Journal of Clinical Pharmacy and Therapeutics (1991) 16,285-289. ADONIS 026947279 1000404
PHARMACOKINETICS OF INTRAVENOUS CAFFEINE IN CRITICALLY ILL PATIENTS A. Shnchez-Alcaraz, P. Ibziiiez* and G. Sangrador Department of Pharmaceutical Services and *Intensive Care Unit, Lluis Alcanyis Hospital, Xativa, Spain
SUMMARY
T h e pharmacokinetics of intravenous caffeine used to measure liver function were evaluated in 20 critically ill patients. Each patient received a single dose of 3.0 mg/kg of caffeine benzoate as a 30-min i.v. infusion. Caffeine serum concentrations were analysed by an enzyme-multiplied immunoassay technique (EMIT). Caffeine pharmacokinetics fitted an open one-compartment model. T h e mean value for the half-life (t,,J was 9.46 k 4.32 h, the volume of distribution was 0.55 k 0 . 1 3 l/kg, and the plasma clearance ( C l ) was 0.85 If:0.44 ml/min/kg. T h e pharmacokinetics parameters of caffeine in critically ill patients compared with normal volunteers were characterized by a reduction in plasma clearance and prolongation in plasma half-life, whereas the volume of distribution remained unchanged. INTRODUCTION Caffeine (1,3,7-trimethylxanthine), an exogenous substance which is extensively consumed, is metabolized almost exclusively in the liver. Its pharmacokinetics in normal man are now well elucidated (1) and its plasma clearance in subjects with cirrhosis is delayed (2). I n recent years it has been recognized that various xenobiotics, particularly those shown to be carcinogenic, are potent inducers of the hepatic cytochrome P-448 system. As a consequence, the availability of a non-invasive test to assess the activity of the cytochrome P-448 system would be highly desirable. Caffeine represents a safe and sensitive indicator of the activity of the cytochrome P-448 system ( 3 ) . Thus, previous studies showed that after administration of a standard dose of caffeine plasma clearance is closely correlated with indicators of hepatic blood flow, microsomal function and acetylator status (4-7). Compared to mean values for caffeine plasma clearance (2.02 & 0.68 ml/min/kg) and plasma half-life (3.8 If:0.9 h) in normal volunteers, cirrhotics were characterized by highly significant reductions in plasma clearance (to 0.67 & 0.42 ml/min/kg) and prolongation in plasma half-life (to 13.7 k 13.0 h), whereas the volume of distribution remained relatively unchanged (0.57 0.16 vs 0.64k0.13 l/kg in normals) (5). T h e aims of the present study were to verify the caffeine liver function test and to compare the pharmacokinetic data obtained in severely ill patients with those available. Correspondence: G. Sangrador Garcia, Servicio de Farmacia, Hospital Lluis Alcanyis, 46800 Xativa, Espafia.
285
286
A. Sanchez-Alcaraz, P . IbaCez and G. Sangrador Table 1. Demographic data for patients
Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Age (years) 68 75 71 71 77 77 41 80 64 72 65 75 75 64 83 75 18 24 68 74
Sex
F F M M M M F F
F M
M F F M M F M M M M
Height (cm)
Weight (kg)
Diagnosis
161 163 180 180 165 165 170 155 165 164 165 165 164 162 169 160 182 170 160 170
58 68 90 90 70 65 80 41 78 65 65 70 71 45 71 80 90 75 67 74
Cardiac insufficiency Abdominal surgery Pneumonia Pneumonia Pneumonia Pneumonia Myocardial infarction Shock Myocardial infarction Myocardial infarction Pneumonia Heart block Heart block Myocardial infarction Cardiac arrest Abdominal surgery Hydrocarbure intoxication Tricyclic intoxication Pneumonia Shock
MATERIALS AND METHODS
Patients T h e study was conducted in 20 patients, admitted to the Intensive Care Unit (Table 1). All patients evaluated had stable renal function and creatinine clearance > 50 ml/min. More of the patients had cirrhosis or chronic liver disease. T h e mean age was 65.9 years (range 18-83 years), the mean height was 1.68 m (range 1.55-1.82 m), and the mean weight was 72.7 kg (range 41-90 kg).
Measurement of cafJeine serum concentration and protocol f o r the pharmacokinetics studies Vials containing a sterile solution of 40 mg caffeine and 40 mg sodium benzoate per 1 ml were prepared in the hospital pharmacy. All the vials were stored at 4°C until used. All patients received a single 3-mg/kg dose of caffeine, infused intravenously over 30 min. Blood samples were obtained from the contralateral arm before caffeine was administered and at the following times after the infusion was completed (time 0): 0.08, 0 5 , 1-0, 4.0, and 8-0 h. T h e blood was allowed to clot, and the resultant serum was separated by centrifugation and stored at - 20°C until analysed. Caffeine serum concentrations were analysed by an enzyme-multiplied immunoassay technique (EMIT, Syva, Inc.). All samples for a given subject were analysed at the same time with a standard curve using an automated centrifugal analyser (Cobas Mira, Roche Diagnostica). T h e pharmacokinetics of caffeine can be described by a one-compartment model. T h e elimination rate constant ( K )was determined from the slope of the natural log
Pharmacokinetics of caffeine
287
ic
f
-E
\ 0
2
E
c
._0 +
e
+ c W
0 c
0
0
4
E,
a
2
I
I
I
I
I
I
I
I
I
4 5 6 7 a 9 Time ( h ) Fig. 1. Plasma caffeine concentration-time data after administration of 150 mg as a 30-min i.v. infusion (11=20). 0
I
2
3
Table 2. Pharmacokinetic parameters for caffeine in critically ill patients (n = 20) k (min
f:
'1
(h)
Mean
0,0015
SD
0.0007
9.5 4.3
Vd (l/kg) 0.55 0.13
AUC CL (mg/min/l) (rnljminlkg) 4840 2817
0.85 0.44
concentration-time curve using a least-squares linear regression method. T h e elimination half-life ( t i )was defined as the ratio of the quotient of the natural log of 2 and K . T h e apparent volume of distribution ( Vd)was calculated using the ratio dose (mglkg)lCo(mgll) where C, is the extrapolated plasma level at time zero.
288
A. Sanchez-Alcaraz, P. Ibanlez and G. Sangrador
T h e area under the concentration-time curve (AUC) was calculated using the trapezoidal rule. Total body clearance (CZ) was defined as the quotient of the intravenous caffeine dose and the AUC. RESULTS AND DISCUSSION Figure 1 shows the mean caffeine plasma level-time curve after the 30-min i.v. infusion. Mean caffeine plasma concentration values at 0,0.5,4 and 8 h after administration were 6.8f 1.8, 5.0f 1.1, 4.8f 1.2, 4.1 1.4 and 3.2+ 1.2 pg/ml, respectively. Caffeine concentrations decreased from a peak concentration of 6.8 pg/ml to a mean concentration of 3-2 pg/ml at 8 h after termination of the i.v. infusion. Table 2 gives the values of rate constant elimination ( K ) ,plasma half-life (ti),distribution volume (V,), area under the curve (AUC) and plasma clearance (CZ).T h e mean value of the ti was 9.5 +_ 4.3 h (range 3.7-20.4 h) with a coefficient of variation of 45.6O/& showing high interpatient variability in the rate of elimination. These results compared with those reported by Renner et aZ. (5) in normal volunteers, indicate that the ti is two to three times higher in our critically ill patients. T h e distribution volume was 0.55 k 0.13 l/kg (range 0.35-0.76 l/kg). These values are similar to those reported by other authors (5); 0.57+0.161/kg in patients with liver disease and 0.64 & 0.13 l/kg in normals. T h e mean value for caffeine plasma clearance was 0.85 5 0.44 ml/min/kg with a high coefficient of variation of 52.1%. T h e caffeine plasma clearance in our critically ill patients is ( < 0-001) therefore significantly different to values reported for normal patients [2.02 i0.68 ml/min/kg (5) or 1-70k0.50 ml/min/kg (7)]. Whether the differences are significant when compared with cirrhotic patients depends on which reported values (5,7) are used [0.39 0.10 ml/min/kg ( P < O . O O l ) and 0.67k0.42 ml/min/kg ( P > O.l)]. T h e distribution volume is the pharmacokinetic parameter with the lowest coefficient of variation (20-24%) and shows minor interpopulation differences (critically ills vs. cirrhotics t = 0.45; P > 0.1 and vs volunteers t = 1.78; P > 0.1). Therefore to evaluate the liver function, the measure of two plasma levels after infusion ( 2 and 8 h) is sufficient to estimate the plasma clearance of caffeine assuming a constant value of 0.6 l/kg for a distribution volume.
+
CONCLUSION T h e pharmacokinetic parameters of caffeine in critically ill patients, compared with normal volunteers, were characterized by a reduction in plasma clearance and a prolongation in plasma half-life, whereas the volume of distribution remained unchanged. T h e plasma clearance of caffeine in critically ill patients without cirrhosis is closer to that of cirrhotic population than to that of normal volunteers. T h e evaluation of caffeine plasma levels and its clearance provides quantitative information on liver function. REFERENCES
1. Bonati, M., Latini, R., Galletti, F., et al. (1982) Caffeine disposition after oral doses. Clinical Pharmacology and Therapeutics, 32,98-106, 2. Desmond, P., Patwardhan, R.V., Johnson,R.F., et al. (1980) Impaired elimination of caffeine in cirrhosis. Digestive Disease Science, 25, 193-197.
Pharmacokinetics of caffeine
289
3. Wietholtz, H., Voegelin, M., Arnaud, M.J., et al. (1981) Assessment of the cytochrome P-448 dependent liver enzyme system by a caffeine breath test. European Journal of Clinical Pharmacology, 21, 53-59. 4. Zysset, T., Wahllander, A. & Preisig, R. (1984) Evaluation of caffeine plasma levels by an automated enzyme immunoassay (EMIT) in comparison with a high-performance liquid chromatographic method. Therapeutic Drug Monitoring, 6,348 - 354. 5. Renner, E., Wietholtz, H., Huguenin, P., et al. (1984) Caffeine: A model compound for measuring liver function. Hepatology, 4,38 -46. 6. Rankin, R.B., Hudson, S.A. & Fell, A.F. (1987) Caffeine as a potential indicator for acetylator status. Journal of Clinical Pharmacy and Therapeutics, 12,47-56. 7. Jost, G., Mandach, U.V., Wahllander, A. & Preisig, R. (1984) Differential effect of liver disease on the cytochrome P-450 and P-448 systems. European Association for the Study of the Liver. 19th Meeting, 6th-8th September 1984. Berne.
PHARMACOKINETICS OF INTRAVENOUS CAFFEINE IN CRITICALLY ILL PATIENTS A. Shnchez-Alcaraz, P. Ibziiiez* and G. Sangrador Department of Pharmaceutical Services and *Intensive Care Unit, Lluis Alcanyis Hospital, Xativa, Spain
SUMMARY
T h e pharmacokinetics of intravenous caffeine used to measure liver function were evaluated in 20 critically ill patients. Each patient received a single dose of 3.0 mg/kg of caffeine benzoate as a 30-min i.v. infusion. Caffeine serum concentrations were analysed by an enzyme-multiplied immunoassay technique (EMIT). Caffeine pharmacokinetics fitted an open one-compartment model. T h e mean value for the half-life (t,,J was 9.46 k 4.32 h, the volume of distribution was 0.55 k 0 . 1 3 l/kg, and the plasma clearance ( C l ) was 0.85 If:0.44 ml/min/kg. T h e pharmacokinetics parameters of caffeine in critically ill patients compared with normal volunteers were characterized by a reduction in plasma clearance and prolongation in plasma half-life, whereas the volume of distribution remained unchanged. INTRODUCTION Caffeine (1,3,7-trimethylxanthine), an exogenous substance which is extensively consumed, is metabolized almost exclusively in the liver. Its pharmacokinetics in normal man are now well elucidated (1) and its plasma clearance in subjects with cirrhosis is delayed (2). I n recent years it has been recognized that various xenobiotics, particularly those shown to be carcinogenic, are potent inducers of the hepatic cytochrome P-448 system. As a consequence, the availability of a non-invasive test to assess the activity of the cytochrome P-448 system would be highly desirable. Caffeine represents a safe and sensitive indicator of the activity of the cytochrome P-448 system ( 3 ) . Thus, previous studies showed that after administration of a standard dose of caffeine plasma clearance is closely correlated with indicators of hepatic blood flow, microsomal function and acetylator status (4-7). Compared to mean values for caffeine plasma clearance (2.02 & 0.68 ml/min/kg) and plasma half-life (3.8 If:0.9 h) in normal volunteers, cirrhotics were characterized by highly significant reductions in plasma clearance (to 0.67 & 0.42 ml/min/kg) and prolongation in plasma half-life (to 13.7 k 13.0 h), whereas the volume of distribution remained relatively unchanged (0.57 0.16 vs 0.64k0.13 l/kg in normals) (5). T h e aims of the present study were to verify the caffeine liver function test and to compare the pharmacokinetic data obtained in severely ill patients with those available. Correspondence: G. Sangrador Garcia, Servicio de Farmacia, Hospital Lluis Alcanyis, 46800 Xativa, Espafia.
285
286
A. Sanchez-Alcaraz, P . IbaCez and G. Sangrador Table 1. Demographic data for patients
Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Age (years) 68 75 71 71 77 77 41 80 64 72 65 75 75 64 83 75 18 24 68 74
Sex
F F M M M M F F
F M
M F F M M F M M M M
Height (cm)
Weight (kg)
Diagnosis
161 163 180 180 165 165 170 155 165 164 165 165 164 162 169 160 182 170 160 170
58 68 90 90 70 65 80 41 78 65 65 70 71 45 71 80 90 75 67 74
Cardiac insufficiency Abdominal surgery Pneumonia Pneumonia Pneumonia Pneumonia Myocardial infarction Shock Myocardial infarction Myocardial infarction Pneumonia Heart block Heart block Myocardial infarction Cardiac arrest Abdominal surgery Hydrocarbure intoxication Tricyclic intoxication Pneumonia Shock
MATERIALS AND METHODS
Patients T h e study was conducted in 20 patients, admitted to the Intensive Care Unit (Table 1). All patients evaluated had stable renal function and creatinine clearance > 50 ml/min. More of the patients had cirrhosis or chronic liver disease. T h e mean age was 65.9 years (range 18-83 years), the mean height was 1.68 m (range 1.55-1.82 m), and the mean weight was 72.7 kg (range 41-90 kg).
Measurement of cafJeine serum concentration and protocol f o r the pharmacokinetics studies Vials containing a sterile solution of 40 mg caffeine and 40 mg sodium benzoate per 1 ml were prepared in the hospital pharmacy. All the vials were stored at 4°C until used. All patients received a single 3-mg/kg dose of caffeine, infused intravenously over 30 min. Blood samples were obtained from the contralateral arm before caffeine was administered and at the following times after the infusion was completed (time 0): 0.08, 0 5 , 1-0, 4.0, and 8-0 h. T h e blood was allowed to clot, and the resultant serum was separated by centrifugation and stored at - 20°C until analysed. Caffeine serum concentrations were analysed by an enzyme-multiplied immunoassay technique (EMIT, Syva, Inc.). All samples for a given subject were analysed at the same time with a standard curve using an automated centrifugal analyser (Cobas Mira, Roche Diagnostica). T h e pharmacokinetics of caffeine can be described by a one-compartment model. T h e elimination rate constant ( K )was determined from the slope of the natural log
Pharmacokinetics of caffeine
287
ic
f
-E
\ 0
2
E
c
._0 +
e
+ c W
0 c
0
0
4
E,
a
2
I
I
I
I
I
I
I
I
I
4 5 6 7 a 9 Time ( h ) Fig. 1. Plasma caffeine concentration-time data after administration of 150 mg as a 30-min i.v. infusion (11=20). 0
I
2
3
Table 2. Pharmacokinetic parameters for caffeine in critically ill patients (n = 20) k (min
f:
'1
(h)
Mean
0,0015
SD
0.0007
9.5 4.3
Vd (l/kg) 0.55 0.13
AUC CL (mg/min/l) (rnljminlkg) 4840 2817
0.85 0.44
concentration-time curve using a least-squares linear regression method. T h e elimination half-life ( t i )was defined as the ratio of the quotient of the natural log of 2 and K . T h e apparent volume of distribution ( Vd)was calculated using the ratio dose (mglkg)lCo(mgll) where C, is the extrapolated plasma level at time zero.
288
A. Sanchez-Alcaraz, P. Ibanlez and G. Sangrador
T h e area under the concentration-time curve (AUC) was calculated using the trapezoidal rule. Total body clearance (CZ) was defined as the quotient of the intravenous caffeine dose and the AUC. RESULTS AND DISCUSSION Figure 1 shows the mean caffeine plasma level-time curve after the 30-min i.v. infusion. Mean caffeine plasma concentration values at 0,0.5,4 and 8 h after administration were 6.8f 1.8, 5.0f 1.1, 4.8f 1.2, 4.1 1.4 and 3.2+ 1.2 pg/ml, respectively. Caffeine concentrations decreased from a peak concentration of 6.8 pg/ml to a mean concentration of 3-2 pg/ml at 8 h after termination of the i.v. infusion. Table 2 gives the values of rate constant elimination ( K ) ,plasma half-life (ti),distribution volume (V,), area under the curve (AUC) and plasma clearance (CZ).T h e mean value of the ti was 9.5 +_ 4.3 h (range 3.7-20.4 h) with a coefficient of variation of 45.6O/& showing high interpatient variability in the rate of elimination. These results compared with those reported by Renner et aZ. (5) in normal volunteers, indicate that the ti is two to three times higher in our critically ill patients. T h e distribution volume was 0.55 k 0.13 l/kg (range 0.35-0.76 l/kg). These values are similar to those reported by other authors (5); 0.57+0.161/kg in patients with liver disease and 0.64 & 0.13 l/kg in normals. T h e mean value for caffeine plasma clearance was 0.85 5 0.44 ml/min/kg with a high coefficient of variation of 52.1%. T h e caffeine plasma clearance in our critically ill patients is ( < 0-001) therefore significantly different to values reported for normal patients [2.02 i0.68 ml/min/kg (5) or 1-70k0.50 ml/min/kg (7)]. Whether the differences are significant when compared with cirrhotic patients depends on which reported values (5,7) are used [0.39 0.10 ml/min/kg ( P < O . O O l ) and 0.67k0.42 ml/min/kg ( P > O.l)]. T h e distribution volume is the pharmacokinetic parameter with the lowest coefficient of variation (20-24%) and shows minor interpopulation differences (critically ills vs. cirrhotics t = 0.45; P > 0.1 and vs volunteers t = 1.78; P > 0.1). Therefore to evaluate the liver function, the measure of two plasma levels after infusion ( 2 and 8 h) is sufficient to estimate the plasma clearance of caffeine assuming a constant value of 0.6 l/kg for a distribution volume.
+
CONCLUSION T h e pharmacokinetic parameters of caffeine in critically ill patients, compared with normal volunteers, were characterized by a reduction in plasma clearance and a prolongation in plasma half-life, whereas the volume of distribution remained unchanged. T h e plasma clearance of caffeine in critically ill patients without cirrhosis is closer to that of cirrhotic population than to that of normal volunteers. T h e evaluation of caffeine plasma levels and its clearance provides quantitative information on liver function. REFERENCES
1. Bonati, M., Latini, R., Galletti, F., et al. (1982) Caffeine disposition after oral doses. Clinical Pharmacology and Therapeutics, 32,98-106, 2. Desmond, P., Patwardhan, R.V., Johnson,R.F., et al. (1980) Impaired elimination of caffeine in cirrhosis. Digestive Disease Science, 25, 193-197.
Pharmacokinetics of caffeine
289
3. Wietholtz, H., Voegelin, M., Arnaud, M.J., et al. (1981) Assessment of the cytochrome P-448 dependent liver enzyme system by a caffeine breath test. European Journal of Clinical Pharmacology, 21, 53-59. 4. Zysset, T., Wahllander, A. & Preisig, R. (1984) Evaluation of caffeine plasma levels by an automated enzyme immunoassay (EMIT) in comparison with a high-performance liquid chromatographic method. Therapeutic Drug Monitoring, 6,348 - 354. 5. Renner, E., Wietholtz, H., Huguenin, P., et al. (1984) Caffeine: A model compound for measuring liver function. Hepatology, 4,38 -46. 6. Rankin, R.B., Hudson, S.A. & Fell, A.F. (1987) Caffeine as a potential indicator for acetylator status. Journal of Clinical Pharmacy and Therapeutics, 12,47-56. 7. Jost, G., Mandach, U.V., Wahllander, A. & Preisig, R. (1984) Differential effect of liver disease on the cytochrome P-450 and P-448 systems. European Association for the Study of the Liver. 19th Meeting, 6th-8th September 1984. Berne.
