Pregnancy-induced increase in metoprolol

metabolism
Five women who developed hypertension during pregnancy received metoprolol, 10 mg iv; 3 days later
they received metoprolol, 100 mg by mouth. Blood and urine samples were collected after each dose. The
same procedure was repeated 3 to 6 months after delivery. The apparent oral clearance of metoprolol
during pregnancy exceeded that after pregnancy by a factor of 2 to 13. As a result, after oral dosing the
peak plasma concentrations during pregnancy were only 12% to 55% those after delivery, and the plasma
AVCs were reduced to the same extent. Oral bioavailability increased by a factor of 1.3 to 3.7 after
pregnancy. Systemic clearance after pregnancy was 26% to 97% that during pregnancy, but this difference
was not significant. Metoprolol plasma protein binding was the same on both study occasions. Our data
cannot be explained by a change in gastrointestinal absorption, because the urinary recovery of metoprolol
and its metabolites was slightly higher during pregnancy. It is concluded that the greater metoprolol
clearance during pregnancy results from increased hepatic metabolism of the drug. (CLIN PHARMACOL
THER 37:688-692, 1985.)

Stellan Hogstedt, M.D., Bo Lindberg, M.D., Ph.D., Dun Ren Peng, M.D.,
Carl-Gunnar Regardh, Phann.D., and Anders Rane, M.D., Ph.D.
Viisteras, Uppsala, Molndal, and Huddinge, Sweden

During the last few years, cardioselective [3-blocking
drugs have largely replaced diuretics in the treatment
of hypertension during pregnancy. II One of the most
widely used drugs in this category is metoprolol, a
cardioselective [31-adrenoceptor blocker." The antihy-
pertensive therapy is often reinforced with hydralazine.
This combination has proved effective in most cases of
pregnancy-induced hypertension and, apart from the
known side effects of these drugs, the treatment appears
safe for both mother and fetus. 8 Despite the extensive
use of metoprolol, very little is known about its kinetics
during pregnancy. Pregnancy is associated with an in-
creased clearance of several drugs that are oxidized or
excreted directly by the kidneys. 1,9,10,13 We have re-
ported a marked increase in the apparent oral clearance
of metoprolol during pregnancy." Metoprolol is almost
entirely metabolized by the liver and, at least in subjects
From the Department of Obstetrics and Gynecology, Vasteras Hos-
pital, Vasteras; Department of Obstetrics and Gynecology, Uni-
versity Hospital, Uppsala; Department of Pharmacokinetics and
Drug Metabolism, AB Hassle, Molndal; and Department of Clin-
ical Pharmacology at the Karolinska Institute, Huddinge Hospital,
Huddinge.
Supported by grants from the Swedish Medical Research Council
(No. 04X-04496), The Expressen Prenatal Research Foundation,
and the Karolinska Institute.
Received for publication Nov. 8, 1984; accepted Feb. 14, 1985.
Reprint requests to: Stellan Hogstedt, M.D., Department of Obstet-
rics and Gynecology, Vasteras Hospital, 72189 Vasteras, Sweden.
688
who are not pregnant, it is completely absorbed after
oral dosing."
We studied metoprolol kinetics in five women during
and after pregnancy. The study was undertaken to elu-
cidate the mechanism behind the increased oral clear-
ance of metoprolol during pregnancy.
METHODS
Our subjects were five women 20 to 36 years old
who were studied the first time while in the third trimes-
ter of pregnancy. The second study was performed be-
tween 3 to 6 months after delivery. Clinical features of
the subjects are listed in Table I. Four of them developed
hypertension before the 20th week of pregnancy, while
the fifth (no. 5) developed a typical preeclampsia with
profound albuminuria (1.2 gm/24 hr the day before
delivery) in the 36th week of pregnancy. Other than
hypertension and albuminuria, subjects had no com-
plications during pregnancy. None was treated with any
drugs during pregnancy or during the postpartum study.
Subject 2 developed rheumatoid arthritis after delivery
but was not taking any drug during the second study.
Four subjects had a spontaneous vaginal delivery; the
subject with preeclampsia underwent Caesarean sec-
tion. All infants were healthy and had normal birth-
weights according to Swedish reference values.
During the first day of the study, 10 mg metoprolol
was injected intravenously into an arm vein in the mom-
VOLUME 37
NUMBER 6 Pregnancy-induced metoprolol metabolism 689

Table I. Clinical features of the subjects

Subject No.

1 2 3 4 5
No. of previous pregnancies/deliveries 0/0 0/0 0/0 0/0 I/O
Smoking (cigarettes per day) 0 <10 0 0 0
Body weight (kg) on both study occasions 80/68 81/71.5 95.5/88 78.5/69.5 74/63
Week in pregnancy/after delivery (on study occasions) 38/25 35/24 35112 37112 38114
Delivery week (completed) 38 36 39 39 38
Birthweight (gm) and sex 3500(F) 2540(M) 2880(M) 3150(M) 3270(M)
First appearance of hypertension (wk of pregnancy) 14 18 11 20 36
Maximum blood pressure during pregnancy (mm Hg) 150/100 150/100 160/105 1501110 1551120
F = Female; M = male.

ing after an overnight fast. Two hours after drug dosing Table II. Metoprolol kinetics after oral and
the subject ate a standardized breakfast. Blood samples intravenous doses in five women during and
were drawn through an indwelling catheter before, 5, after pregnancy
15, 30, and 45 minutes, and 1, 1.5, 2, 3, 4, 6, 8, and
During After
12 hours after dosing. Three days later, when virtually pregnancy delivery
the entire intravenous dose had been eliminated from
the body, 15 the subject received an oral dose of 100 mg Oral (100 mg)
tY2 (hr) 3.42 ± 0.80 3.42 ± 0.31
metoprolol. The same procedure as before was fol-
AUC (nmol . hr/L) 998 ± 517 3562 ± 809*
lowed, except that blood samples were drawn before, CIa (Umin) 9.56 ± 2.70 1.71 ± 0.39*
20 and 30 minutes, and 1, 1.5, 3, 4.5, 6, 8, and 12 CIa (Umin . kg) 0.118 ± 0.034 0.024 ± 0.006*
hours after dosing. All blood samples were centrifuged F 0.21 ± 0.06 0.42 ± 0.05t
within 10 to 60 minutes of sampling and the plasma Intravenous (10 mg)
tY2 (hr) 5.38 ± 1.36 5.36 ± 1.18
was frozen at - 20° C until analyzed. A small portion
Vd (Llkg) 6.87 ± 1.32 3.85 ± 0.43
of the blank plasma (collected before metoprolol dos- C!;, (Umin) 1.38 ± 0.26 0.65 ± 0.08
ing) was saved for the assay of the metoprolol plasma C!;, (Llmin . kg) 0.017 ± 0.003 0.009 ± 0.001
protein binding. Plasma protein bind- 9.18 ± 1.70 10.86 ± 1.06
Urine was collected over 24-hour intervals for 3 days ing (%)
after both doses of metoprolol. The volume of each Data are X ± SE.
portion was measured and 30 ml was frozen at F = Bioavailability.
*p < 0.05, tP < 0.01 compared with value during pregnancy.
- 20° C until analyzed.
The concentration of metoprolol in plasma was
determined by gas chromatography." Metoprolol and acid (H 104/83) is also determinable by this method,
the metabolites a-hydroxymetoprolol (H 119/66) and but it is not fully optimized and, therefore, we did not
O-demethylmetoprolol (H 105/22) were quantified in calculate the amount of this metabolite excreted in the
urine by gas chromatography-mass spectrometry (Fig. urine. Metoprolol plasma protein binding was deter-
1).3 The lowest measurable concentration at a standard mined at 37° C by an equilibrium dialyzing cell system
deviation of 10% is 1 nmollL for all three substances. with a cellulose membrane."
The acid metabolite (H 117/04) was analyzed in urine The disposition kinetics of intravenous metoprolol
according to a recently published gas chromatographic were evaluated by a three-compartment model with non-
method" based on a cyclization of the aminoalcohol side linear regression analysis. The program ELSFIT* was
chain with phosgene at alkaline pH. After completion used in the calculations. Systemic clearance (CI;J was
of the reaction, the oxazolidineone derivative is ex- determined by the ratio: Cl., = Dose.i/ AUC~.iV' in
tracted to an organic phase at acid pH. Before capillary which AUCoc,iv is the plasma AUC obtained by inte-
column gas chromatography with flame ionization de- grating the computer-derived triexponential equation.
tection, the remaining carboxylic group is trimethyl- The elimination rate constant of the terminal phase of
silylated. The minimum determinable concentration of
H 117/04 in urine was about 20 urnol/L. The hydroxy *Beal and Shiner, University of California at San Francisco.
 CLIN PHARMACOL THER
690 Hagstedt et al. JUNE 1985

3 hours after the oral dose during as well as after preg-

nancy. In most subjects the peak concentration occurred
within I to 2 hours of dosing (Fig. 2). Individual peak
concentrations were 1.8 to 8.4 times higher after than
during pregnancy, and the AUCs were 1.9 to 12.8 times
higher after than during pregnancy. In contrast, there
H 117/04
were no differences in the plasma tYzs during and after
pregnancy. As a result of the change in AUC, the ap-
p~rent metoprolol CIa fell after pregnancy to 8% to 53%
OH (X ± SE = 25.3% ± 7.8%) of the value during
I
O'CH 'CH'COOH pregnancy (Table II). Consistent with the results after

~:"OCH
-> 223
METOPROLOL H 104/83
"" OH CH3
9A
"'\. O'CH 'CH'CH 'NH'CH
2 2
r H ' CH ' OCH
2 3
OH
H 119/66
Fig. 1. Major metabolic pathways of metoprolol: a-hydroxy-
metoprolol (H 119/66), O-demethylmetoprolol (H 105/22),
an acid metabolite (H 117/04), and a hydroxy acid metabolite
(H 104/83).
the plasma concentration-time curve of the oral dose
(13) was calculated by linear regression analysis of the
log plasma concentration-time curve. The terminal tYz
was determined as: tYz = 0.693/13. The plasma AUC
of the oral dose (AUCor• l ) was calculated by the trap-
ezoidal rule. The residual area was calculated as C/I3,
where C is the plasma concentration at the last sampling
time. The apparent oral clearance (CIa) was calculated
as: CIa = Dose/ AUCor• l .
Results are presented as X ± SE. Statistical analyses
were with Student's paired t test.
RESULTS
No subject reported any adverse effects on any of
the four study occasions. After the intravenous dose,
the plasma tYzs were of the same order during and after
pregnancy. Metoprolol Cl., fell after delivery to 26%
to 97% (X ± .SE = 55.0% ± 12.6%) of the value
during pregnancy (calculated on the non-weight-related
clearance values). The apparent volume of distribution
(Vd) was higher during pregnancy, although not sig-
nificantly so (Table II).
The peak plasma concentration was reached 0.5 to
oral and intravenous dosing, oral bioavailability was
lower during (0.08 to 0.43) than after (0.30 to 0.56)
pregnancy, with a mean difference of 260% (Table 11).
Metoprolol plasma protein binding during and after
pregnancy did not differ, but the binding was only 3%
in subject 3 during pregnancy and II % after parturition
(Table II).
I
The total urinary recoveries of metoprolol and its
CH3
metabolites during the first 24 hours after the oral dose
were higher during pregnancy than after delivery (ex-
cept in subject 3 [Table III]). After the intravenous dose,
urinary metabolite concentrations were probably over-
estimated because of low urine metabolite concentra-
tions and cochromatographing peaks. Therefore, these
results are not reported.
DISCUSSION
We have reported' and now confirm that the me-
toprolol apparent CIa is more than four times higher
during than after pregnancy. Our study was designed
to explain the mechanism for this difference. Metopro-
101 is completely absorbed in man. 15 In all but one
subject the urinary recovery of the oral dose was higher
during pregnancy than after delivery. The reason for
this is not clear but might be a consequence of either
more extensive formation of the metabolites, mainly
the carboxylic acid, or more rapid excretion of the me-
tabolites by the kidneys during pregnancy. On the basis
of our findings, we believe that the absorption is, if
anything, higher during pregnancy than in the non-
pregnant state. These findings would thus rule out im-
paired absorption as the reason for the large difference
in apparent CIa and the reduction in systemic avail-
ability.
Several drugs, such as phenytoin and salicylates,
have lowered plasma protein binding during preg-
nancy.v" Changes in the plasma protein binding of
metoprolol cannot explain our findings, because we
found this to be very low. Therefore, alterations in this
parameter will not affect metoprolol kinetics. Further-
more, there was no clear difference in binding during
and after pregnancy.
VOLUME 37
NUMBER 6 Pregnancy-induced metoprolol metabolism 691

".t.""

-------

2 3 .. 5 • 7 • (t 10 11 12 hr 2 3 .. 5 e 7 8 , 10 11 12 h' 1 2 3 .. 5 IS 7 II , 10 11 12 hr

.................... Plasma concentration profile

--..... .......
-.
after metoprolol administration
< ,_ - - - - - - - - - -

_ o r a l dose pregnant
oral dose not pregnant
0- - - -0

_ i . v . dose pregnant
6 - - - - 6 i.v. dose not pregnant
2 3 .. 5 • 1 • • 10 11 12 h. 2 3 .. 5 e 7 8 , 10 11 12 hr

Fig. 2. Metoprolol plasma concentrations after single intravenous (10 mg) and oral (100 mg) doses
in five women with hypertension in the third trimester of pregnancy and after delivery.

Table III. Urinary excretion of metoprolol and its metabolites during and after pregnancy over the first 24 hours
after an oral dose of 100 mg (292 umol) metoprolol

Subject Metoprolol H 105122 H 119166 H 117104 Total* Recovery

No. (umol) (umol) ( pmol) (umol) (pmol) (% of dose)

During 2.35 2.37 26.4 186 217 74.3

After 7.68 1.04 26.9 144 180 61.6
2
During 3.35 0.59 27.0 173 204 69.9
After 16.3 0.47 12.4 106 135 46.2
3
During 11.3 0.67 28.8 99.2 140 47.9
After 14.2 0.57 22.1 174 211 72.3
4
During 1.58 0.48 26.1 138 166 56.8
After 4.76 0.28 18.9 127 151 51.7
5
During 12.9 0.90 18.6 180 212 72.6
After 24.1 0.80 17.7 137 180 61.6
Data are X ± SE.
"X ± SE values are 187.8 ± 14.9 and 171.4 ± 13.1 urnol during and after pregnancy. respectively. The difference is not significant.

692 Hiigstedt et at.
We have presented arguments against extracellular
fluid volume changes as a cause of our findings." Two
of our subjects had a 300% to 400% increase in CIa
during pregnancy. This change could be partly ac-
counted for by the increase in Yd, which was of a
similar magnitude, but it is not plausible that the Vd
explains all the increase in Cia, particularly since three
of the subjects had a large increase in Cia that far
exceeded their increase in Yd. On the basis of our
present results, we suggest that the greater clearance of
metoprolol in pregnancy results from increased hepatic
metabolism. Similar findings for other drugs, such as
phenytoin'? and carbamazepine, I support our conclu-
sion. At present, there is no teleologic explanation for
the increased metabolic capacity of the liver in preg-
nancy, but our results should lead to further studies of
this phenomenon. It is possible that the increased levels
of steroid hormones in pregnancy act as stimulants of
the hepatic monooxygenase system. These hormones
serve as substrates of the same enzyme system that is
responsible for the oxidation of drugs.
Although the effect of pregnancy on the 24-hour uri-
nary recovery of the active metabolites O-demethyl-
metoprolol and a-hydroxymetoprolol was small, the
increased metabolic capacity during pregnancy might
lead to higher metabolite blood concentrations. It seems
unlikely, however, that the concentrations of these two
metabolites, which are 5 to 10 times less active than
metoprolol itself, would reach high enough levels to
influence significantly the l3-blocking effect of meto-
prolo!. O-Dealkylated metoprolol is normally found at
very low concentrations over a short period, while a-
hydroxymetoprolol reaches about the same blood levels
as the parent drug in healthy young subjects.":" The
acid and the hydroxy acid metabolites are both devoid
of l3-adrenoceptor-blocking activity.
The recommended doses of metoprolol for hyperten-
sion do not differ for different patient categories, and
no special recommendations have been made for preg-
nant women. The high CIa in such patients will lead to
low plasma concentrations. It is therefore of great im-
portance to study the dose- and plasma concentration-
effect relationships. Such studies are now in progress.
We thank Mr. M. Ervik, Mrs. M. Gabrielsson, and Dr.
G. Gyllenhal for invaluable assistance in the analysis of
metoprolol and its metabolites.
References
1. Dam M, Christiansen J, Munck 0, Mygind KI: Anti-
epileptic drugs: Metabolism in pregnancy, Clin Phar-
macokinet 4:53-62, 1979.
CLIN PHARMACOL THER
JUNE 1985
2. Ervik M: Quantitative determination of metoprolol in
plasma and urine by gas chromatography. Acta Phar-
macol ToxicoI36:(suppl 5):136-144, 1975.
3. Ervik M, Hoffmann KJ, Kylberg-Hansson K: Selected
ion monitoring of metoprolol and two metabolites in
plasma and urine using deuterated internal standards.
Biomed Man Spectrom 8:322-326, 1981.
4. Gyllenhal 0, Hoffmann K-J: Simultaneous determination
of metoprolol and metabolites in urine by capillary col-
umn gas chromatography as oxazolidineone and trimeth-
ylsilyl derivatives. J Chromatogr 309:317-328, 1984.
5. Hamer C, Levy G: Serum protein binding of drugs and
bilirubin in newborn infants and their mothers. CLIN
PHARMACOL THER 28:58-63, 1980.
6. Hoffmann K-J, Regardh C-G, Aurell M, Ervik M, Jorde
L: The effect of impaired renal function on the plasma
concentrations and urinary excretion of metoprolol me-
tabolites. Clin Pharmacokinet 5: 181-191, 1980.
7. Hogstedt S. Lindberg BS, Rane A: Increased oral clear-
ance of metoprolol in pregnancy. Eur J Clin Pharmacol
24:217-220, 1983.
8. Hogstedt S, Lindeberg S, Axelsson 0, Lindmark G, Rane
A, Sandstrom B, Lindberg BS: A prospective controlled
trial of metoprolol-hydralazine treatment in hypertension
during pregnancy. Acta Obstet Gynecol Scand. (In
press.)
9. Lander CM, Edwards WE, Eadie MJ, Tyrer JH: Plasma
anticonvulsant concentrations during pregnancy. Neu-
rology 27:128-131, 1977.
10. Landon MJ, Kirkley M: Metabolism of diphenylhydan-
toin (phenytoin) during pregnancy. Br J Obstet Gynaecol
86:125-132, 1979.
11. Lindberg BS, Sandstrom B: How Swedish obstetricians
manage hypertension in pregnancy. Acta Obstet Gynecol
Scand 60:327-331, 1981.
12. Perucca E, Crema A: Plasma protein binding of drugs in
pregnancy. Clin Pharmacokinet 7:336-352, 1982.
13. Philipsson A: Pharmacokinetics of ampicillin during
pregnancy. J Infect Dis 136:370-376, 1977.
14. Piafsky KM, Borga 0: Plasma protein binding of basic
drugs. II. Importance of cxI-acid glycoprotein for indi-
vidual variation. CLIN PHARMACOL THER 22:545-549,
1977.
15. Regardh CG, Borg KO, Johansson R, Johnsson G, Pal-
mer L: Pharmacokinetic studies on the selective [3,-re-
ceptor antagonist metoprolol in man. J Pharmacokinet
Biopharm 2:347-364, 1974.
16. Regardh CG, Johnsson G: Clinical pharmacokinetics of
metoprolol. Clin Pharmacokinet 5:557-569, 1980.
17. Regardh CG, Landahl S, Lundborg P, Steen B, Hoffmann
K-J, Lagerstrom P-O: Pharmacokinetics of metoprolol
and its metabolite cx-OH-metoprolol in healthy, non-
smoking, elderly individuals. Eur J Clin Pharmacol
24:221-226, 1983.