European Journal of Clinical Pharmacology
Europ. J. clin. Pharmacol. 12, 387-392 (1977)
© by Springer-Verlag 1977
Overall Pharmacokinetics During Prolonged Treatment of Healthy Volunteers with Digoxin and fl-Methyldigoxin* F. Keller, H. P. Blumenthal, K. Maertin, and N. Rietbrock Department of Clinical Pharmacology, Free University of Berlin, Federal Republic of Germany, and Department of Pharmaceutics, Clinical Pharmacokinetics Laboratory, State University of New York at Buffalo, NY, USA
Summary. Five healthy volunteers received digoxin 0.4 mg or fl-methyldigoxin 0.4 mg i.v., daily for 14 days, in a randomized cross-over arrangement. By monitoring minimal plasma concentrations during multiple dosing, it was found that the steady state pharmacokinetics of digoxin and /3-methyldigoxin could be estimated even better by a one-compartment than by a two-compartment model. The following mean parameters were calculated: the half life of digoxin of 1.54 + 0.31 days was significantly shorter than the half life of 2.29 _+ 0.34 days for /3-methyldigoxin. The distribution volume of 807 + 187 liters for digoxin was not significantly larger than the 735 _+ 227 liters for/3-methyldigoxin. Renal digoxin clearance of 191 + 25 ml/min was significantly higher than both the renal clearance of/3methyldigoxin of 111 + 23 ml/min and also the creatinine clearance, which indicates tubular secretion of digoxin. There was a 2.8-fold accumulation of/3-methyldigoxin injected once a day, which was significantly higher than the 1.8-fold accumulation of digoxin. Key words: Digoxin, /3-methyldigoxin, prolonged administration, pharmacokinetics.
From single dose studies a two-compartment model (Reuning et al., 1973) and even a three-compartment model (Kramer et al., 1974) have been found necessary to describe the pharmacokinetics of digoxin. The existence of a deep compartment for digoxin in the CNS has been demonstrated in the dog
* This study was supported by Bundesministerium fiir Jugend, Familie und Gesundheit, Federal Republic of Germany.
(Rietbrock and Kuhlmann, 1977). On the other hand, digoxin is mainly employed for extended treatment, and the physician should have simple but correct equations which permit ready calculation of the required dose. A pharmacokinetic concept of digitalis glycoside accumulation may be used to advantage in clinical practice if the glycoside concentration in plasma and its elimination can be described by equations appropriate to a one-compartment model (Dettli, 1975). The underlying assumptions are: 1. a constant and time-independent relationship between the plasma concentration and the total amount of drug in the body; 2. very rapid uptake compared to very slow elimination rate, and 3. proportional relationship between the amount of drug excreted over a period of time and the amount remaining in the body (Dettli, 1975).
Methods Five healthy volunteers (Table 1) took part in the study, and they gave their informed consent to the investigation. To use the same units as the later digoxin clearance evaluated subsequently the creatinine clearance of the volunteers was not standardized to a 1.73 m 2 body surface area. The other Table 1. Details of healthy subjects studied Subject
P.B.F.K.B.H.
B.K. W.S. mean_+SD
Age (years) Body weight (kg) Height(cm) SEX Creatinineclearance (ml/min)
22 59 175 ~
27 58 175 CY
27 70 178 CY
25 69 175 CY
24 65 t77 CY
25+2 64+6 176+1
88
92
186
126
188
136+44
388
F. Keller et al.: Pharmacokinetics of Digoxin and ~Methytdigoxin 4
In[10xncj;/ml]
1
7
14
days
19
Fig. 1. Plasma concentrations (ng/ml) in 5 volunteers after multiple doses of digoxin 0.4 mg ( * ~ * ) or fl-methyldigoxin 0.4 mg (o . . . . . . o) daily for 14 days i. v.
6
I. [ ~/d~] 5
4
3
2
i
1
7
14
2t
28
Fig. 2. Renal excretion 0tg/die) by 5 volunteers after multiple doses of digoxin 0.4 mg (**) or fl-methyldigoxin 0.4 mg (o . . . . . . o) daily for 14 days i.v.
Table 2. Pharmacokinetic parameters of digoxin and fi-methyldigoxin using a one-compartment open model Digoxin
fi-methyldigoxin
Subject
P.B.
F.K.
B. H~
B, K.
W, S.
mean _+ SD
EST KTOT KREN VOL CTOT CREN CCR CR/CT CR/CCR T/2 BMIN BMIN/D
0.50 0.581 0.369 595 240 153 88 0.64 1.74 1.19 507 1.27
0,44 0.490 0.320 814 277 181 92 0.65 1,96 1.41 632 1.58
0.39 0,477 0,409 686 227 195 186 0.86 1.05 1.45 655 1.64
0.33 0.343 0,273 1085 258 206 126 0.80 1.63 2.02 979 2.45
0.46 0.422 0,369 854 250 219 188 0.87 1.16 1.64 761 1.90
0.42 0,423 0.348 807 250 191 136 0.76 1.51 1,54 707 1,77
0.07 0.088 0,053 187 19 25 49 0.11 0.39 0.31 177 0.44
P.B.
F.K.
B.H.B.K.
W. S, mean -+ SD
0.36 0,372 0.255 482 t24 85 88 0.69 0.97 1.86 887 2.22
0.36 0.350 0.248 529 128 91 92 0.71 0.99 1.98 955 2.39
0.38 0.280 0,214 755 147 112 186 0.77 0,60 2.48 1239 3.10
0.26 0.274 0.205 920 175 131 188 0.75 0.69 2.53 1269 3.17
0.22 0.266 0,197 991 183 135 126 0,74 1.07 2.60 1310 3.28
0.32 0.308 0.224 735 151 111 136 0.73 0.86 2.29 1132 2.83
0.07 0,049 0.026 227 27 23 49 0,03 0.21 0.34 196 0.49
p L
0.005 0.0t n.s. 0.005 0.001 n.s. 0.005 0.005 0.005
389
F. Keller et al.: Pharmacokinetics of Digoxin and/3-Methyldigoxin
reason for the large discrepancies between the individual creatinine clearances, even though they were estimated three times, was the fact that B.H. and W.S. physically exerted themselves in the morning before coming to the laboratory (Table 1). Digoxin 0.4 mg or fl-methyldigoxin 0.4 mg were given intravenously once daily for 14 days. Each subject received both glycosides in a random sequence at a 14 day interval to ensure almost complete elimination of glycoside from the preceding study. Blood samples from each subject were collected in heparinized tubes 24 h after administration. Samples were immediately centrifuged and the plasma removed and stored at - 2 0 ° C until analyzed. Urine was collected for periods of 24 h for the 14 days of treatment and for the 14 days after medication had closed. In subjects P.B. and F.K. urine was collected for 24 days after intravenous administration of fl-methyldigoxin had ceased. After measurement of total volume, aliq u o t s of urine were stored at - 2 0 ° C . Plasma glycoside concentrations were measured by 125Iradioimmunoassay utilizing a commercial kit from Clinical Assays Inc., which had a standard sample reproducibility of 9%, a precision of 6% (for concentrations below 1.0 ng/ml of 16%), a recovery of 5% (for concentrations below 1.0 ng/ml of 12%), a sensitivity (expressed as maximal binding capacity) of 35-40%, and a 12% coefficient of variation in normal plasma, which indicates its specificity. Urinary glycoside concentrations were measured by the 3H-radioimmunoassay reported by Maertin and Rietbrock (1977). The pharmacokinetic parameters were estimated by means of a nonlinear, least squares regression program, NONLIN (Metzler, 1969). The calculations employed equations describing the time course of the plasma concentration after multiple dosing for a one- or two-compartment model and the urinary excretion rate of the drug (Gibaldi et al., 1975, equations 359, 419 and 14). Both types of data were fitted sumultaneously and all data points were given an equal numerical weight. Statistical analysis included mean, standard deviation and Student's t-test for paired trials.
life (6 days for digoxin and 9 days for fl-methyldigoxin, see Table 2) the plasma concentrations were within 10% of their steady state levels. The estimated steady state plasma levels of fl-methyldigoxin were significantly higher (p < 0.005) than those after digoxin, 1.81 + 0.33 ng/ml and 0.90 + 0.17 ng/ml, respectively. But the average renal excretion of digoxin and fl-methyldigoxin during a steady state interval did not differ significantly, 298 + 30 ~tg and 296 _+ 27 ~tg, respectively. Urinary excretion rates over a period of 24 days after cessation of/3-methyldigoxin treatment in subjects P.B. and F. K. are shown in Figure 3. From the 28th day the terminal excretion of glycoside was retarded, which reflects a slowly equilibrating tissue pool. The pharmacokinetic parameters computed with a one-compartment model showed very close correlation to the empirical data (Figs. 4 and 5). Urinary excretion rate data yielded better fits (r = 0.996 for digoxin and/3-methyldigoxin) than did plasma concentration data (r -- 0.974 for fl-methyldigoxin and r = 0.939 for digoxin), indicating greater reliability of urinary excretion rate data. The average half life of digoxin was 1.54 days, which was distinctly lower than the 2.29 day half life of fl-methyldigoxin (Table 2). Since the volumes of distribution did not differ, the total clearance of the latter was 1.66 times that of digoxin. The renal clearance of fl-methyldigoxin was distinctly lower than the creatinine clearance, whereas the renal clearance of digoxin exceeded the creatinine clearance by about 40 percent. Comparison of the ratio of the minimum amount at steady state in the body and the intravenous dose of digoxin and fl-methyldigoxin permitted estimation of the extent of accumulation. The accumulation factor which is usually calculated from the elimination rate constant and the dosing interval, the ratio of BMAX/D or (BMIN + D)/D (Gibaldi et al., 1975), should not be identified with the extent of accumulation as used here.
Discussion Results
The plasma concentrations and the renal excretion of digoxin and/3-methyldigoxin in 5 volunteers are shown in Figs. 1 and 2. The rate and extent of accumulation of digoxin and fl-methyldigoxin was a function of the magnitude of the dosing interval and the half life of the two glycosides. After multiple dosing for a period equal to four times the half
The present study is believed to be one of the first to present empirical and pharmacokinetic data about plasma levels and renal excretion of digoxin during prolonged treatment as employed in clinical practice, and which verifies Augsberger's concept based upon pharmacodynamic observations. The intention was to reassess the pharmacokinetic characteristics of digoxin and /3-methyldigoxin in healthy volunteers after multiple dosing. In regard to clinical
390
F. Keller et al.: Pharmacokinetics of Digoxin and/3-Methyldigoxin 6
In[~g/die]
5
"'::':i"..
. . . . .
"'..
.
"'""i'. %1. ::,
':::ii/:;:::: .................... "..... ........ i:...... :].-"
14
21
28
digoxin
B-methyldigoxin
2.0
2.0
NG/ML
/
r = .939
~,0
0.5
/ ' 01,5
NGIML r = .974
1.5
jo/S
'L.O
0.5
0.0
0.0
0
35
days
Fig. 3. Renal excretion (~tg/die) by subjects P.B. and F.K. after cessation of multiple doses of fl-methyldigoxin
ll.O
II.5
2i, 0
0
0i. 5
11,0
NG/ML
21,0
NG/ML
digoxin
B-methyldigoxin
400
400
~JGIDIE
~1,5
~JGIDIrE
Fig. 4. Correlation of plasma concentrations (ng/ml) between observed (x) and calculated (y) values using a one-compartment open model
.//° /
r = . 996 300
3O0
200
20(]
"1_.00
100
/
4
0 1001
2001
3001 400 I )JG/DIE
'tOOI
200 ~
3001
400 I
,,UO/O[E
Fig. 5. Correlation of renal excretion (btg/day) between observed (x) and calculated (y) values using a one-compartment open model
F. Keller et al.: Pharmacokinetics of Digoxin and/3-Methyldigoxin
391
applicability a one-compartment model will be more successful than a complex multi-compartment model. Data supporting use of such a one-compartment model include the correlation between computed body glycoside level and measured plasma glycoside concentration, and the correlation between computed body glycoside level and measured myocardial digitalis concentration at autopsy (Jelliffe, 1967; Jelliffe et al., 1972). From single dose studies Reuning et at., (1973) postulated that a two-compartment open model was necessary to account for digoxin pharmacokinetics and Kramer et al., (1974) indicated that a threecompartment open model gave a significantly improved fit. However, after multiple doses and by monitoring minimal plasma concentrations as well as urinary excretion rates, the ability to discern two-compartment kinetics of digoxin and /3-methyldigoxin is decreased. Since the rattio of zero time intercepts A/B = ( a -- k21)/(k21 -- / 3 ) f r o m a two compartment model is low (2.8 for ~methyldigoxin), the plasma concentration time curve is more nearly described by a one-compartment model. In usingthe NONLIN least squares regression program of Metzler (1969) the /3-value in the two-compartment model revealed a very poor coefficient of variation, average 85.5 _+ 54.1 percent for /3-methyldigoxin, whereas the corresponding KTOT-value in the onecompartment model showed a much smaller variation coefficient averaging 3.1 _+ 1.6 percent. Therefore, in the present study a one-compartment model offered a distinct improvement over a two-compartment open model. At the same time, the design of the study came very close to the therapeutic use and clinical monitoring of digitalis glycosides. The observed half lives of digoxin and/3-methyldigoxin encompass the range reported by Rietbrock et al., (1976), which showed a longer half life of/3methyldigoxin than of digoxin. The volumes of distribution, renal clearance and total clearance of digoxin were within the range calculated by a twocompartment open model (Koup et al., 1975). Marked differences were observed between the elimination kinetics of/3-methyldigoxin and digoxin, apparently due to the higher renal clearance of digoxin, which was also significantly (p < 0.025) greater than the creatinine clearance. The assumption of tubular secretion of digoxin is based on data from Steiness (1973), Falch (1973) and Koup et al. (1975). The extent of accumulation of digoxin and fl-methyldigoxin, according to the ratio of the calculated body stores and the maintenance dose just before receiving the next dose during steady state, was much lower for digoxin than for /3-methytdigoxin.
From the urinary excretion rate of digoxin and ~methyldigoxin on the first day, which amounted to 37 + 0.08 percent of the dose after digoxin and 22 + 0.03 percent of the dose after/3-methyldigoxin (p < 0.02), it may be suggested that the elimination rate on the first day was an essential indicator of glycoside accumulation in the body. Therefore, a reduction in the dose of fi-methyldigoxin compared to digoxin is necessary to prevent the greater risk of developing symptoms of glycoside toxicity. The reduction of glycoside pharmacokinetics to a one-compartment model is not devoid of disadvantages. From studies in beagle dogs (Rietbrock and Kuhlmann, 1977) it must be concluded that the brain is a deep compartment. Therefore, glycoside side effects resulting from CNS accumulation may be masked in using only the elimination rate constants of a one-compartment model. The slow release of /3-mthyldigoxin from the brain may have contributed to the increased half life (5.2 days for P.B. and 5.8 days for F. K.) of the urinary excretion rates from the 14th day after cessation of/3-methyldigoxin administration (Fig. 3). The calculated renal excretion rates of fl-methyldigoxin and digoxin agreed well with the observed data (Fig. 5). However, in the case of fi-methyldigoxin calculated plasma concentrations less than 0.78 ng/ml were higher and concentrations more than 0.78 ng/ml were lower than the observed values (Fig. 4). This error probably reflects a rate limiting step in demethylation by liver enzymes (Rietbrock et al., 1972; Zilly et al., 1975). With an increasing plasma concentration of total glycosides, the relative amount of digoxin decreases. These findings would also indicate that unchanged fl-methyldigoxin has a considerably smaller volume of distribution than digoxin. Abbreviations of parameters D EST KTOT KREN VOL ClOT CREN BMIN CCR CR/CT CR/CCR T/2 BMIN/D
dose estimated disposition rate constant (day -~) semilog curve fit total disposition rate constant (day -I) NONLIN fit renal disposition rate constant (day -1) NONLIN fit volume of distribution (liters) NONLIN fit total clearance (ml/min) renal clearance (ml/min) body stores (~tg) before next dose in steady state creatinine clearance (ml/min) fraction renally excreted renal drug clearance versus creatinine clearance half life (days) extent of accumulation
392
F. Keller et al.: Pharmacokinetics of Digoxin and fl-Methyldigoxin
References
Number 7292/69/7292/005. The Upjohn Company, Kalamazoo, Mich. 1969 10. Reuning, R.H., Sams, R.A., Notari, R.E.: Role of pharmacokinetics in drug dosage adjustment. I. Pharmacologic effect kinetics and apparent volume of distribution of digoxin. J. clin. Pharmacol. 13, 127-141 (1973) 11. Rietbrock, N., Rennekamp, C., Rennekamp, H., Bergmann, K.v., Abshagen, U.: Demethylation and cleavage of glycosidic bonds of 4'"-methyldigoxin in man. NaunynSchmiedeberg's Arch. Pharmacol. 272, 450--453 (1972) 12. Rietbrock, N., Guggenmos, J., Kuhlmann, J., Hess, U.: Bioavailability and pharmacokinetics of fl-methyldigoxin after multiple oral and intravenous doses. Europ. J.clin. Pharmacol. 9, 373-379 (1976) 13. Rietbrock, N., Kuhlmann, J.: Clinical significance of distribution and elimination of cardiac glycosides. In: Clinical Pharmacokinetics (ed. Ritschel, W.A.). pp. 57-65. Stuttgart: Gustav Fischer 1977 14. Steiness, E.: Renal excretion of digoxin. In: Symposium on Digitalis. (ed. Storstein, O., Nitter-Hauge, S., Storstein, L.) pp. 178-183. Oslo, Norway: Gyldendal 1973 15. Zilly, W., Richter, E., Rietbrock, N.: Pharmacokinetics and metabolism of digoxin- and /3-methyldigoxin-12-3H in patients with acute hepatitis. Clin. Pharmacol. Ther. 17, 302-309 (1975).
1. Dettli, L.: Pharmacokinetische Grundlagen fiir die optimale Dosierung bei repetierter Applikation yon Arzneimitteln. In: Klinische Arzneimittelpriifung. (ed. Eickstedt, K.W. und Gross, F.) 1. Symposium fiir Klinische Pharmakologie des Bundesgesundheitsamtes Berlin, 1974. pp. 59-68. Stuttgart: Gustav Fischer 1975 2. Falch, D., Teien, A.: The influence of kidney function on the plasma level and urinary excretion of digoxin. In: Symposium on Digitalis (ed. Storstein, O.; Nitter-Hauge, S. and Storstein L.) pp. 183-189. Oslo: Norway, Gyldendal 1973 3. Gibaldi, M., Perrier, D.: Pharmacokinetics. New York: Marcel Dekker, 1975 4. Jelliffe, R.W.: A mathematical analysis of digitalis kinetics in patients with normal and reduced renal function. Math. Biosci. 1, 305-325 (1967) 5. Jelliffe, R.W., Buell, J., Kalaba, R.: Reduction of digitalis toxicity by computer-assisted glycoside dosage regimens. Ann. int. Med. 77, 891-906 (1972) 6. Koup, J.R., Greenblatt, D.J., Jusko, W.J., Smith, T.W., Koch-Weser, J.: Pharmacokinetics of digoxin in normal subjects after intravenous bolus and infusion doses. J. Pharmakokinet. Biopharm. 3, 181-192 (1975) 7. Kramer, W.G., Lewis, R.P., Cobb, T.C., Forester, W.F.Jr., Visconti, J.A., Wanke, L.A., Boxenbaum, H.G., Reuning, R.H.: Pharmacokinetics of digoxin: comparison of a two- and a three-compartment model in man. J. Pharmakokinet. Biopharm. 2, 299-312 (1974) 8. Maertin, K., Rietbrock, N.: Radioimmunologische Bestimmung von Digoxin in Urin. Klin. Wschr. 55, 429--438 (1977) 9. Metzler, C.M.: NONLIN; A computer program for parameter estimation in nonlinear situations. Technical Report
Received; April 4, 1977, in revised form: Juni 28, 1977, accepted: July 4, 1977 Prof. Dr. N. Rietbrock Department of Clinical Pharmacology Hindenburgdamm 30 D-1000 Berlin 45 Federal Republic of Germany
Europ. J. clin. Pharmacol. 12, 387-392 (1977)
© by Springer-Verlag 1977
Overall Pharmacokinetics During Prolonged Treatment of Healthy Volunteers with Digoxin and fl-Methyldigoxin* F. Keller, H. P. Blumenthal, K. Maertin, and N. Rietbrock Department of Clinical Pharmacology, Free University of Berlin, Federal Republic of Germany, and Department of Pharmaceutics, Clinical Pharmacokinetics Laboratory, State University of New York at Buffalo, NY, USA
Summary. Five healthy volunteers received digoxin 0.4 mg or fl-methyldigoxin 0.4 mg i.v., daily for 14 days, in a randomized cross-over arrangement. By monitoring minimal plasma concentrations during multiple dosing, it was found that the steady state pharmacokinetics of digoxin and /3-methyldigoxin could be estimated even better by a one-compartment than by a two-compartment model. The following mean parameters were calculated: the half life of digoxin of 1.54 + 0.31 days was significantly shorter than the half life of 2.29 _+ 0.34 days for /3-methyldigoxin. The distribution volume of 807 + 187 liters for digoxin was not significantly larger than the 735 _+ 227 liters for/3-methyldigoxin. Renal digoxin clearance of 191 + 25 ml/min was significantly higher than both the renal clearance of/3methyldigoxin of 111 + 23 ml/min and also the creatinine clearance, which indicates tubular secretion of digoxin. There was a 2.8-fold accumulation of/3-methyldigoxin injected once a day, which was significantly higher than the 1.8-fold accumulation of digoxin. Key words: Digoxin, /3-methyldigoxin, prolonged administration, pharmacokinetics.
From single dose studies a two-compartment model (Reuning et al., 1973) and even a three-compartment model (Kramer et al., 1974) have been found necessary to describe the pharmacokinetics of digoxin. The existence of a deep compartment for digoxin in the CNS has been demonstrated in the dog
* This study was supported by Bundesministerium fiir Jugend, Familie und Gesundheit, Federal Republic of Germany.
(Rietbrock and Kuhlmann, 1977). On the other hand, digoxin is mainly employed for extended treatment, and the physician should have simple but correct equations which permit ready calculation of the required dose. A pharmacokinetic concept of digitalis glycoside accumulation may be used to advantage in clinical practice if the glycoside concentration in plasma and its elimination can be described by equations appropriate to a one-compartment model (Dettli, 1975). The underlying assumptions are: 1. a constant and time-independent relationship between the plasma concentration and the total amount of drug in the body; 2. very rapid uptake compared to very slow elimination rate, and 3. proportional relationship between the amount of drug excreted over a period of time and the amount remaining in the body (Dettli, 1975).
Methods Five healthy volunteers (Table 1) took part in the study, and they gave their informed consent to the investigation. To use the same units as the later digoxin clearance evaluated subsequently the creatinine clearance of the volunteers was not standardized to a 1.73 m 2 body surface area. The other Table 1. Details of healthy subjects studied Subject
P.B.F.K.B.H.
B.K. W.S. mean_+SD
Age (years) Body weight (kg) Height(cm) SEX Creatinineclearance (ml/min)
22 59 175 ~
27 58 175 CY
27 70 178 CY
25 69 175 CY
24 65 t77 CY
25+2 64+6 176+1
88
92
186
126
188
136+44
388
F. Keller et al.: Pharmacokinetics of Digoxin and ~Methytdigoxin 4
In[10xncj;/ml]
1
7
14
days
19
Fig. 1. Plasma concentrations (ng/ml) in 5 volunteers after multiple doses of digoxin 0.4 mg ( * ~ * ) or fl-methyldigoxin 0.4 mg (o . . . . . . o) daily for 14 days i. v.
6
I. [ ~/d~] 5
4
3
2
i
1
7
14
2t
28
Fig. 2. Renal excretion 0tg/die) by 5 volunteers after multiple doses of digoxin 0.4 mg (**) or fl-methyldigoxin 0.4 mg (o . . . . . . o) daily for 14 days i.v.
Table 2. Pharmacokinetic parameters of digoxin and fi-methyldigoxin using a one-compartment open model Digoxin
fi-methyldigoxin
Subject
P.B.
F.K.
B. H~
B, K.
W, S.
mean _+ SD
EST KTOT KREN VOL CTOT CREN CCR CR/CT CR/CCR T/2 BMIN BMIN/D
0.50 0.581 0.369 595 240 153 88 0.64 1.74 1.19 507 1.27
0,44 0.490 0.320 814 277 181 92 0.65 1,96 1.41 632 1.58
0.39 0,477 0,409 686 227 195 186 0.86 1.05 1.45 655 1.64
0.33 0.343 0,273 1085 258 206 126 0.80 1.63 2.02 979 2.45
0.46 0.422 0,369 854 250 219 188 0.87 1.16 1.64 761 1.90
0.42 0,423 0.348 807 250 191 136 0.76 1.51 1,54 707 1,77
0.07 0.088 0,053 187 19 25 49 0.11 0.39 0.31 177 0.44
P.B.
F.K.
B.H.B.K.
W. S, mean -+ SD
0.36 0,372 0.255 482 t24 85 88 0.69 0.97 1.86 887 2.22
0.36 0.350 0.248 529 128 91 92 0.71 0.99 1.98 955 2.39
0.38 0.280 0,214 755 147 112 186 0.77 0,60 2.48 1239 3.10
0.26 0.274 0.205 920 175 131 188 0.75 0.69 2.53 1269 3.17
0.22 0.266 0,197 991 183 135 126 0,74 1.07 2.60 1310 3.28
0.32 0.308 0.224 735 151 111 136 0.73 0.86 2.29 1132 2.83
0.07 0,049 0.026 227 27 23 49 0,03 0.21 0.34 196 0.49
p L
0.005 0.0t n.s. 0.005 0.001 n.s. 0.005 0.005 0.005
389
F. Keller et al.: Pharmacokinetics of Digoxin and/3-Methyldigoxin
reason for the large discrepancies between the individual creatinine clearances, even though they were estimated three times, was the fact that B.H. and W.S. physically exerted themselves in the morning before coming to the laboratory (Table 1). Digoxin 0.4 mg or fl-methyldigoxin 0.4 mg were given intravenously once daily for 14 days. Each subject received both glycosides in a random sequence at a 14 day interval to ensure almost complete elimination of glycoside from the preceding study. Blood samples from each subject were collected in heparinized tubes 24 h after administration. Samples were immediately centrifuged and the plasma removed and stored at - 2 0 ° C until analyzed. Urine was collected for periods of 24 h for the 14 days of treatment and for the 14 days after medication had closed. In subjects P.B. and F.K. urine was collected for 24 days after intravenous administration of fl-methyldigoxin had ceased. After measurement of total volume, aliq u o t s of urine were stored at - 2 0 ° C . Plasma glycoside concentrations were measured by 125Iradioimmunoassay utilizing a commercial kit from Clinical Assays Inc., which had a standard sample reproducibility of 9%, a precision of 6% (for concentrations below 1.0 ng/ml of 16%), a recovery of 5% (for concentrations below 1.0 ng/ml of 12%), a sensitivity (expressed as maximal binding capacity) of 35-40%, and a 12% coefficient of variation in normal plasma, which indicates its specificity. Urinary glycoside concentrations were measured by the 3H-radioimmunoassay reported by Maertin and Rietbrock (1977). The pharmacokinetic parameters were estimated by means of a nonlinear, least squares regression program, NONLIN (Metzler, 1969). The calculations employed equations describing the time course of the plasma concentration after multiple dosing for a one- or two-compartment model and the urinary excretion rate of the drug (Gibaldi et al., 1975, equations 359, 419 and 14). Both types of data were fitted sumultaneously and all data points were given an equal numerical weight. Statistical analysis included mean, standard deviation and Student's t-test for paired trials.
life (6 days for digoxin and 9 days for fl-methyldigoxin, see Table 2) the plasma concentrations were within 10% of their steady state levels. The estimated steady state plasma levels of fl-methyldigoxin were significantly higher (p < 0.005) than those after digoxin, 1.81 + 0.33 ng/ml and 0.90 + 0.17 ng/ml, respectively. But the average renal excretion of digoxin and fl-methyldigoxin during a steady state interval did not differ significantly, 298 + 30 ~tg and 296 _+ 27 ~tg, respectively. Urinary excretion rates over a period of 24 days after cessation of/3-methyldigoxin treatment in subjects P.B. and F. K. are shown in Figure 3. From the 28th day the terminal excretion of glycoside was retarded, which reflects a slowly equilibrating tissue pool. The pharmacokinetic parameters computed with a one-compartment model showed very close correlation to the empirical data (Figs. 4 and 5). Urinary excretion rate data yielded better fits (r = 0.996 for digoxin and/3-methyldigoxin) than did plasma concentration data (r -- 0.974 for fl-methyldigoxin and r = 0.939 for digoxin), indicating greater reliability of urinary excretion rate data. The average half life of digoxin was 1.54 days, which was distinctly lower than the 2.29 day half life of fl-methyldigoxin (Table 2). Since the volumes of distribution did not differ, the total clearance of the latter was 1.66 times that of digoxin. The renal clearance of fl-methyldigoxin was distinctly lower than the creatinine clearance, whereas the renal clearance of digoxin exceeded the creatinine clearance by about 40 percent. Comparison of the ratio of the minimum amount at steady state in the body and the intravenous dose of digoxin and fl-methyldigoxin permitted estimation of the extent of accumulation. The accumulation factor which is usually calculated from the elimination rate constant and the dosing interval, the ratio of BMAX/D or (BMIN + D)/D (Gibaldi et al., 1975), should not be identified with the extent of accumulation as used here.
Discussion Results
The plasma concentrations and the renal excretion of digoxin and/3-methyldigoxin in 5 volunteers are shown in Figs. 1 and 2. The rate and extent of accumulation of digoxin and fl-methyldigoxin was a function of the magnitude of the dosing interval and the half life of the two glycosides. After multiple dosing for a period equal to four times the half
The present study is believed to be one of the first to present empirical and pharmacokinetic data about plasma levels and renal excretion of digoxin during prolonged treatment as employed in clinical practice, and which verifies Augsberger's concept based upon pharmacodynamic observations. The intention was to reassess the pharmacokinetic characteristics of digoxin and /3-methyldigoxin in healthy volunteers after multiple dosing. In regard to clinical
390
F. Keller et al.: Pharmacokinetics of Digoxin and/3-Methyldigoxin 6
In[~g/die]
5
"'::':i"..
. . . . .
"'..
.
"'""i'. %1. ::,
':::ii/:;:::: .................... "..... ........ i:...... :].-"
14
21
28
digoxin
B-methyldigoxin
2.0
2.0
NG/ML
/
r = .939
~,0
0.5
/ ' 01,5
NGIML r = .974
1.5
jo/S
'L.O
0.5
0.0
0.0
0
35
days
Fig. 3. Renal excretion (~tg/die) by subjects P.B. and F.K. after cessation of multiple doses of fl-methyldigoxin
ll.O
II.5
2i, 0
0
0i. 5
11,0
NG/ML
21,0
NG/ML
digoxin
B-methyldigoxin
400
400
~JGIDIE
~1,5
~JGIDIrE
Fig. 4. Correlation of plasma concentrations (ng/ml) between observed (x) and calculated (y) values using a one-compartment open model
.//° /
r = . 996 300
3O0
200
20(]
"1_.00
100
/
4
0 1001
2001
3001 400 I )JG/DIE
'tOOI
200 ~
3001
400 I
,,UO/O[E
Fig. 5. Correlation of renal excretion (btg/day) between observed (x) and calculated (y) values using a one-compartment open model
F. Keller et al.: Pharmacokinetics of Digoxin and/3-Methyldigoxin
391
applicability a one-compartment model will be more successful than a complex multi-compartment model. Data supporting use of such a one-compartment model include the correlation between computed body glycoside level and measured plasma glycoside concentration, and the correlation between computed body glycoside level and measured myocardial digitalis concentration at autopsy (Jelliffe, 1967; Jelliffe et al., 1972). From single dose studies Reuning et at., (1973) postulated that a two-compartment open model was necessary to account for digoxin pharmacokinetics and Kramer et al., (1974) indicated that a threecompartment open model gave a significantly improved fit. However, after multiple doses and by monitoring minimal plasma concentrations as well as urinary excretion rates, the ability to discern two-compartment kinetics of digoxin and /3-methyldigoxin is decreased. Since the rattio of zero time intercepts A/B = ( a -- k21)/(k21 -- / 3 ) f r o m a two compartment model is low (2.8 for ~methyldigoxin), the plasma concentration time curve is more nearly described by a one-compartment model. In usingthe NONLIN least squares regression program of Metzler (1969) the /3-value in the two-compartment model revealed a very poor coefficient of variation, average 85.5 _+ 54.1 percent for /3-methyldigoxin, whereas the corresponding KTOT-value in the onecompartment model showed a much smaller variation coefficient averaging 3.1 _+ 1.6 percent. Therefore, in the present study a one-compartment model offered a distinct improvement over a two-compartment open model. At the same time, the design of the study came very close to the therapeutic use and clinical monitoring of digitalis glycosides. The observed half lives of digoxin and/3-methyldigoxin encompass the range reported by Rietbrock et al., (1976), which showed a longer half life of/3methyldigoxin than of digoxin. The volumes of distribution, renal clearance and total clearance of digoxin were within the range calculated by a twocompartment open model (Koup et al., 1975). Marked differences were observed between the elimination kinetics of/3-methyldigoxin and digoxin, apparently due to the higher renal clearance of digoxin, which was also significantly (p < 0.025) greater than the creatinine clearance. The assumption of tubular secretion of digoxin is based on data from Steiness (1973), Falch (1973) and Koup et al. (1975). The extent of accumulation of digoxin and fl-methyldigoxin, according to the ratio of the calculated body stores and the maintenance dose just before receiving the next dose during steady state, was much lower for digoxin than for /3-methytdigoxin.
From the urinary excretion rate of digoxin and ~methyldigoxin on the first day, which amounted to 37 + 0.08 percent of the dose after digoxin and 22 + 0.03 percent of the dose after/3-methyldigoxin (p < 0.02), it may be suggested that the elimination rate on the first day was an essential indicator of glycoside accumulation in the body. Therefore, a reduction in the dose of fi-methyldigoxin compared to digoxin is necessary to prevent the greater risk of developing symptoms of glycoside toxicity. The reduction of glycoside pharmacokinetics to a one-compartment model is not devoid of disadvantages. From studies in beagle dogs (Rietbrock and Kuhlmann, 1977) it must be concluded that the brain is a deep compartment. Therefore, glycoside side effects resulting from CNS accumulation may be masked in using only the elimination rate constants of a one-compartment model. The slow release of /3-mthyldigoxin from the brain may have contributed to the increased half life (5.2 days for P.B. and 5.8 days for F. K.) of the urinary excretion rates from the 14th day after cessation of/3-methyldigoxin administration (Fig. 3). The calculated renal excretion rates of fl-methyldigoxin and digoxin agreed well with the observed data (Fig. 5). However, in the case of fi-methyldigoxin calculated plasma concentrations less than 0.78 ng/ml were higher and concentrations more than 0.78 ng/ml were lower than the observed values (Fig. 4). This error probably reflects a rate limiting step in demethylation by liver enzymes (Rietbrock et al., 1972; Zilly et al., 1975). With an increasing plasma concentration of total glycosides, the relative amount of digoxin decreases. These findings would also indicate that unchanged fl-methyldigoxin has a considerably smaller volume of distribution than digoxin. Abbreviations of parameters D EST KTOT KREN VOL ClOT CREN BMIN CCR CR/CT CR/CCR T/2 BMIN/D
dose estimated disposition rate constant (day -~) semilog curve fit total disposition rate constant (day -I) NONLIN fit renal disposition rate constant (day -1) NONLIN fit volume of distribution (liters) NONLIN fit total clearance (ml/min) renal clearance (ml/min) body stores (~tg) before next dose in steady state creatinine clearance (ml/min) fraction renally excreted renal drug clearance versus creatinine clearance half life (days) extent of accumulation
392
F. Keller et al.: Pharmacokinetics of Digoxin and fl-Methyldigoxin
References
Number 7292/69/7292/005. The Upjohn Company, Kalamazoo, Mich. 1969 10. Reuning, R.H., Sams, R.A., Notari, R.E.: Role of pharmacokinetics in drug dosage adjustment. I. Pharmacologic effect kinetics and apparent volume of distribution of digoxin. J. clin. Pharmacol. 13, 127-141 (1973) 11. Rietbrock, N., Rennekamp, C., Rennekamp, H., Bergmann, K.v., Abshagen, U.: Demethylation and cleavage of glycosidic bonds of 4'"-methyldigoxin in man. NaunynSchmiedeberg's Arch. Pharmacol. 272, 450--453 (1972) 12. Rietbrock, N., Guggenmos, J., Kuhlmann, J., Hess, U.: Bioavailability and pharmacokinetics of fl-methyldigoxin after multiple oral and intravenous doses. Europ. J.clin. Pharmacol. 9, 373-379 (1976) 13. Rietbrock, N., Kuhlmann, J.: Clinical significance of distribution and elimination of cardiac glycosides. In: Clinical Pharmacokinetics (ed. Ritschel, W.A.). pp. 57-65. Stuttgart: Gustav Fischer 1977 14. Steiness, E.: Renal excretion of digoxin. In: Symposium on Digitalis. (ed. Storstein, O., Nitter-Hauge, S., Storstein, L.) pp. 178-183. Oslo, Norway: Gyldendal 1973 15. Zilly, W., Richter, E., Rietbrock, N.: Pharmacokinetics and metabolism of digoxin- and /3-methyldigoxin-12-3H in patients with acute hepatitis. Clin. Pharmacol. Ther. 17, 302-309 (1975).
1. Dettli, L.: Pharmacokinetische Grundlagen fiir die optimale Dosierung bei repetierter Applikation yon Arzneimitteln. In: Klinische Arzneimittelpriifung. (ed. Eickstedt, K.W. und Gross, F.) 1. Symposium fiir Klinische Pharmakologie des Bundesgesundheitsamtes Berlin, 1974. pp. 59-68. Stuttgart: Gustav Fischer 1975 2. Falch, D., Teien, A.: The influence of kidney function on the plasma level and urinary excretion of digoxin. In: Symposium on Digitalis (ed. Storstein, O.; Nitter-Hauge, S. and Storstein L.) pp. 183-189. Oslo: Norway, Gyldendal 1973 3. Gibaldi, M., Perrier, D.: Pharmacokinetics. New York: Marcel Dekker, 1975 4. Jelliffe, R.W.: A mathematical analysis of digitalis kinetics in patients with normal and reduced renal function. Math. Biosci. 1, 305-325 (1967) 5. Jelliffe, R.W., Buell, J., Kalaba, R.: Reduction of digitalis toxicity by computer-assisted glycoside dosage regimens. Ann. int. Med. 77, 891-906 (1972) 6. Koup, J.R., Greenblatt, D.J., Jusko, W.J., Smith, T.W., Koch-Weser, J.: Pharmacokinetics of digoxin in normal subjects after intravenous bolus and infusion doses. J. Pharmakokinet. Biopharm. 3, 181-192 (1975) 7. Kramer, W.G., Lewis, R.P., Cobb, T.C., Forester, W.F.Jr., Visconti, J.A., Wanke, L.A., Boxenbaum, H.G., Reuning, R.H.: Pharmacokinetics of digoxin: comparison of a two- and a three-compartment model in man. J. Pharmakokinet. Biopharm. 2, 299-312 (1974) 8. Maertin, K., Rietbrock, N.: Radioimmunologische Bestimmung von Digoxin in Urin. Klin. Wschr. 55, 429--438 (1977) 9. Metzler, C.M.: NONLIN; A computer program for parameter estimation in nonlinear situations. Technical Report
Received; April 4, 1977, in revised form: Juni 28, 1977, accepted: July 4, 1977 Prof. Dr. N. Rietbrock Department of Clinical Pharmacology Hindenburgdamm 30 D-1000 Berlin 45 Federal Republic of Germany
