Effects of Magnesium Oxide on Trichlormethiazide Bioavailability

Effects of Magnesium Oxide on Trichlormethiazide BioavailabiIity HARUMI TAKAHASHI',YASHUSHI WATANABE, HIDEOSHIMAMURA, AND KEIKOSUGlTO Received August 20, 1984, from the Department of Pharmacy, Mevi College of Pharmacy, 1-22-1, Yato-cho, Tanashi-shi, Tokyo 188, Accepted for publication March 13, 1985. Japan.

Abstract 0 The effect of an antacid, magnesium oxide (MgO), on the bioavailabilityof a thiazide diuretic, trichlormethiazide(l), was studied in 10 healthy subjects who fasted overnight. A single oral dose of 4 mg of 1 alone or in combination with 0.5 g of MgO was given in a two-way Latinsquare crossover design. Urine concentrationsof lduring the 24 h after each dose were determined by an HPLC method. There were no significant differences for drug alone versus drug with MgO in the mean percentage recovery (60 versus 62%) and the cumulative amount excreted unchanged in urine over 24 h (2408 versus 2463 pg). However, coadministrationof MgO increasedthe mean excretion rate of 1 at 0.75 h and 1.5 h (p < 0.05), the cumulative amount excreted unchanged in urine over 2 h (p < 0.05), and the absorption rate constant (p < 0.05).Therefore, the extent of bioavailability was not influenced by MgO, but the rate of absorption was enhanced. The solubility of 1 increased remarkably by changing from pH 1.2 to 8.0 (141 to 1365 pg/mL). The dissolution rate of 4 mg of 1 in 500 mL of medium was not affected by an increase in pH. However, a 1.5-fold increase of the dissolution rate in 20 mL of medium was observed by changing from pH 1.2 to 7.3. The results suggest that the GI absorption of 1, a poorly soluble weak acid, is hastened by MgO, which probably raises gastric pH sufficiently to increase not only the solubility but also the dissolution rate of 1 in the small volume of the gastric fluids. The clinical significance of the above findings, however, remains uncertain.

The current approach to the management o f hypertension involves the use o f a thiazide diuretic as t h e drug o f choice.' Trichlormethiazide (1), a thiazide diuretic that i s administered at relatively l o w doses (2-4 mg daily), i s often coadministered w i t h other drugs (85% o f 713 prescriptions; average, 2.3 drugs).2 A wide variety of medications used w i t h 1in t h i s 1983 survey included antihypertensive agents such as methyldopa, c a l c i u m - c h a n n e l antagonists, h y d r a l a z i n e , pblockers, and spironolactone (53% o f 713 prescriptions), potassium supplements (21%), peptic ulcer medications such as sucralfate, antacids such as magnesium oxide (MgO) (14%), and m u l t i v i t a m i n s (11%). Antacids often alter the absorption and excretion o f coadministered agents by reducing o r enhancing drug dissolution due to pH changes, interacting w i t h drugs in t h e GI tract, delaying gastric emptying, and/or elevating u r i n e pH,= the l a t t e r o f w h i c h may influence tubular reabsorption. W h i l e a number o f pharmacokinetic studies o f thiazides and t h e i r interaction w i t h other drugs have been performed,6*7 there are l i m i t e d data o n the concurrent administration o f antacids w i t h thiazide diuretics. Therefore, t h i s study was conducted to clarify t h e effect of an antacid, MgO, o n the pharmacokinetics and the possible alteration in pharmacodynamics o f the thiazide diuretic, trichlormethiazide.

Experimental Section Material-Trichlormethiazide was kindly supplied by Shionogi Co., Ltd. (Osaka, Japan). Acetonitrile was HPLC grade and all other chemicals were reagent grade. Subjects-Ten healthy volunteers (1 female and 9 males, aged 862 / Journal of Pharmaceutical Sciences Vol. 74,No. 8, August 1985

between 20 and 24 years, mean 21.9 years, and weighing 48-78 kg, mean 62.9 kg) participated in the study. Before the study, all subjects underwent a physical examination and routine laboratory tests including urinalysis, biochemical examination, and a hemogram. All tests showed values within the normal limits. Clinical Protocol-The study was carried out in a single-dose, two-way, Latin-square crossover fashion. The drug was administered on two separate occasions a week apart according to the dosage schedule. The subjects had fasted overnight and until 2 h postdose. Each subject took two tablets (4 mg in total) of either 1 alone or 1 combined with 0.5 g of MgO, orally, with 200 mL of water. Urine samples were collected just prior to the dose and at the following time intervals: 0-0.5, 0.5-1, 1-2, 2 3 , 3-4, 4-6, 6-8, 8-10, 10-13, and 13-24 h postdose. The volume and pH of each urine sample was measured. An aliquot was frozen until the time of analysis. Sodium and potassium ions were determined in urine samples using selective ion-sensitive electrodes (NOVA-1, Niko Bioscience, Ltd.). The urine volume, electrolytes excretion, and NaiK excretion ratio between the two treatments during the first 2 h and at 24 h were compared by employing the paired t test. Trichlormethiazide Analysis-A volume (0.5 mL) of urine was placed in a 10-mL glass-stoppered centrifuge tube and extracted by the addition of 1.0 mL of 0.01 M phosphate buffer (pH 7.0),3.0 g of sodium chloride, and 9.0 mL of ether. The tube was vigorously shaken for 10 min, then centrifuged at 2500 rpm for 10 min. A 7.0mL aliquot of the organic layer was transferred to a 30-mL centrifuge tube and evaporated to dryness in a centrifugal evaporator at 50°C. The residue was reconstituted with 100 pL of 80% methanol containing methyl p-hydroxybenzoate (50 pg/mL) which was used as an external standard. A 20-pL aliquot was then injected into the HPLC. The chromatograph consisted of a pump (Shimadzu LC-3A) fitted to a reversed-phase column (Zorbax ODs, 15 cm x 4.6 mm). The sample was injected using a n automatic sampler (Waters 710B). A variable-wavelength UV detector (Shimadzu SPD-2A) was set at 274 nm, a t a range of 0.04 AUFS. The mobile phase consisted of 20% acetonitrile in 0.01 M sodium phosphate buffer (pH 7.0). The flow rate was set a t 0.8 mL/min and the column temperature was kept at 35°C. The amount of 1 was estimated by a peak height ratio method by using a n integrator (Shimadzu C-RlA). Solubility-The solubility of 1 was spectrophotometrically determined at 37°C after filtration of 1 through a 0.45-pm membrane filter (TM-2, Toyo Roshi Co., Ltd.). Measurements were repeated 6 times every 24 h until the solution had been equilibrated. The pH values of the media used in determining solubility were adjusted to pH 1.2 with dilute HC1, to pH 3.0 and 5.0 with 0.1 M NaOAc buffer, to pH 6.8 with 0.05 M phosphate buffer, and to pH 8.0 with 0.1 M phosphate buffer. Dissolution-The dissolution rate of 1 was determined at 37°C with two tablets of 1 (4 mg in total). The amount of the drug dissolved was monitored spectrophotometrically at 274 nm. The following two methods were used for assessing the dissolution rate. Paddle Method-According to the JP X procedure: 500 mL of dissolution medium (pH 1.2, dilute hydrochloric acid solution; pH 6.5,0.05 M sodium phosphate buffer) was used. The stirring rate was 100 rpm. A sample (20 mL) of the medium was passed through a glass filter (3G) to a flow cell at appropriate intervals and returned to the dissolution vessel. The dissolution rate was determined as a percentage of the asymptotic absorbance which was taken as the value of 100%dissolution. An average dissolution rate was obtained in triplicate.

0022-3549/85/0800-0862$0 1.OO/O 0 1985, American PharmaceuticalAssociation

Results

Beaker Method-The beaker method employed 20 mL of dissolution medium (pH 1.2; dilute HCl; pH 4.3, 0.1 M NaOAc buffer; pH 6.6, 7.3, and 8.0, 0.05 M sodium phosphate buffer) was used. A cylindrical, flat-bottomed vessel (4.0 cm, i.d.) was employed and the mixture was stirred a t 100 rpm. The dissolution rate after 30 min was expressed as a percentage of dissolved amount to 4 mg of 1. Pharmacokinetic Analysis-The following parameters were compared between tHe two treatments with and without MgO by using the paired t test: 1. The mean excretion rates at the midpoint time of urine collection periods (0.25, 0.75, 1.5, 2.5, 3.5, 5.0, 7.0, 9.0, 11.5, and 18.5 h). 2. The mean percentage recovery over 24 h, [f,lg4. 3. The cumulative amount excreted unchanged in urine over 2.0 h, [X,lg ', and 24 h, 4. The mean maximum urinary excretion rate, [dX,ldt],,,. 5. Time of maximum urinary excretion rate, t,,,. 6. The absorption rate constant, k,. 7. The elimindtion rate constant, kel. The k, and kel values were calculated on the basis of a one. ~ all the compartment open model using the NONLIN p r ~ g r a m For analyses, the minimal level of statistical significance was taken as p < 0.05. The power analysis of variance was also employed to estimate the minimum detectable difference (a = 0.05, 1 - p = 0.8) in a Latin-square crossover design.10

The percent recovery of 1 from urine over a concentration range of 10-40 pg/mL averaged 99.3 5.6% ( * SD) in our assay. The calibration curves over a concentration range of 10-40 pgimL exhibited excellent linearity with correlation coefficients of at least 0.997 for each curve. The detection limit (2 x baseline) was 2 pgfmL for 1. Within- and betweenassay coefficients of variation (CV) were 1.1 and 1.6%, respectively, a t a concentration of 40 pg/mL. The mean urinary excretion rate and the cumulative amount excreted unchanged in the urine are shown in Fig. 1 for 1 after the oral administration of 1 (4 mg) alone or combined with MgO (0.5 g). The pharmacokinetic parameters of 1 are listed in Table I. The interindividual variability of each parameter was large. The percentage and absolute amounts of 1 recovered in urine 24 h postdose ([ful:4, percent of dose, and [Xu1g4,pg) were similar in both cases (1 alone versus 1 with MgO, averaging 60 versus 62% and 2408 versus 2463 pg). However, the absorption rate of 1 represented by the urinary amount excreted unchanged in the early phase [Xultoand the mean excretion rates at 0.75 and 1.5 h were significantly (p < 0.05)

*

CI

m

3 .-e 5 .3 9

0,250,151.52.53.5 5.0

7.0

11.5

Figure 1-Mean urinary excretion rate and amount of trichlormethiazide(7) excreted unchanged in urine of 10 healthy subjects after an oral administrationof 4 mg of 7 alone (0)and with magnesium oxide (0).The data are plotted as mean f SEM. *Between the two treatment groups, p < 0.05.

18.5

Time (h)

Table I-Pharmacokinetic Parameters of TrichlormethiaziUe (1) Obtained After an Oral Administration of 4 mg With and Without Magnesium Oxide (MgO) to 10 Healthy Subjects' ~~

Kinetic Parameters

[f,J204

%

[xu12!pCLS

[Xu1 M [dXu/dtImax, d h 4naX3 h ka, h-' ki, h-' 01

Trichlormethiazide Alone

*

5.7 60.2 623.7 f 51.4 2408.3 2 226.7 669.5 95.6 2.35 f 0.3 1.05 0.2 0.64 2 0.1

* *

Trichlormethiazide With Magnesium Oxide 61.6 1159.8 2462.6 891.5 1.43 3.42 0.59

*

7.6 182.8 302.2 165.4 f 0.1 0.8 f 0.1 2 2

*

Significance NS p < 0.05 NS NS p < 0.05 p < 0.05 NS

~~

Minimum Detectable Difference (a = 0.05, 1 - p = 0.8) 0.306 0.851 0.306 0.649 0.441 2.864 0.377

= BThedata are expressedas mean 2 SEM (n = 10).Abbreviations: [fu]',"= mean percentage of 1 recoveredin urine during 24 h postdose; amounts of 1 recovered,inurine during 2.0 h postdose; [Xu]':= amounts of 1 recovered in urine during 24 h postdose; [dX,/dt], = mean maximum urinary excretion rate; tmaX= time of maximum urinary excretion rate; k, = absorption rate constant; kl = elimination rate constant; NS = not statistically significant.

Journal of PharmaceuticalSciences / 863 Vol. 74, No. 8, August 1985

increased with the coadministration of MgO. Although k,, showed similar values in both treatments, a significant increase in k , (p < 0.05) and a significant reduction oft,,, (p < 0.05) were found in the combined treatment of 1 with MgO. The differences of urine volume over 2-24 h did not reach a statistically significant level between 1 alone and 1 with MgO. Sodium and potassium excretion within 2 and 24 h remained a t almost the same levels for both treatments. However, the sodium-potassium excretion ratio was increased significantly within 24 h (p < 0.05) in the group receiving 1 concomitantly with MgO (Table 11). Changes in the solubility of 1 remained relatively minor with a change in pH from 1.2 to 6.8 (141 to 380 pg/mL), whereas a remarkable increase in the solubility (1365 pg/mL) was found when the pH was fixed a t 8.0. These findings are shown in Fig. 2. Although 1 is a poorly soluble drug, the dissolution of 4 mg of 1 in 500 mL of test solution was rather rapid a t both pH 1.2 and 6.5. Increasing the pH did not produce an effect on the dissolution rates of 1 in 500 mL of dissolution medium, as shown in Fig. 3. Table I11 summarizes the dissolution rates of 4 mg of 1 in 20 mL of medium. A 1.5-fold increase in the dissolution rate was observed by increasing the pH from 1.2 to 7.3.

Discussion Thiazides, primarily excreted unchanged in urine, enhance the renal excretion of sodium and chloride ions along with accompanying water. This occurs because of inhibition of ion reabsorption in the distal tubule of the nephron.11 The measurement of drug at or near the site of action is preferable in assessing pharmacological and therapeutic effects. Therefore, the urinary determination of thiazides, instead of measurement in the blood, appears to be the logical choice for

bioavailability studies of diuretics.'z For this reason, the urinarj; concentration of 1 has been measured in this study. The absorption rate constant, k,, estimated from a few data points,.when the rate of absorption is high, may not be reliable for assessing the rate of absorption. However, the significant increase of k, with MgO was associated with an 33% increase in [dXu/dtl,,,, a reduction of tmax, and an enhanced excretion rate and amount of 1 in the early phase (Table I, Fig. 1).All of these findings indicate a hastened absorption of 1 by the addition of MgO. As [Xu]:* was similar for both treatments, MgO appears not to affect the extent of bioavailability of 1. Regular administration of certain antacids can increase urine pH sufficiently to produce substantial effects on the elimination kinetics of certain acidic and basic drugs.13J4 Gibaldi et a1.16 reported that a usual dose of MgO administered for 7 d increased urine pH from 0.4 to 0.5. In this study a single dose of MgO (0.5 g), which did not induce any sighificant effect on urine pH (data not shown), resulted in no change in the elimination rate constant (k,1) of 1 (Table I). Since MgO appears to affect the rate but not the extent of bioavailability, we assessed the pharmacodynamic effects by studying the data after the first 2 h of 1 dosing, as well as the 24-h data. The greater interindividual variability in these data might be responsible for the inability to detect an early increase in the urine volume or the excretion of sodiumpotassium a t 2 h. The sodium-potassium excretion ratio indicates the net balance between sodium diuresis and hypokalemia, a clinically important side effect of thiazides.16The exact mechanism for an increase in the sodium-potassium excretion ratio after coadministration of the single dose of MgO is totally u&nown from this study. Most drugs must be dissolved in the milieu of the duodenum to be absorbed effectively. Maekawa et al.17observed an

Table il-Comparison of Urine Volume, Electrolytes Excretion, and NdK Excretion Ratio Between the Two Treatmentse

Urine Volume, mL Drug Alone Drugbwith MgO Significance

Na Excretion, mEq

24 h

2h

24 h

2h

24 h

2h

24 h

223.4 2 35.1 199.7 f 22.7

1336.12 97.2 1469.3 & 119.8

23.8 2 3.1

195.1 -r- 13.8 191.3 2 9.9

4.7 -r- 0.7 4.3 f 0.6

32.6 2 3.2

27.4 k 3.2

5.5 0.6 7.0 2 0.7 NS

6.2 2 0.2 7.3 f 0.4 p < 0.05

NS

NS f

SEM (n

=

NS

NS

10). Abbreviations: MgO

=

\

-::

C

.-w0

-

-

H I

400 -

3

200

s

I

-

50-

,- 800 U

not statistically significant. bDrug refers to

o!

m

.-

=

v

1000 -

600

NS

-x

3 1200 -

0

r

26.9 2 1.9

NS

magnesium oxide; NS

100

c

Na/K Excretion Ratio

2h

aThe data are given as mean trichlormethiazide.

fi

K Excretion, mEq

'

0 .I

20

10

30

Time (Min)

Figure 3-Dissolution curves of 4 mg of trichlormethiazide (1) at iwo different pH values in 500 mL of the medium. Key: (0) pH 1.2;(0)pH 6.5.

864 / Journal of Pharmaceutical Sciences Vol. 74, No. 8, August 1985

Table Ill-Effect of pH on Dissolution Rate of 4 mg of Trichlormethiazidein 20 mL of the Medium Dissolution Rate at 30 min, YO

62.3 67.3 83.6 92.3 94.7

1.2 4.3 6.6 7.3 8.0

early increase in gastric pH at -8 by the oral administration of a single 0.3-g MgO dose. The results of our solubility studies suggested that the increase of the absorption rate might be attributed to the increased solubility of 1by MgO,lE which raises gastric pH sufficiently to exceed the pKal (6.8)of 1.19

The dissolution of a poorly soluble weak acid like 1 is thought to be a rate-limiting step in its absorption. The dissolution rate of 4 mg of 1in 500 mL, however, did not show any increase but was influenced greatly in 20 mL by increasing the pH. Water alone transferred rapidly from the stomach when solid preparations were administered along with water.20,21Hence, in the fasting state, the volume of the gastric fluids which was used to dissolve 4 mg of 1 appears to be much smaller than the volume of water taken with 1 (200 mL) in this study. Magnesium oxide in the small volume of the medium showed a remarkable effect as an antacid.17 Therefore, it is reasonable to speculate that the gastric pH may be increased by MgO to enhance not only the solubility but also the dissolution rate cf 1 in the smaller volume of gastric fluids.

References and Notes 1. Moser, M. Post rad Med 1983, 73, 199-210. 2. Watanabe, Y. A n . Clin. Pharmacol. Ther. 1984,15,255-258.

i.

3. Hurwitz, A. Clin. Phurmacokinet. 1977,2, 269-280. 4. Greenblatt, D. J.; Allen, M. D.; MacLaughlin, D. S.; Harmatz, J. S.; Shader, R. I. Clin. Pharmacol. Ther. 1978,24, 600-609. 5. Osman, M. A.; Patel, R. B.; Schuna, A.; Sundstrom, W. R.; Welling, P. G. Clin. Pharmacol. Ther. 1983, 33, 465-470. 6. Jordo, L.; Johnsson, G.; Lundborg, P.; Persson, B. A.; Regardh, C. G.; Rtinn, 0. Br. J . Clin. Pharmacol. 1979, 7, 563-567. 7. Sundquist, H.; Anttila, M.; Simon, A.; Reich, J. W. J . Clin. Pharmacol. 1979,19, 557-564. 8. “The Japanese Pharmacopeia,” 10th rev.; Society of Japanese Pharmaco oeia: Tokyo, 1981; pp 760-763. 9. Metzler, M. “A User’s Manual for NONLIN,” Technical Report 7292/69/7292/005; The Upjohn Co.: Kalamazoo, MI, 1969. 10. Wagner, J. G. “Fundamentals of Clinical Pharmacokinetics”; Dru Intelligence Publications, Inc.: Hamilton, IL, 1975;pp 285. 11. Gooiman, L. S.; Gilman, A. “The Pharmacological Basis of Therapeutics”; Macmillan: New York, 1980; pp 899-902. 12. Shah, V. P.; Hunt, J. P.; Prasad, V. K.; Cabana, B. E. J . Phurm. Sci. 1981, 70, 833-835. 13. Verbeeck, R. K.; Branch, R. A.; Wilkinson, G. R. Clin.Pharmacol. Ther. 1981, 30, 619-628. 14. Perez-Reyes, M.; Di Guisep i, S.; Brine, D. R.; Smith, H.; Edgar Cook, C. Clin. Pharmacol. 8her. 1982,32, 635-641. 15. Gibaldi, M.; Grundhofer, B.; Levy, G. Clin. Pharmacol. Ther. 1974,16, 520-525. 16. Perez-Stable, E.; Caralis, P. V. A m . Heart J . 1983,106,245-251. 17. Maekawa, H.; Takagishi, Y.; Doi, Y.; Ohsumi, K.; Otomune, T. Yakuzaiaku 1971.31.266-272. 18. HurwitzyA. J . Phrmacol. Exp. Ther. 1971,179, 485-489. 19. Goto, S.; Odawara, Y.; Nakano, M.; Araki, Y. Yakugaku Zasshi 1978.98 236-241 1978,98,236-241. , - - , -- - -- . 20. AoGgi, Aoyagi, N.; Ogata, H.; Kaniwa, N.; Koibuchi, M.; Shibazaki, T.; Eiima, A. J . Phurm. Sci. 1982. 71. 1165-1169. 21. Aoya ‘i, N.; 0 ata, H.; Kaniwa,’N.; Ejima, A.; Yasuda, Y.; Taniota, Y. J . %harm. Dyn. 1984, 7, 7-14.

d:

Acknowledgments We wish to thank Miss Kayoko Itoh for her excellent technical assistance and Dr. Hiroyasu Ogata, Division of Drugs, National Institute of Hygienic Sciences, and Dr. Takashi Ishizaki, Division of Clinical Research Institute, National MediClinical Pharmacolo cal Center, for their g l p f u l discussion.

Journal of Pharmaceutical Sciences / 865 Vol. 74, No. 8,August 1985