European Journal of Clinical Pharmacology
Europ. J. clin. Pharmacol. 11, 14%154 (1977)
©
by Springer-Verlag 1977
Huorometric Determination of Hydrofhmethiazide in Human Plasma and Urine after Its 0ral Administration O. Brors, S. Jacobsen and E. Arnesen Institute of Pharmacology, University of Oslo, and Medical Department VII, Ullevgd Hospital, Osto, Norway
Summary. A spectrofluorometfic method for determination of hydroflumethiazide in human plasma and urine has been developed. The limit of detection was 10 ng/ml plasma and 100 ng/ml urine. The plasma concentration of hydroflumethiazide was determined for 9-11 hours and excretion in urine for 24-37 hrs after oral administration of about 1 mg/kg body weight to 7 subjects. Plasma half life in healthy subjects was 1.9-2.1 h, and 2.7-8.6 h in patients during the period 4-9 hrs after dosing. Cumulative excretion in urine was 67-79% of the dose during 31-37 hrs in 6 subjects; one patient with renal disease was found to excrete only 25.8% of dose during 24 hours. Renal clearance of hydroflumethiazide was higher in the healthy subjects (0.29-0.44 1 h -~ kg-1) than in the patients (0.040-0.15 1 h -1 kg-l). Plasma half life of hydroflumethiazide was not closely correlated with renal clearance of the drug, which suggests that other factors may play a role in determining plasma half life.
Key words" Hydroflumethiazide, spectrofluorometry, pharmacokinetics, plasma half life, renal excretion, renal disease.
The thiazide or benzothiadiazine diuretics are extensively used in the treatment of arterial hypertension. Although it is more than 15 years since their introduction, knowledge of the human pharmacokinetics of these drugs is still incomplete, being limited mainly to the study of their excretion in urine. As to the pharmacodynamics of thiazide diuretics, dose response relationships have been reported by several investigators [1, 2]. However, no study appears to have been published on plasma or blood concentration in man, or on its correlation, if any, with the pharmacological effects.
The duration of effect of a single oral dose of a thiazide diuretic on the renal excretion of electrolytes and water varies considerably; e. g. from 12-24 hrs for chlorothiazide, hydrochlorothiazide, and hydroflumethiazide [3, 4, 5] to 36-48 hrs for polythiazide [6, 7, 8]. In relation to regular daily doses, these figures seem to indicate that significant amounts of drug may still remain in the body when the next dose is given. An interesting question, therefore, is whether the occurrence of side effects, such as hypokalaemia [9, 10, 11], uric acid retention [12, 13], and impaired glucose tolerance [14, 15] during chronic treatment with thiazide diuretics is due to a high steady state concentration in the body. To answer this problem, further knowledge of the pharmacokinetics and pharmacodynamics of these drugs must be obtained. The aim of the present investigation was to reveal some of the pharmacokinetic characteristics of hydroflumethiazide, a thiazide which has been in use since 1958. For this purpose a spectrofluorometrie method of analysis of the drug in plasma and urine has been developed, and it has been used to determine the plasma concentration and renal excretion of hydroflumethiazide in 7 subjects after a single oral dose. No attempt was made to correlate diuretic activity with plasma concentration, as the dose of hydroflumethiazide administered would be expected to have little diuretic activity (1), and salt and water balance in the subjects was not standardized.
Material and Methods Material
Hydroflumethiazide was obtained as a pure substance from Leo, Copenhagen, Ballerup, Denmark. Tablets containing hydroflumethiazide 25 mg (Rontyl ®, Leo)
150 all from the same batch, were obtained commercially. Human serum albumin was obtained as a lyophAlized powder from Kabi AB, Stockholm, Sweden. Heparin was obtained as a powder from Novo Industri A/S, Copenhagen, Denmark. All chemicals used in the analyses were analytical reagent grade. Vacutainer tubes were obtained from Becton, Dickinson-France, Grenoble, France.
Analytical Procedure Analysis of Plasma. Atiquots of plasma 1 ml were placed in glass-stoppered extraction tubes. Saturated (NH4)2SO 4 1.5 ml, 5 N H2SO4 0.2 ml and 24% (w/v) N a z W O 4 • 2 H 2 0 0.3 ml were added, the solution being mixed for about 5 seconds on a Vortex mixer after addition of each compound. The solution was extracted for 15 minutes with toluene-acetone (3 : 1 v/v) 8 ml using a mechanical shaker. After centrifugation, 7 ml of the organic layer was transferred to an extraction tube and evaporated to dryness at room temperature by blowing an air stream into the tube. The dry residue was dissolved in 0.1 N H2SO4 0.6 ml, and washed with chloroform 1.2 ml for 15 minutes in a mechanical shaker. After centrifugation, the sulphuric acid phase was isolated for fluorescence measurement. In every analysis, blanks and standards of hydroflumethiazide in plasma and in 3% human albumin dissolved in a Krebs-Ringer-phosphate buffer (pH 7.4) were also extracetd. Analysis of Urine. To cover the wide concentration range encountered in urine, sample volumes ranging from 50-500 microliters were analysed. 0.15 M citrate-phosphate buffer, pH 7.0, was added to make a total volume of 1 ml, followed by saturated (NH4)2SO4 1 ml, and the mixture was extracted for 15 minutes with toluene-acetone (3:1 v/v) 5 ml in a mechanical shaker. After centrifugation, 0.5-2 ml of the organic layer was transferred to an extraction tube and evaporated to dryness. The dry residue was dissolved in 0.1 N H2SO4 0.6 ml and washed for 15 rain with chloroform 3 ml in a mechanical shaker. After centrifugation, the sulphuric acid phase was isolated for fluorescence measurement. In every analysis, appropriate blanks and standards were also extracted.
Fluorescence Measurement This was performed in a Farrand spectrofluorometer, using microcuvettes of internal diameter 3 mm, which required 0.2 ml for a reading. Excitation was at 279 nm, emission at 390 nm (uncorrected).
O. Brors et al.: Hydroflumethiazidein Human Plasma and Urine
Search for Metabolites Urine samples were extracted with dichloromethaneacetone (3:1), chloroform-acetone (3:1), and toluene-acetone (3 : 1), evaporated to dryness and redissolved in a small volume of methanol. TLC on Silica Gel 1 B-F, "Baker-flex" (J. T. Baker Chemical Co., PhiUipsburg, New Jersey) was performed with three different chromatographic solutions: ethyl acetate-methanol-0.1 N NH3 48 : 1 : 1 (volumes), toluene-acetone 3 : 1 and chloroform-acetone-methanol 8 : 1 : 1. Spots containing hydroflumethiazide were located in UV light.
Subjects, Drug Administration and Samples Two healthy volunteers and 5 patients with a tendency to develop oedema were included in the study. Routine clinical laboratory methods were used to evaluate liver and renal function. Heart function was classified according to the system of the New York Heart Association. All subjects received hydroflumethiazide orally in the morning, after an overnight fast. The drug was administered as tablets which were swallowed with a glass of water. The subjects then fasted for a further 2 hours to reduce the influence of food on absorption. During the study, the healthy volunteers performed normal indoor activities, whilst the patients were confined to bed and were only allowed to wash themselves and go to the toilet as usual. Other diuretics were withheld for at least 3 days before hydroflumethiazide was administered, and throughout the whole sampling period. Blood was collected immediately before and hourly for about 9 hours after administration of hydroflumethiazide. The blood from the patients was taken in glass Vacutainer tubes containing sodium heparin, and from the healthy subjects in tubes heparinized in our laboratory. The blood was centrifuged immediately to obtain plasma, which was kept frozen at - 2 0 ° until analysed. Urine was collected 0-3, 3-6, 6-9, 9-24, and 24-36 hours after dosing. The urine volume was measured and an aliquot frozen at - 2 0 ° until analysed.
Calculations After correction for blank values, a least squares regression line through zero was calculated for every standard curve, using reciprocal weighing. Absolute recovery of hydroflumethiazide from plasma and urine was estimated by extraction of spiked aliquots of plasma and urine, and the amount of drug in the final sulphuric acid phase expressed as a percentage of the amount in the spiked aliquot.
O. Brors et al.: Hydroflumethiazide in Human PIasma and Urine
151
Table 1. Mean coefficient of variation a of duplicate analysis of plasma and urine samples Plasma
Urine
Concentration range (ng/ml)
Mean coefficient of variation (%)
Number of samples
Concentration range (~,g/ml)
Sample size 0xl)
Mean coefficient of variation (%)
Number of samples
4 0 - 100 t 0 0 - 200 2 0 0 - 500 500-1200
5.4 2.4 3.2 2.0
6 6 11 13
4 - 10 4 - 10 10- 50 50-150
50 500 50 50
19.7 25 2.0 1.4
6 5 14 9
100 ,n a Calculated by the formula - - ] / 2 Z [A, - B i / (A i + Bi)] where Ai and B~are the results of duplicate analysis of the same sample, and n the number of samples n i= I Table 2. Sex, age, principal diagnosis, functional heart group (New York Heart Association classification), and serum creatinine of the
subjects investigated Subject no
Sex
Age (yrs)
Main diagnosis
Functional heart group
Serum creatinine (mg/lO0 ml)
1 2
M F
65 74
IV IV
1.4 1.1
3
F
70
III
1.0
4
M
60
II
2.4
5 6 7
F M M
61 38 29
Cqronary h.d. Arteriosclerotic and hypertonic h. d. Arteriosclerotic and hypertonic h. d. Subacute glomerulonephritis Coronary h.d. Healthy Healthy
(II)-III I I
1.5 0,9 1.0
M = male, F = female, h. d. = heart disease
The slopes of the straight part of the plasma concentration curves, in a semilogarithmic plot, were calculated by least squares regression lines, and plasma half lives were calculated from the slopes so obtained.
Results
Analysis At excitation and emission wavelengths of 279 and 390 nm, respectively, there was a linear relationship between the concentration of hydroflumethiazide in the solution and the intensity of fluorescence, up to a concentration of 6 ~tg/ml. All the measurements were performed within this concentration range. Absolute recovery of hydrofiumethiazide from plasma was 83%, and from urine 33%, when 2 ml of the organic phase was evaporated. Ommission of (NH4)2SO4 from the analysis without other changes in
the procedure reduced the recovery from plasma to 68% and from urine to 24%. The slopes of standard curves run in plasma and in 3% albumin solution were not significantly different. If the slope of the plasma standard curve was taken as 1.000, the relative slope of the corresponding albumin curves, run in the same analysis, as 1.007 +_ 0.017 (mean + standard deviation, 4 pairs of curves). The mean correlation coefficient of the standard curves was 0.9987, and that of the urine standard curves was 0.9990. Standard curves for plasma and 3% albumin were linear down to 10 ng/ml when 1 ml aliquots were analysed. This corresponded to the lower limit of detection of hydroflumethiazide in plasma, as the blank reading for plasma corresponded to 2 0 2 5 ng/ml of drug. Using a 500 ~tl sample, the limit of detection of hydroflumethiazide in urine was 100 ng/ml, the blank value being 200-250 ng/ml. The mean coefficient of variation between duplicate determinations at different concentration ranges is listed in Table 1.
152
O. Brors et al.: Hydroflumethiazide in Human Plasma and Urine
Table 3. Plasma concentration (ng/ml) and plasma half life of hydroflumethiazide Hours after dose
Plasma half life (h)
Subject no
1
2
3
4
5
6
7
8
9
1 2 3 4 5 6 7
440 236 499 788 713 e 403 171 k
902 574 871 836 727 f 259 446
1073 688 645 702
989 549 525 512 a 690g 202 384
846 430 362 355 b 516 h 117 190 m
783 363 254 334 436 89 159
744 280 211 270 c 362 63 105
694 233 168 2090 222 47 88
640 202 147
198 i 658
8.6 3.4 2.7 3.8 3.5 2.1 1.9
220 50 89
Time point of sampling (h): a = 3.8, b = 5.3, c = 7.2, d = 8.5, e = 1.4, f = 2.5, g = 3.6, h = 4.7, i = 3.2, k = 1.3, m = 5 . 6
EXCRETION RATE mglh
Plasma concentration
subject 2
ng/ml 2000-
subject 3
subject/*
10005000 12 2f, EXCRETION RATE mg/h
200I0050-
subject '~ 1
• o
36
subject 5
0
12 24 12 2/* 36 0 TIME AFTER DOSE (h)
subject 6
36
0
12
2t,
subject 7
2
3 4
subject o
5
a
7
20:
4
6
8
10
12 0
2
4
6
8
10
Time after dose (hi
0~ 0
12 24
0 12 24 316 0 36 TIME AFTER DOSE (h)
12
2/*
316
Fig. 1. Plasma concentrations of hydroflumethiazide in 7 subjects after a single oral dose. The left side shows curves from the patients and the right side those from healthy volunteers
Fig. 2. Excretion rate of hydroflumethiazide in urine in 7 subjects after a single oral dose
Table 4. Urinary excretion rates of hydroflumethiazide (mg/h)
PharrnacokineticStudy
Subject
Time interval (hours after dose)
no
0-3
3-6
6-9
9-24
24-36
1 2 3 4 5 6 7
1.1 1,8 0,83 1.4 3.7 7.4 9.9
4.6 6.0
1.3 2.7 2.5" 0.76 2.0 1.4 2.5
1.8 0.45
1.1 0.19 0.80 b
1.8 5.3 4.4 7.1
0.51 0.81 0.38 0.67
Time point of sampling (h): a = 3-9, b = 9-33
0.25 0.12 0.11
Details of the subjects in the study have been listed in Table 2. It can be seen that renal function, as judged by serum creatinine, was considerably reduced in subject 4 and slightly reduced in subjects 1 and 5. Liver function (not listed) was normal in all patients except subject 1, who had a serum bilirubin of 3.1 mg/100 ml. Plasma concentrations measured in the 7 subjects after a single oral dose of hydroflumethiazide are listed in Table 3, and in Figure 1 the data are shown as a semilogarithmic plot. The maximum plasma con-
O. Brors et al.: Hydroflumethiazide in Human Plasma and Urine
153
Table 5. Body weight, dose, cumulative urinary excretion and renal clearancea of hydroflumethiazide Subject no
1 2 3 4 5 6 7
Body weight (kg)
75 62 46 75 49 75 90
Dose (mg)
75 50 50 75 75 75 100
Cumulative urinary excretion
Renal clearance
(% of dose)
(h)
(1 h -1)
(1 h -1 kg -1)
78.3 74.8 65.9 25.8 66.3 67.7 72.9
35.0 35.3 32.4 24.0 36.0 37.1 30.8
3.1 9.1 4.8 3.0 6.9 32.8 26.0
0.041 0.15 0.10 0.040 0.14 0.44 0.29
" Calculated by the formula U0-x •whereUisthecumu•ativeexcreti•ninurineandAUCtheareaunderthep•asmac•ncentrati•ncurve•x 9 hrs AUC0_x
centration was obtained 1-3 hours after dosing. The plasma half life was calculated for the last part of the curves (Table 3). They were longer in patients than in the healthy subjects, and there was considerable variation in half life amongst the patients. The urine concentration of hydroflumethiazide varied over the range 2-140 ~tg/ml. Urinary excretion rates are listed in Table 4 and are shown graphically in Fig. 2. The cumulative excretion of hydroflumethiazide in the urine is shown in Table 5, together with the calculated renal clearance of the drug during the first 9 hours after the single dose. Renal clearance in patients was much lower than in the healthy subjects, even when corrected for differences in body weight. No metabolite was detected by TLC of urine extracts.
Discussion
The spectrofluorometric method described has the same limit of detection as that reported previously for a method based on spectroftuorodensitometry [16]. The method proved successful in determining plasma concentrations for 9 hours after dosing in the seven subjects investigated. As judged from the plasma concentration curves, it might be possible to determine plasma concentration for a much longer time in patients after a single close of hydroflumethiazide. Plasma levels of hydrochlorothiazide, determined with a GLC method after a single oral dose of 50 rag, were in the range 50-350 ng/ml 1.5 to 4 hours after dosing, after 100 mg a maximal concentration of 710 ng/ml was found after 1.5 hours [17]. These concentrations were in the same range as was observed in the present study, and these two thiazides are reported to have the same potency [3].
The finding that all the subjects, except subject 4, excreted 65-79% of the oral dose in urine during the first 31-37 hours after dosing (Table 5), is in agreement with earlier studies [3], and confirms renal excretion as the main elimination pathway for hydroflumethiazide in subjects with normal renal function. The plasma half life of hydroflumethiazide might be expected to be correlated closely with renal function, provided other routes of elimination remain unchanged. As the plasma half life of subject 4 was found to be as short as 3.8 hours, in spite of reduced renal function, increased biliary excretion or metabolism must be considered in this patient. In subject 1, the reduction in both renal and hepatic function was associated with more than doubling of the plasma half life of hydroflumethiazide, presumably because biliary excretion or metabolism could not compensate for the reduced renal elimination. The renal clearance of hydroflumethiazide was considerably lower in the patients than in the healthy subjects, in whom renal clearance was almost as high as the expected renal plasma flow, and several times higher than the anticipated glomerular filtration rate. The high renal clearances must be explained, therefore, by tubular secretion of hydroflumethiazide. Impairment of the renal circulation, known to occur in heart failure, may have been responsible for lower renal clearance of hydroflumethiazide in the patients with heart failure in comparison to the healthy subjects. The high cumulative excretion in urine over 36 hours, as seen in subject 1, may be explained by the reduced liver function and prolonged half life of the drug in plasma. Judged by the area under the plasma concentration curve and the plasma half life of the drug in subject 4, the amount absorbed from the gastrointestinal tract in this patient should be comparable to that in the other subjects. Therefore, normal liver function in this patient was probably responsible for elimination
154
of the greater part of the drug, presumably by excretion in the bile or by metabolism. The latter is not known to occur in man or animals, and no metabolite was detected in this study. The urinary excretion rate of hydroflumethiazide observed in the healthy subjects 24-36 hrs after dosing is higher than might be expected from a plasma half life of 2 hrs. The reason for the discrepancy might be enterohepatic recirculation of the drug, or a later phase of elimination with a longer half life than the initial phase. Further investigations are under way for evaluation of this phenomenon. Based on observed interindividual differences in plasma concentration levels and plasma half life after a single oral dose of hydroflumethiazide, a variable intensity and duration of pharmacological effect might be anticipated in different subjects. Especially when multiple dosing is concerned, as in the treatment of hypertension and chronic oedema, this may be of relevance for individualization of drug therapy. Therefore, in order to establish a rational therapeutic regime for the thiazides, further studies are required.
Acknowledgements. O. Brors was a Research Fellow of the Norwegian Research Council for Science and the Humanities. Financial support from this Council, from The Norwegian Drug Monopoly, and from Anders Jahre's Foundation is gratefully acknowledged. We wish to thank Mrs. Eva Sigfridsson and Renuka Patel for skilfuI technical assistance.
References 1. Blagg, C. R.: Hydroflumethiazide- a new oral diuretic. Lancet 1959 II, 311-313 2. Dettli, L., Spring, P.: Vergleichende pharrnakologische Priifung eines neuen Salidiureticums (Cyclopenthiazid) am gesunden Menschen. Z. ges. exp. Med. exp. Chir. 134, 310-322 (1960) 3. Young, D. S., Forrester, T. M., Morgan, T. N.: A comparison of the diuretic action of the chlorothi~zide analogues. Lancet 1959 II, 765-768
O. Brors et al.: Hydroflumethiazide in Human Plasma and Urine 4. Matheson, N. A., Morgan, T. N.: Diuretic action of chlorothiazide. Lancet 1958 I, 1195-1199 5. Ford, R. V., Moyer, J. H., Spurr, C. L.: Clinical and laboratory observations on chlorothiazide (Diuril). Arch. int. Med. 100, 582-592 (1957) 6. Tapia, F. A.: Diuretic and antihypertensive effects of polythiazide. Curr. ther. Res. 3, 365-377 (1961) 7. Ford, R. V.: Clinical pharmacological investigation of polythiazide, a potent oral diuretic agent. Curr. ther. Res. 3, 320-329 (1961) 8. Hutcheon, D. E., Leonard, G. B.: Therapeutic efficacy of diuretics with different durations of action. J. Amer. reed. Ass. 185, 6 4 0 6 4 2 (1963) 9. Beevers, D. G., Harpur, J. E., Hamilton, M.: The long-term treatment of hypertension with thiazide diuretics. Postgrad. med. J. 47, 639-643 (1971) 10. Maronde, R. F., Milgrom, M., Dickey, J. M.: Potassium loss with thiazide therapy. Amer. Heart. J. 78, 16-21 (1969) 11. Skovbo, P., Bjerregaard, P., Hvidt, R., Jesting, E , Lorentzen, F., Hvidt, S.: Diuretika og hypokali~emi. Ugeskr. L~g. 134, 1043-1047 (1972) 12. Reutter, F., Schaub, F.: Harnsiiurestoffwechsel and Salidiuretica. Dtsch. med. Wschr. 89, 1101-1104 (1964) 13. Heimsoth, V., Harmann, F.: Untersuchungen zur St6rung des Harns~iurestoffwechsets nach Saluretica-Verabreichung. Dtsch. reed. Wschr. 90, 1905-1908 (1965) 14. Roediger, P. M., Alden, J., Beardwood, D., Hutchison, J. C.: Benzothiadiazines and diabetes mellitus. II. Comparison of hyperglycemic effects of bendroflumethiazide and chlorothiazide in hypertensive patients with diabetes mellitus. Curr. tiler. Res. 6, 670--677 (1964) 15. Runyan, J. W.: Influence of thiazide diuretics on carbohydrate metabolism in patients with mild diabetes. New Engl. J. Med. 267, 541-543 (1962) 16. Garceau, Y., Davis, J., Hasegawa, J.: Quantitative TLC procedure for determination of hydroflumethiazide in biological fluids. J. pharm. Sci. 63, 1793-1795 (1974) 17. Vandenheuvel, W. J. A., Gruber, V. F., Walker, R. W., Wolf, F. J.: GLC analysis of hydrochlorothiazide in blood and plasma. J. pharm. Sci. 64, 1309-1312 (1975)
Received: May 17, 1976, and in revised form: August 26, 1976, accepted: September 2, 1976 Dr. O. Brors Institute of Pharmacology University of Oslo Box 1057, Blindem N-Oslo 3 Norway
Responsiblefor the text: Prof. Dr. H. J. Dengler,Bonn and Prof. Dr. F. Gross, Heidelberg Responsiblefor advertisements:L. Siegel,Kurfiirstendamm237, D-1000 Berlin 15, Springer-Verlag Beriin-Heidelberg-NewYork, Printed in Germanyby aprinta, Wemding Copyright© by Springer-VerlagBerlin. Heidelberg 1977
Europ. J. clin. Pharmacol. 11, 14%154 (1977)
©
by Springer-Verlag 1977
Huorometric Determination of Hydrofhmethiazide in Human Plasma and Urine after Its 0ral Administration O. Brors, S. Jacobsen and E. Arnesen Institute of Pharmacology, University of Oslo, and Medical Department VII, Ullevgd Hospital, Osto, Norway
Summary. A spectrofluorometfic method for determination of hydroflumethiazide in human plasma and urine has been developed. The limit of detection was 10 ng/ml plasma and 100 ng/ml urine. The plasma concentration of hydroflumethiazide was determined for 9-11 hours and excretion in urine for 24-37 hrs after oral administration of about 1 mg/kg body weight to 7 subjects. Plasma half life in healthy subjects was 1.9-2.1 h, and 2.7-8.6 h in patients during the period 4-9 hrs after dosing. Cumulative excretion in urine was 67-79% of the dose during 31-37 hrs in 6 subjects; one patient with renal disease was found to excrete only 25.8% of dose during 24 hours. Renal clearance of hydroflumethiazide was higher in the healthy subjects (0.29-0.44 1 h -~ kg-1) than in the patients (0.040-0.15 1 h -1 kg-l). Plasma half life of hydroflumethiazide was not closely correlated with renal clearance of the drug, which suggests that other factors may play a role in determining plasma half life.
Key words" Hydroflumethiazide, spectrofluorometry, pharmacokinetics, plasma half life, renal excretion, renal disease.
The thiazide or benzothiadiazine diuretics are extensively used in the treatment of arterial hypertension. Although it is more than 15 years since their introduction, knowledge of the human pharmacokinetics of these drugs is still incomplete, being limited mainly to the study of their excretion in urine. As to the pharmacodynamics of thiazide diuretics, dose response relationships have been reported by several investigators [1, 2]. However, no study appears to have been published on plasma or blood concentration in man, or on its correlation, if any, with the pharmacological effects.
The duration of effect of a single oral dose of a thiazide diuretic on the renal excretion of electrolytes and water varies considerably; e. g. from 12-24 hrs for chlorothiazide, hydrochlorothiazide, and hydroflumethiazide [3, 4, 5] to 36-48 hrs for polythiazide [6, 7, 8]. In relation to regular daily doses, these figures seem to indicate that significant amounts of drug may still remain in the body when the next dose is given. An interesting question, therefore, is whether the occurrence of side effects, such as hypokalaemia [9, 10, 11], uric acid retention [12, 13], and impaired glucose tolerance [14, 15] during chronic treatment with thiazide diuretics is due to a high steady state concentration in the body. To answer this problem, further knowledge of the pharmacokinetics and pharmacodynamics of these drugs must be obtained. The aim of the present investigation was to reveal some of the pharmacokinetic characteristics of hydroflumethiazide, a thiazide which has been in use since 1958. For this purpose a spectrofluorometrie method of analysis of the drug in plasma and urine has been developed, and it has been used to determine the plasma concentration and renal excretion of hydroflumethiazide in 7 subjects after a single oral dose. No attempt was made to correlate diuretic activity with plasma concentration, as the dose of hydroflumethiazide administered would be expected to have little diuretic activity (1), and salt and water balance in the subjects was not standardized.
Material and Methods Material
Hydroflumethiazide was obtained as a pure substance from Leo, Copenhagen, Ballerup, Denmark. Tablets containing hydroflumethiazide 25 mg (Rontyl ®, Leo)
150 all from the same batch, were obtained commercially. Human serum albumin was obtained as a lyophAlized powder from Kabi AB, Stockholm, Sweden. Heparin was obtained as a powder from Novo Industri A/S, Copenhagen, Denmark. All chemicals used in the analyses were analytical reagent grade. Vacutainer tubes were obtained from Becton, Dickinson-France, Grenoble, France.
Analytical Procedure Analysis of Plasma. Atiquots of plasma 1 ml were placed in glass-stoppered extraction tubes. Saturated (NH4)2SO 4 1.5 ml, 5 N H2SO4 0.2 ml and 24% (w/v) N a z W O 4 • 2 H 2 0 0.3 ml were added, the solution being mixed for about 5 seconds on a Vortex mixer after addition of each compound. The solution was extracted for 15 minutes with toluene-acetone (3 : 1 v/v) 8 ml using a mechanical shaker. After centrifugation, 7 ml of the organic layer was transferred to an extraction tube and evaporated to dryness at room temperature by blowing an air stream into the tube. The dry residue was dissolved in 0.1 N H2SO4 0.6 ml, and washed with chloroform 1.2 ml for 15 minutes in a mechanical shaker. After centrifugation, the sulphuric acid phase was isolated for fluorescence measurement. In every analysis, blanks and standards of hydroflumethiazide in plasma and in 3% human albumin dissolved in a Krebs-Ringer-phosphate buffer (pH 7.4) were also extracetd. Analysis of Urine. To cover the wide concentration range encountered in urine, sample volumes ranging from 50-500 microliters were analysed. 0.15 M citrate-phosphate buffer, pH 7.0, was added to make a total volume of 1 ml, followed by saturated (NH4)2SO4 1 ml, and the mixture was extracted for 15 minutes with toluene-acetone (3:1 v/v) 5 ml in a mechanical shaker. After centrifugation, 0.5-2 ml of the organic layer was transferred to an extraction tube and evaporated to dryness. The dry residue was dissolved in 0.1 N H2SO4 0.6 ml and washed for 15 rain with chloroform 3 ml in a mechanical shaker. After centrifugation, the sulphuric acid phase was isolated for fluorescence measurement. In every analysis, appropriate blanks and standards were also extracted.
Fluorescence Measurement This was performed in a Farrand spectrofluorometer, using microcuvettes of internal diameter 3 mm, which required 0.2 ml for a reading. Excitation was at 279 nm, emission at 390 nm (uncorrected).
O. Brors et al.: Hydroflumethiazidein Human Plasma and Urine
Search for Metabolites Urine samples were extracted with dichloromethaneacetone (3:1), chloroform-acetone (3:1), and toluene-acetone (3 : 1), evaporated to dryness and redissolved in a small volume of methanol. TLC on Silica Gel 1 B-F, "Baker-flex" (J. T. Baker Chemical Co., PhiUipsburg, New Jersey) was performed with three different chromatographic solutions: ethyl acetate-methanol-0.1 N NH3 48 : 1 : 1 (volumes), toluene-acetone 3 : 1 and chloroform-acetone-methanol 8 : 1 : 1. Spots containing hydroflumethiazide were located in UV light.
Subjects, Drug Administration and Samples Two healthy volunteers and 5 patients with a tendency to develop oedema were included in the study. Routine clinical laboratory methods were used to evaluate liver and renal function. Heart function was classified according to the system of the New York Heart Association. All subjects received hydroflumethiazide orally in the morning, after an overnight fast. The drug was administered as tablets which were swallowed with a glass of water. The subjects then fasted for a further 2 hours to reduce the influence of food on absorption. During the study, the healthy volunteers performed normal indoor activities, whilst the patients were confined to bed and were only allowed to wash themselves and go to the toilet as usual. Other diuretics were withheld for at least 3 days before hydroflumethiazide was administered, and throughout the whole sampling period. Blood was collected immediately before and hourly for about 9 hours after administration of hydroflumethiazide. The blood from the patients was taken in glass Vacutainer tubes containing sodium heparin, and from the healthy subjects in tubes heparinized in our laboratory. The blood was centrifuged immediately to obtain plasma, which was kept frozen at - 2 0 ° until analysed. Urine was collected 0-3, 3-6, 6-9, 9-24, and 24-36 hours after dosing. The urine volume was measured and an aliquot frozen at - 2 0 ° until analysed.
Calculations After correction for blank values, a least squares regression line through zero was calculated for every standard curve, using reciprocal weighing. Absolute recovery of hydroflumethiazide from plasma and urine was estimated by extraction of spiked aliquots of plasma and urine, and the amount of drug in the final sulphuric acid phase expressed as a percentage of the amount in the spiked aliquot.
O. Brors et al.: Hydroflumethiazide in Human PIasma and Urine
151
Table 1. Mean coefficient of variation a of duplicate analysis of plasma and urine samples Plasma
Urine
Concentration range (ng/ml)
Mean coefficient of variation (%)
Number of samples
Concentration range (~,g/ml)
Sample size 0xl)
Mean coefficient of variation (%)
Number of samples
4 0 - 100 t 0 0 - 200 2 0 0 - 500 500-1200
5.4 2.4 3.2 2.0
6 6 11 13
4 - 10 4 - 10 10- 50 50-150
50 500 50 50
19.7 25 2.0 1.4
6 5 14 9
100 ,n a Calculated by the formula - - ] / 2 Z [A, - B i / (A i + Bi)] where Ai and B~are the results of duplicate analysis of the same sample, and n the number of samples n i= I Table 2. Sex, age, principal diagnosis, functional heart group (New York Heart Association classification), and serum creatinine of the
subjects investigated Subject no
Sex
Age (yrs)
Main diagnosis
Functional heart group
Serum creatinine (mg/lO0 ml)
1 2
M F
65 74
IV IV
1.4 1.1
3
F
70
III
1.0
4
M
60
II
2.4
5 6 7
F M M
61 38 29
Cqronary h.d. Arteriosclerotic and hypertonic h. d. Arteriosclerotic and hypertonic h. d. Subacute glomerulonephritis Coronary h.d. Healthy Healthy
(II)-III I I
1.5 0,9 1.0
M = male, F = female, h. d. = heart disease
The slopes of the straight part of the plasma concentration curves, in a semilogarithmic plot, were calculated by least squares regression lines, and plasma half lives were calculated from the slopes so obtained.
Results
Analysis At excitation and emission wavelengths of 279 and 390 nm, respectively, there was a linear relationship between the concentration of hydroflumethiazide in the solution and the intensity of fluorescence, up to a concentration of 6 ~tg/ml. All the measurements were performed within this concentration range. Absolute recovery of hydrofiumethiazide from plasma was 83%, and from urine 33%, when 2 ml of the organic phase was evaporated. Ommission of (NH4)2SO4 from the analysis without other changes in
the procedure reduced the recovery from plasma to 68% and from urine to 24%. The slopes of standard curves run in plasma and in 3% albumin solution were not significantly different. If the slope of the plasma standard curve was taken as 1.000, the relative slope of the corresponding albumin curves, run in the same analysis, as 1.007 +_ 0.017 (mean + standard deviation, 4 pairs of curves). The mean correlation coefficient of the standard curves was 0.9987, and that of the urine standard curves was 0.9990. Standard curves for plasma and 3% albumin were linear down to 10 ng/ml when 1 ml aliquots were analysed. This corresponded to the lower limit of detection of hydroflumethiazide in plasma, as the blank reading for plasma corresponded to 2 0 2 5 ng/ml of drug. Using a 500 ~tl sample, the limit of detection of hydroflumethiazide in urine was 100 ng/ml, the blank value being 200-250 ng/ml. The mean coefficient of variation between duplicate determinations at different concentration ranges is listed in Table 1.
152
O. Brors et al.: Hydroflumethiazide in Human Plasma and Urine
Table 3. Plasma concentration (ng/ml) and plasma half life of hydroflumethiazide Hours after dose
Plasma half life (h)
Subject no
1
2
3
4
5
6
7
8
9
1 2 3 4 5 6 7
440 236 499 788 713 e 403 171 k
902 574 871 836 727 f 259 446
1073 688 645 702
989 549 525 512 a 690g 202 384
846 430 362 355 b 516 h 117 190 m
783 363 254 334 436 89 159
744 280 211 270 c 362 63 105
694 233 168 2090 222 47 88
640 202 147
198 i 658
8.6 3.4 2.7 3.8 3.5 2.1 1.9
220 50 89
Time point of sampling (h): a = 3.8, b = 5.3, c = 7.2, d = 8.5, e = 1.4, f = 2.5, g = 3.6, h = 4.7, i = 3.2, k = 1.3, m = 5 . 6
EXCRETION RATE mglh
Plasma concentration
subject 2
ng/ml 2000-
subject 3
subject/*
10005000 12 2f, EXCRETION RATE mg/h
200I0050-
subject '~ 1
• o
36
subject 5
0
12 24 12 2/* 36 0 TIME AFTER DOSE (h)
subject 6
36
0
12
2t,
subject 7
2
3 4
subject o
5
a
7
20:
4
6
8
10
12 0
2
4
6
8
10
Time after dose (hi
0~ 0
12 24
0 12 24 316 0 36 TIME AFTER DOSE (h)
12
2/*
316
Fig. 1. Plasma concentrations of hydroflumethiazide in 7 subjects after a single oral dose. The left side shows curves from the patients and the right side those from healthy volunteers
Fig. 2. Excretion rate of hydroflumethiazide in urine in 7 subjects after a single oral dose
Table 4. Urinary excretion rates of hydroflumethiazide (mg/h)
PharrnacokineticStudy
Subject
Time interval (hours after dose)
no
0-3
3-6
6-9
9-24
24-36
1 2 3 4 5 6 7
1.1 1,8 0,83 1.4 3.7 7.4 9.9
4.6 6.0
1.3 2.7 2.5" 0.76 2.0 1.4 2.5
1.8 0.45
1.1 0.19 0.80 b
1.8 5.3 4.4 7.1
0.51 0.81 0.38 0.67
Time point of sampling (h): a = 3-9, b = 9-33
0.25 0.12 0.11
Details of the subjects in the study have been listed in Table 2. It can be seen that renal function, as judged by serum creatinine, was considerably reduced in subject 4 and slightly reduced in subjects 1 and 5. Liver function (not listed) was normal in all patients except subject 1, who had a serum bilirubin of 3.1 mg/100 ml. Plasma concentrations measured in the 7 subjects after a single oral dose of hydroflumethiazide are listed in Table 3, and in Figure 1 the data are shown as a semilogarithmic plot. The maximum plasma con-
O. Brors et al.: Hydroflumethiazide in Human Plasma and Urine
153
Table 5. Body weight, dose, cumulative urinary excretion and renal clearancea of hydroflumethiazide Subject no
1 2 3 4 5 6 7
Body weight (kg)
75 62 46 75 49 75 90
Dose (mg)
75 50 50 75 75 75 100
Cumulative urinary excretion
Renal clearance
(% of dose)
(h)
(1 h -1)
(1 h -1 kg -1)
78.3 74.8 65.9 25.8 66.3 67.7 72.9
35.0 35.3 32.4 24.0 36.0 37.1 30.8
3.1 9.1 4.8 3.0 6.9 32.8 26.0
0.041 0.15 0.10 0.040 0.14 0.44 0.29
" Calculated by the formula U0-x •whereUisthecumu•ativeexcreti•ninurineandAUCtheareaunderthep•asmac•ncentrati•ncurve•x 9 hrs AUC0_x
centration was obtained 1-3 hours after dosing. The plasma half life was calculated for the last part of the curves (Table 3). They were longer in patients than in the healthy subjects, and there was considerable variation in half life amongst the patients. The urine concentration of hydroflumethiazide varied over the range 2-140 ~tg/ml. Urinary excretion rates are listed in Table 4 and are shown graphically in Fig. 2. The cumulative excretion of hydroflumethiazide in the urine is shown in Table 5, together with the calculated renal clearance of the drug during the first 9 hours after the single dose. Renal clearance in patients was much lower than in the healthy subjects, even when corrected for differences in body weight. No metabolite was detected by TLC of urine extracts.
Discussion
The spectrofluorometric method described has the same limit of detection as that reported previously for a method based on spectroftuorodensitometry [16]. The method proved successful in determining plasma concentrations for 9 hours after dosing in the seven subjects investigated. As judged from the plasma concentration curves, it might be possible to determine plasma concentration for a much longer time in patients after a single close of hydroflumethiazide. Plasma levels of hydrochlorothiazide, determined with a GLC method after a single oral dose of 50 rag, were in the range 50-350 ng/ml 1.5 to 4 hours after dosing, after 100 mg a maximal concentration of 710 ng/ml was found after 1.5 hours [17]. These concentrations were in the same range as was observed in the present study, and these two thiazides are reported to have the same potency [3].
The finding that all the subjects, except subject 4, excreted 65-79% of the oral dose in urine during the first 31-37 hours after dosing (Table 5), is in agreement with earlier studies [3], and confirms renal excretion as the main elimination pathway for hydroflumethiazide in subjects with normal renal function. The plasma half life of hydroflumethiazide might be expected to be correlated closely with renal function, provided other routes of elimination remain unchanged. As the plasma half life of subject 4 was found to be as short as 3.8 hours, in spite of reduced renal function, increased biliary excretion or metabolism must be considered in this patient. In subject 1, the reduction in both renal and hepatic function was associated with more than doubling of the plasma half life of hydroflumethiazide, presumably because biliary excretion or metabolism could not compensate for the reduced renal elimination. The renal clearance of hydroflumethiazide was considerably lower in the patients than in the healthy subjects, in whom renal clearance was almost as high as the expected renal plasma flow, and several times higher than the anticipated glomerular filtration rate. The high renal clearances must be explained, therefore, by tubular secretion of hydroflumethiazide. Impairment of the renal circulation, known to occur in heart failure, may have been responsible for lower renal clearance of hydroflumethiazide in the patients with heart failure in comparison to the healthy subjects. The high cumulative excretion in urine over 36 hours, as seen in subject 1, may be explained by the reduced liver function and prolonged half life of the drug in plasma. Judged by the area under the plasma concentration curve and the plasma half life of the drug in subject 4, the amount absorbed from the gastrointestinal tract in this patient should be comparable to that in the other subjects. Therefore, normal liver function in this patient was probably responsible for elimination
154
of the greater part of the drug, presumably by excretion in the bile or by metabolism. The latter is not known to occur in man or animals, and no metabolite was detected in this study. The urinary excretion rate of hydroflumethiazide observed in the healthy subjects 24-36 hrs after dosing is higher than might be expected from a plasma half life of 2 hrs. The reason for the discrepancy might be enterohepatic recirculation of the drug, or a later phase of elimination with a longer half life than the initial phase. Further investigations are under way for evaluation of this phenomenon. Based on observed interindividual differences in plasma concentration levels and plasma half life after a single oral dose of hydroflumethiazide, a variable intensity and duration of pharmacological effect might be anticipated in different subjects. Especially when multiple dosing is concerned, as in the treatment of hypertension and chronic oedema, this may be of relevance for individualization of drug therapy. Therefore, in order to establish a rational therapeutic regime for the thiazides, further studies are required.
Acknowledgements. O. Brors was a Research Fellow of the Norwegian Research Council for Science and the Humanities. Financial support from this Council, from The Norwegian Drug Monopoly, and from Anders Jahre's Foundation is gratefully acknowledged. We wish to thank Mrs. Eva Sigfridsson and Renuka Patel for skilfuI technical assistance.
References 1. Blagg, C. R.: Hydroflumethiazide- a new oral diuretic. Lancet 1959 II, 311-313 2. Dettli, L., Spring, P.: Vergleichende pharrnakologische Priifung eines neuen Salidiureticums (Cyclopenthiazid) am gesunden Menschen. Z. ges. exp. Med. exp. Chir. 134, 310-322 (1960) 3. Young, D. S., Forrester, T. M., Morgan, T. N.: A comparison of the diuretic action of the chlorothi~zide analogues. Lancet 1959 II, 765-768
O. Brors et al.: Hydroflumethiazide in Human Plasma and Urine 4. Matheson, N. A., Morgan, T. N.: Diuretic action of chlorothiazide. Lancet 1958 I, 1195-1199 5. Ford, R. V., Moyer, J. H., Spurr, C. L.: Clinical and laboratory observations on chlorothiazide (Diuril). Arch. int. Med. 100, 582-592 (1957) 6. Tapia, F. A.: Diuretic and antihypertensive effects of polythiazide. Curr. ther. Res. 3, 365-377 (1961) 7. Ford, R. V.: Clinical pharmacological investigation of polythiazide, a potent oral diuretic agent. Curr. ther. Res. 3, 320-329 (1961) 8. Hutcheon, D. E., Leonard, G. B.: Therapeutic efficacy of diuretics with different durations of action. J. Amer. reed. Ass. 185, 6 4 0 6 4 2 (1963) 9. Beevers, D. G., Harpur, J. E., Hamilton, M.: The long-term treatment of hypertension with thiazide diuretics. Postgrad. med. J. 47, 639-643 (1971) 10. Maronde, R. F., Milgrom, M., Dickey, J. M.: Potassium loss with thiazide therapy. Amer. Heart. J. 78, 16-21 (1969) 11. Skovbo, P., Bjerregaard, P., Hvidt, R., Jesting, E , Lorentzen, F., Hvidt, S.: Diuretika og hypokali~emi. Ugeskr. L~g. 134, 1043-1047 (1972) 12. Reutter, F., Schaub, F.: Harnsiiurestoffwechsel and Salidiuretica. Dtsch. med. Wschr. 89, 1101-1104 (1964) 13. Heimsoth, V., Harmann, F.: Untersuchungen zur St6rung des Harns~iurestoffwechsets nach Saluretica-Verabreichung. Dtsch. reed. Wschr. 90, 1905-1908 (1965) 14. Roediger, P. M., Alden, J., Beardwood, D., Hutchison, J. C.: Benzothiadiazines and diabetes mellitus. II. Comparison of hyperglycemic effects of bendroflumethiazide and chlorothiazide in hypertensive patients with diabetes mellitus. Curr. tiler. Res. 6, 670--677 (1964) 15. Runyan, J. W.: Influence of thiazide diuretics on carbohydrate metabolism in patients with mild diabetes. New Engl. J. Med. 267, 541-543 (1962) 16. Garceau, Y., Davis, J., Hasegawa, J.: Quantitative TLC procedure for determination of hydroflumethiazide in biological fluids. J. pharm. Sci. 63, 1793-1795 (1974) 17. Vandenheuvel, W. J. A., Gruber, V. F., Walker, R. W., Wolf, F. J.: GLC analysis of hydrochlorothiazide in blood and plasma. J. pharm. Sci. 64, 1309-1312 (1975)
Received: May 17, 1976, and in revised form: August 26, 1976, accepted: September 2, 1976 Dr. O. Brors Institute of Pharmacology University of Oslo Box 1057, Blindem N-Oslo 3 Norway
Responsiblefor the text: Prof. Dr. H. J. Dengler,Bonn and Prof. Dr. F. Gross, Heidelberg Responsiblefor advertisements:L. Siegel,Kurfiirstendamm237, D-1000 Berlin 15, Springer-Verlag Beriin-Heidelberg-NewYork, Printed in Germanyby aprinta, Wemding Copyright© by Springer-VerlagBerlin. Heidelberg 1977
