica cartridge column, 100 by 4.6 mm (Chrompack International B.V., Middelburg, The Netherlands), was used for separation. A Kratos Spectroflow 783 detector (Kratos Analytical Instruments, Ramsey, N.J.) set at 280 nm was used with a Spectra-Physics SP 4290 integrator (SpectraPhysics Inc., San Jose, Calif.) for quantitation. The mobile phase consisted of a mixture of methanol, acetic acid, tetrahydrofuran, chloroform, and n-heptane (80:1:12.5:1100:820 vol/vol), which was filtered and degassed. The flow rate was 1.0 ml min-1. Retention times for caffeine, theophylline, D-4030 (internal standard), theobromine, paraxanthine, and 13-hydroxyethyltheophylline were 2.3, 3.0, 3.7, 4.0, 4.8, and 5.6 minutes, respectively. Calibration curves were obtained from blank plasma samples spiked with standard of caffeine (0.5 to 12.0 mg L-1). The method shows the following characteristics: lower limit of quantitation, 0.5 mg L-1; interday reproducibility (coefficient of variation), less than 3.2%; and accuracy, less than 0.9%. Theophylline and major metabolites in urine. The method consists of a rapid and highly selective procedure for the determination of theophylline and its major metabolites (i.e., 1,3-dimethyluric acid, 1-methyluric acid, 3-methylxanthine, and 1-methylxanthine). A combination of normal and ion-pair liquid-liquid extraction followed by reversed-phase ion-pair gradient elution HPLC with a 3 p.m Pecosphere C-8 cartridge column 83 by 4.6 mm (Perkin-Elmer, Norwalk, Conn.) for separation was used. Detection was performed by measurement of the ultraviolet absorbance at 280 nm. The method is based on one developed earlier by Muir et al. 17 The method shows the following characteristics: lower limit of quantitation of all compounds (2 mg L-1), recovery (3-methylxanthine, 86%; 1,3dimethylxanthine, 86%; 1-methyluric acid, 67%; and 1,3-dimethyluric acid, 78%), and intraday reproducibility (coefficient of variation) (3-methylxanthine, 6.6%; 1,3-dimethylxanthine, 7.7%; 1-methyluric acid, 9.1%; and 1,3-dimethyluric acid, 8.3%).
Pharmacokinetics and statistics The following pharmacokinetic characteristics were determined for theophylline in plasma: the average
VOLUME 49 NUMBER 3
Effect of caffeine on theophylline
steady-state concentration from 1 to 24 hours after the start of the first theophylline infusion Cav= C av(1-24 hr) and, to detect possible day and night differences in clearance Cav(1-12 hr) and Cav(12-24 hr), the Cav values were obtained by dividing the respective areas under the curve (AUCs) (trapezoidal method) by the corresponding time intervals. Because of the superposition of the short-term infusion for 1/2 hour and the subsequent long-term infusion for 231/2 hours, steady-state concentrations cannot be expected within the first hour, which therefore were skipped in the calculation of the average steady-state concentrations. The theophylline elimination rate constant (k,z) was estimated by monoexponential curve fitting from 24 to at least 46 hours (generally 60 hours) after start of the infusion (at least 12 and generally 16 data points), the t2 was calculated as t12 = 1n2/k,,,. The AUC was determined by the trapezoidal method up to 60 hours after start of the infusion and then extrapolated to infinity by adding C6oh,./k,. The mean residence time (MRT) was calculated as ratio of the area under the moment curve (AUMC) and AUC: MRT = AUMC0f/AUC0-1f in which AUMCO-Inf
= AUMC0_60 hr
60 C60 hlikm,z
C60
MRT was then adjusted for the residence time in the infusion device's: MRTadjusted = MRT
MRTinfusion device
In the case of two subsequent infusions of doses D, and D2 during infusion times T, and T2, respectively, one obtains the followine:
MRT;
device
= D i/(D + D2) T112 + DAD, + D2) (T, + T2/2)
Theophylline CL was calculated as CL = dose/ AUC, the apparent volume of distribution at steady state as Vss = CL MRTadjusted 18 Pharmacokinetic characteristics as obtained with (test) and without caffeine administration (reference) were analyzed statistically by the distribution-free crossover analysis for the multiplicative model, including 90% confidence intervals for the ratio 100 test/reference.2° Equivalence between test and reference can be concluded if the 90% confidence interval is in the bioequivalence range (parametric2I and nonparametric20).
251
RESULTS Plasma concentration data. Individual pharmacokinetic characteristics together with the ratios 100 test/ reference for theophylline are presented in Table I. During concomitant administration of theophylline and caffeine, a mutual influence on the pharmacokinetic characteristics of both drugs can be seen (Fig. 1). Steady-state levels of caffeine, as indicated by predose plasma levels at 8 AM on days 1 to 6, showed about a twofold rise during the theophylline infusion. On the other hand, the average theophylline plasma concentration during the 1- to 24-hour interval after the start of the infusion showed an increase of about 23% after concomitant caffeine administration (p < 0.01). Under the same conditions the (apparent) CL of theophylline was reduced by 29% (p < 0.01), the (apparent) elimination rate constant was reduced by 31% (p < 0.01), and hence the (apparent) elimination t,,2 was prolonged by 44% (p < 0.01). The MRTadjusted reflected the change in t2 and was prolonged by 42%. The (apparent) Vss was unchanged. Comparison of the average theophylline plasma concentration during the 1- to 12-hour and 12- to 24hour intervals after start of the infusion did not reveal a significant diurnal effect for either treatment. For the treatment without caffeine the Cav(12-24 hr) was somewhat lower than the Cav(1-12 hr) (means of 8.97 and 9.15 mg respectively), whereas for the treatment with caffeine the opposite was found (means of 11.71 and 10.67 mg In the latter case a contribution of the caffeine administration to the higher value of Cav( 12-24 hr) of theophylline seemed likely. However, day and night equivalence can be concluded because the 90% confidence intervals for the ratio of expected medians of night and day ranged from 0.91 to 1.03 and from 1.01 to 1.15, respectively. Urinary excretion data. Urine collections were complete during a 60-hour period, thus allowing for calculation of the cumulative fraction of the dose excreted in the form of unchanged drug and particular metabolites. Although the cumulative excretion-time curves were generally "plateauing" about 48 hours after the start of the infusion, the 1-methylxanthine and 1-methyluric acid metabolites (and to a lesser degree the 3-methylxanthine metabolite) after coadministration of caffeine were excreted at significant amounts 60 hours after the start of the infusion (Figs. 2 and 3). The total urinary recovery of theophylline and its metabolites amounted on a molar basis to 77.2% (67.0% to 87.0%) (median; range) without caffeine and 135.1% (124.2% to 145.5%) with caffeine coadminis-
k,
L,
L',
CLIN EHARMACOL THER MARCH 1991
252 Jonkman et al.
12
24
36
60
48
Time (h) Theophylline with caffeine
e-e-e Theophylline without caffeine Caffeine
Fig. 1. Mean plasma concentrations of theophylline (1,3-dimethylxanthine) and caffeine (1,3,7trimethylxanthine) after a total of 1200 mg anhydrous theophylline, administered as a short-term loading infusion followed by a constant-rate infusion during the 24-hour interval starting at 8 AM on day 6, with and without 300 mg caffeine t.i.d., administered orally from day until 2 PM on day 8. (From 48 to 56 hours, hypothetical curve for caffeine; for details see Study Design and Dosage.) 1
0
12
36
24
48
60
Time (h) 0
-
E.-6-e
1,3 1,3
DMX no caffeine
1,3-DMX caffeine
- DMU no caffeine *-*- 1,3- DMU caffeine
3-MX no
caffeine
A-*-i 3-MX
caffeine
Fig. 2. Cumulative renal excretion of theophylline (1,3-DMX), 1,3-dimethyluric acid (1,3DMU), and 3-methylxanthine (3-MX). (For details see legend of Fig. 1.)
VOLUME 49 NUMBER 3
Effect of caffeine on theophylline
0
12
36
24
48
253
60
Time (h) s,
a
1
Jan H. G. Jonkman, PhD, Frans A. E. Sollie, MSc, Rita Sauter, MRA, and Volker W. Steinijans, PhD Assen, The Netherlands, and Konstanz, Germany Theophylline (1,3-dimethylxanthine) and caffeine (1,3,7-trimethylxanthine) are metabolized extensively by hepatic cytochrome P-450-dependent enzymes. Only about 8% to 15% of a theophylline dose is excreted unchanged in urine. About 15% to 30% is metabolized by demethylation to 3-methylxanthine, approximately 16% to 40% is excreted as the C8 oxidation product 1,3-dimethyluric acid, and about 20% to 40% is excreted as 1-methyluric acid." about After an oral dose of 5 mg kg' 0.8% to 1.8% is excreted unchanged.2 The main metabolite, paraxanthine (1,7-dimethylxanthine), is found in urine up to about 80%,9 whereas paraxanthine, theobromine (3,7-dimethylxanthine), and 1,3dimethylxanthine time-averaged plasma concentrations after multiple oral caffeine-containing beverages are found to be 67%, 24%, and 8%, respectively. Many factors affecting hepatic uptake and biotransformation in the liver can influence theophylline and caffeine elimination. 10 Various interactions of theo1 1
From Pharma Bio-Research International B. V., Assen, and Byk Gulden Pharmaceuticals, Konstanz. Received for publication June 6, 1990; accepted Oct. 14, 1990. Reprint requests: Jan H. G. Jonkman, Pharma Bio-Research International B.V., P.O. Box 147, 9400 AC Assen, The Netherlands. 13/1/26092
248
phylline with various co-medications have been described. 12.13 The interaction between theophylline and caffeine in healthy humans has been described previously.14.15 Monks et al.14 found no effect on theophylline metabolism in a study in three healthy male subjects after supplemental caffeine intake (270 mg daily) to a usual caffeine-containing diet (300 to 700 mg daily) during 9 days. Theophylline was administered as an intravenous infusion of 100 mg theophylline on day 8 and urine samples were collected for 48 hours and analyzed for theophylline and major metabolites. Zilly et al.15 detected considerable change in elimination half-life (t1,2) and total clearance (CL) in a study in six healthy subjects. Here theophylline was administered as an intravenous infusion of 193.2 mg theophylline with or without coadministration of a single oral 400 mg dose of caffeine. The objective of our study was to investigate the pharmacokinetics of theophylline with and without concomitant multiple-dose administration of caffeine under intravenous steady-state conditions of theophylline.
SUBJECTS AND METHODS Subjects Eight healthy, nonsmoking adult male volunteers took part in this study. Ages ranged from 20 to 26
VOLUME 49 NUMBER 3
Effect of caffeine on theophylline
249
Table I. Demographic and anthropometric data and pharmacokinetic characteristics as derived from theophylline plasma concentration data obtained without and with concomitant caffeine administration in eight male subjects Subject
Sequence of administration Age (yr) Weight (kg)
1
2
3
4
5
6
7
8
Mean ± SD (range)
RT 25 86
TR 26 64
TR
RT 23 88
TR 26 76
RT 20 69
TR
21
21
RT 22
23(20-26)
77
71
77 (64-88)
87
k,,z (hr-') 0.1450 0.1088 0.1112 0.1054 0.1349 0.1202 0.0788 0.0999 0.0999 0.0753 0.0857 0.0692 0.1009 0.0771 0.0501 0.0675 100 T/R t12 (hr)
100 T/R
69
69
77
66
75
64
64
4.8 6.9
6.4 9.2
6.2
6.6 10.0
5.1 6.9
5.8
8.1
9.0
8.8 13.8
144
144
131
152
135
155
157
10.3 149
6.83 7.84
10.87 13.54
8.47 9.73
8.62 10.49
7.72 9.20
9.13 12.30
11.46 14.92
9.35 11.68
115
125
115
122
119
135
130
125
68
69 (65, 72)*
6.9
6.3 ± 1.2 9.3 ± 2.2 144 (139, 154)*
L-')
Cav (mg
100 T/R AUC (mg L-1
100 T/R
hr-1
kg')
123
364.61 285.33 295.40 237.34 296.49 445.92 331.41 528.55 353.66 434.18 314.77 468.46 713.26 488.38 145
124
147
133
158
160
147
0.0650 0.0514 0.0483 0.0462 0.0665 0.0587 0.0349 0.0510 0.0527 0.0355 0.0390 0.0314 0,0502 0.0371 0.0218 0.0346 100 T/R MRTaaJusted (hr)
81
7.8 10.0 100 T/R (L kg- `)
128
0.5071 0.5285 100 T/R
9.06 ± 1.53 11.21 ± 2.35 123 (117, 130)*
hr)
214.64 264.57 CL (L
0.1130 ± 0.021 0.0782 ± 0.017
104
308.89 ± 72.98 445.73 ± 141.09 140 (128, 154)*
0.0527 ± 0.0104 0.0378 ± 0.0099
81
68
75
63
62
68
71(65, 78)*
9.9
9.7
14.1 142
12.2
10.4 15.7
9.3 13.4
13.2 20.3
11.2 15.7
125
152
7.3 10.8 148
144
154
141
9.8 ± 1.9 14.0 ± 3.3 142 (133, 150)*
69
0.5091 0.4694 0.4778 0.4830 0.5455 0.4606 0.5691 0.4984 0.4739 0.4928 0.5407 0.4960 0.4425 0.5430 91 95 101 103 112 96 98
R, Reference; T, test; k,, elimination rate constant; t,,,, elimination half-life; residence time; Vss, steady-state volume of distribution. *Distribution-free point estimate and 90% confidence interval.
years (median, 23 years) and weights ranged from 64 to 88 kg (median, 77 kg). They signed a written informed consent form in accordance with the signed informed consent recommendations of the U.S. Food and Drug Administration. All participants were judged to be healthy by a physician on the basis of prestudy medical history, normal physical prestudy and poststudy examinations, electrocardiogram, and laboratory tests, including a test on creatinine clearance. No other medication was taken for 2 weeks before and during the study. No alcohol or xanthine-containing food or drinks were taken for 72 and 48 hours, respectively, before and during the study. On entry to the study, a urine sample was collected for a drug screening. All test results were negative.
0.5027 ± 0.0382 0.5020 ± 0.0345 100 (94, 108)*
C. average concentration; AUC, area under the curve: CL. clearance: MAT, mean
Study design The study was an open, two-period crossover trial with two study periods of 8 days and no washout. In each study period a total of 1200 mg anhydrous theophylline was administered as a rapid-loading infusion of 370 mg during 30 minutes, starting at 8 AM on day 6, followed by a constant-rate infusion of 830 mg during 231/2 hours. In one of the study periods 300 mg caffeine was administered orally as immediate-release 50 mg tablets at 8 AM, 2 PM, and 8 PM from day 1 until 2 PM on day 8 (test). The caffeine-free period served as reference. Blood samples (6 ml) were obtained (before administration) at 8 AM on days I to 6 and at 10-, 20-, and 30-minute and 1-hour intervals until 12 hours, 2-hour intervals until 48 hours, and 4-hour intervals until 60
CLIN PHARMACOL THER MARCH 1991
250 Jonkman et al. hours after the start of the infusion (at 8 AM on day 6). Plasma was collected after centrifugation at 1500 g and stored at -20° C until analysis. Plasma samples were analyzed for theophylline and caffeine. Urine samples were obtained from the 24-hour interval starting at 8 AM on day 5 and from 2- to 4-hour intervals starting at 8 AM on day 6 until 8 PM on day 8. Samples were stored at 4° to 6° C until the end of a urine-sampling interval and then stored at -20° C until analysis. Urine samples were analyzed for theophylline and major metabolites. To facilitate collection of urine samples, the volunteers were given 150 ml water every 2 hours starting at 8 AM on day 6. All participants were given a (no xanthine-containing) standard diet throughout the trial. The study protocol was reviewed and approved by the Medical Ethical Committee 'Stichting Beoordeling Ethiek Bio-Medisch Onderzoek' before the beginning of the study.
Study medications Intravenous aminophylline solutions prepared from Euphyllin ampules (240 mg aminophylline/10 ml), batch number 85K14, were from Byk Nederland B.V., Zwanenburg, The Netherlands. Caffeine immediaterelease 50 mg tablets, batch number 85126-F5192, were from Pharmachemie B. V., Haarlem, The Netherlands. Bioanalytic methods Theophylline in plasma. A selective and sensitive HPLC method was used for the determination of theophylline in plasma.16 The method results in good separation of theophylline from other potentially interfering xanthine compounds such as paraxanthine and caffeine. Typical retention times for caffeine, theophylline, theobromine, paraxanthine, and P-hydroxyethyltheophylline were 3.0, 3.6, 4.7, 5.2, and 6.0 minutes, respectively. Characteristic features of the assay method were: detection limit, 0.05 mg L-'; lower limit of quantitation, 0.25 mg L-1; recovery, 88.3% ± 1.8% (mean ± SD); intraday reproducibility (coefficient of variation), 0.8%; interday reproducibility (coefficient of variation), 1.0%; and accuracy, 0.8%. Caffeine in plasma. A selective and sensitive straight-phase HPLC method was used for the determination of caffeine in plasma. To a plasma sample of 200 1.11, 200 ill internal standard solution (D4030, 5 mg L-1, A.B. Draco, Lund, Sweden), 200 pi calibration solution or water, and 200 ill saturated ammonium sulfate was added. The mixed sample preparation was added to 1.6 ml chloroform/isopropanol
(9.5:0.5 vol/vol) and mechanically agitated and centrifuged (500 rpm). The organic layer was evaporated under nitrogen with a Techne Sample Concentrator (Techne, Inc., Princeton, N.J.) at 35° C. After adding 1 ml mobile phase and mixing, 50 ill was injected into the column with a Waters Intelligent Sample Processor, model M710B (Waters Associates, Milford, Mass.). A 3 pt.m microsphere silica cartridge column, 100 by 4.6 mm (Chrompack International B.V., Middelburg, The Netherlands), was used for separation. A Kratos Spectroflow 783 detector (Kratos Analytical Instruments, Ramsey, N.J.) set at 280 nm was used with a Spectra-Physics SP 4290 integrator (SpectraPhysics Inc., San Jose, Calif.) for quantitation. The mobile phase consisted of a mixture of methanol, acetic acid, tetrahydrofuran, chloroform, and n-heptane (80:1:12.5:1100:820 vol/vol), which was filtered and degassed. The flow rate was 1.0 ml min-1. Retention times for caffeine, theophylline, D-4030 (internal standard), theobromine, paraxanthine, and 13-hydroxyethyltheophylline were 2.3, 3.0, 3.7, 4.0, 4.8, and 5.6 minutes, respectively. Calibration curves were obtained from blank plasma samples spiked with standard of caffeine (0.5 to 12.0 mg L-1). The method shows the following characteristics: lower limit of quantitation, 0.5 mg L-1; interday reproducibility (coefficient of variation), less than 3.2%; and accuracy, less than 0.9%. Theophylline and major metabolites in urine. The method consists of a rapid and highly selective procedure for the determination of theophylline and its major metabolites (i.e., 1,3-dimethyluric acid, 1-methyluric acid, 3-methylxanthine, and 1-methylxanthine). A combination of normal and ion-pair liquid-liquid extraction followed by reversed-phase ion-pair gradient elution HPLC with a 3 p.m Pecosphere C-8 cartridge column 83 by 4.6 mm (Perkin-Elmer, Norwalk, Conn.) for separation was used. Detection was performed by measurement of the ultraviolet absorbance at 280 nm. The method is based on one developed earlier by Muir et al. 17 The method shows the following characteristics: lower limit of quantitation of all compounds (2 mg L-1), recovery (3-methylxanthine, 86%; 1,3dimethylxanthine, 86%; 1-methyluric acid, 67%; and 1,3-dimethyluric acid, 78%), and intraday reproducibility (coefficient of variation) (3-methylxanthine, 6.6%; 1,3-dimethylxanthine, 7.7%; 1-methyluric acid, 9.1%; and 1,3-dimethyluric acid, 8.3%).
Pharmacokinetics and statistics The following pharmacokinetic characteristics were determined for theophylline in plasma: the average
VOLUME 49 NUMBER 3
Effect of caffeine on theophylline
steady-state concentration from 1 to 24 hours after the start of the first theophylline infusion Cav= C av(1-24 hr) and, to detect possible day and night differences in clearance Cav(1-12 hr) and Cav(12-24 hr), the Cav values were obtained by dividing the respective areas under the curve (AUCs) (trapezoidal method) by the corresponding time intervals. Because of the superposition of the short-term infusion for 1/2 hour and the subsequent long-term infusion for 231/2 hours, steady-state concentrations cannot be expected within the first hour, which therefore were skipped in the calculation of the average steady-state concentrations. The theophylline elimination rate constant (k,z) was estimated by monoexponential curve fitting from 24 to at least 46 hours (generally 60 hours) after start of the infusion (at least 12 and generally 16 data points), the t2 was calculated as t12 = 1n2/k,,,. The AUC was determined by the trapezoidal method up to 60 hours after start of the infusion and then extrapolated to infinity by adding C6oh,./k,. The mean residence time (MRT) was calculated as ratio of the area under the moment curve (AUMC) and AUC: MRT = AUMC0f/AUC0-1f in which AUMCO-Inf
= AUMC0_60 hr
60 C60 hlikm,z
C60
MRT was then adjusted for the residence time in the infusion device's: MRTadjusted = MRT
MRTinfusion device
In the case of two subsequent infusions of doses D, and D2 during infusion times T, and T2, respectively, one obtains the followine:
MRT;
device
= D i/(D + D2) T112 + DAD, + D2) (T, + T2/2)
Theophylline CL was calculated as CL = dose/ AUC, the apparent volume of distribution at steady state as Vss = CL MRTadjusted 18 Pharmacokinetic characteristics as obtained with (test) and without caffeine administration (reference) were analyzed statistically by the distribution-free crossover analysis for the multiplicative model, including 90% confidence intervals for the ratio 100 test/reference.2° Equivalence between test and reference can be concluded if the 90% confidence interval is in the bioequivalence range (parametric2I and nonparametric20).
251
RESULTS Plasma concentration data. Individual pharmacokinetic characteristics together with the ratios 100 test/ reference for theophylline are presented in Table I. During concomitant administration of theophylline and caffeine, a mutual influence on the pharmacokinetic characteristics of both drugs can be seen (Fig. 1). Steady-state levels of caffeine, as indicated by predose plasma levels at 8 AM on days 1 to 6, showed about a twofold rise during the theophylline infusion. On the other hand, the average theophylline plasma concentration during the 1- to 24-hour interval after the start of the infusion showed an increase of about 23% after concomitant caffeine administration (p < 0.01). Under the same conditions the (apparent) CL of theophylline was reduced by 29% (p < 0.01), the (apparent) elimination rate constant was reduced by 31% (p < 0.01), and hence the (apparent) elimination t,,2 was prolonged by 44% (p < 0.01). The MRTadjusted reflected the change in t2 and was prolonged by 42%. The (apparent) Vss was unchanged. Comparison of the average theophylline plasma concentration during the 1- to 12-hour and 12- to 24hour intervals after start of the infusion did not reveal a significant diurnal effect for either treatment. For the treatment without caffeine the Cav(12-24 hr) was somewhat lower than the Cav(1-12 hr) (means of 8.97 and 9.15 mg respectively), whereas for the treatment with caffeine the opposite was found (means of 11.71 and 10.67 mg In the latter case a contribution of the caffeine administration to the higher value of Cav( 12-24 hr) of theophylline seemed likely. However, day and night equivalence can be concluded because the 90% confidence intervals for the ratio of expected medians of night and day ranged from 0.91 to 1.03 and from 1.01 to 1.15, respectively. Urinary excretion data. Urine collections were complete during a 60-hour period, thus allowing for calculation of the cumulative fraction of the dose excreted in the form of unchanged drug and particular metabolites. Although the cumulative excretion-time curves were generally "plateauing" about 48 hours after the start of the infusion, the 1-methylxanthine and 1-methyluric acid metabolites (and to a lesser degree the 3-methylxanthine metabolite) after coadministration of caffeine were excreted at significant amounts 60 hours after the start of the infusion (Figs. 2 and 3). The total urinary recovery of theophylline and its metabolites amounted on a molar basis to 77.2% (67.0% to 87.0%) (median; range) without caffeine and 135.1% (124.2% to 145.5%) with caffeine coadminis-
k,
L,
L',
CLIN EHARMACOL THER MARCH 1991
252 Jonkman et al.
12
24
36
60
48
Time (h) Theophylline with caffeine
e-e-e Theophylline without caffeine Caffeine
Fig. 1. Mean plasma concentrations of theophylline (1,3-dimethylxanthine) and caffeine (1,3,7trimethylxanthine) after a total of 1200 mg anhydrous theophylline, administered as a short-term loading infusion followed by a constant-rate infusion during the 24-hour interval starting at 8 AM on day 6, with and without 300 mg caffeine t.i.d., administered orally from day until 2 PM on day 8. (From 48 to 56 hours, hypothetical curve for caffeine; for details see Study Design and Dosage.) 1
0
12
36
24
48
60
Time (h) 0
-
E.-6-e
1,3 1,3
DMX no caffeine
1,3-DMX caffeine
- DMU no caffeine *-*- 1,3- DMU caffeine
3-MX no
caffeine
A-*-i 3-MX
caffeine
Fig. 2. Cumulative renal excretion of theophylline (1,3-DMX), 1,3-dimethyluric acid (1,3DMU), and 3-methylxanthine (3-MX). (For details see legend of Fig. 1.)
VOLUME 49 NUMBER 3
Effect of caffeine on theophylline
0
12
36
24
48
253
60
Time (h) s,
a
1
