Placental Transfer and Maternal Midazolam Kinetics

Placental transfer and maternal midazolam kinetics Midazolam was given in a single 15-mg oral dose as a sedative the evening before elective cesarean section. Twelve hours later, levels of this new benzodiazepine were measureable in the fetomaternal entity in only one of 13 cases. After 15 mg midazolam orally or 0.05 mg /kg midazolam intramuscularly 15 to 60 min before elective cesarean section, there was evident transfer of drug into the placenta, but transfer took place more slowly than with diazepam. On the basis of kinetics derivedfrom maternal serum concentrations after oral, intramuscular, or intravenous dosing, midazolam appears to have a rapid onset and short duration of action, which was also evident from subjective assessments by the patients. There was wide

interindividual variation in the gastrointestinal absorption of midazolam in full-term pregnant women. Clinically, midazolam nevertheless seemed to be very useful for nocturnal sedation before elective cesarean section; it ensures a mean duration of sleep of about 6 hr and there are virtually no detectable levels of drug in the fetomaternal entity the next morning.

J. Kanto, M.D., S. Sjtivall, M.D., R. Erkkola, M.D., J-J Himberg, M.D., Ph.D., and L. Kangas, M.Sc. Helsinki and Turku, Finland Departments of Anesthesiology and Obstetrics and Gynecology, Turku University Central Hospital, and FRC Blood Transfusion Service, Helsinki and Turku Research Centre, Farmos Group Ltd.

Benzodiazepine kinetics during pregnancy, labor, and lactation have recently been extensively reviewed.' There are distinct and clinically significant differences in the degrees of transfer of the various derivatives into the placenta and in the maternal kinetics of the various derivatives. We studied these properties of the benzodiazepine midazolam. Midazolam (see structure), the first watersoluble benzodiazepine, has a number of unique properties. After oral" or parentera12-12 dosing, it is absorbed quickly and, in contrast to

Received for publication Dec. 7, 1982. Accepted for publication Dec. 17, 1982. Reprint requests to: Jussi Kanto, M.D., Department of Anesthesiology, Turku University Central Hospital, Kiinamyllynkatu 4-8, SF-20520 Turku 52, Finland.

786

older derivatives, it is very readily excreted, with a t1/2 of only about 2 hr.5." After a single dose, correlation between plasma levels and clinical response is reasonably good.2.6.'2 This new drug therefore seems attractive for use as a sedative-anxiolytic during childbirth. Methods All the subjects were in class I of the classification published by the American Society of Anesthesiologists, and their pregnancies at 38 to 41 wk of gestation seemed uncomplicated. Fetopelvic disproportion was the most common reason for elective cesarean section. Each patient was placed on an operating table tilted 20 to 30 degrees to the left, and 100% oxygen was administered for about 5 min. No premedication other than midazolam was used.

Volume 33 Number 6

Anesthesia was induced with thiopental (4 to 5 mg/kg IV). Endotracheal intubation was facilitated with suxamethonium (1 to 1.3 mg/kg IV) followed by pancuronium (0.1 mg/kg IV). The lungs were ventilated with a 60% mixture of nitrous oxide in oxygen and halothane 0.5% to 1% until delivery. After clamping of the umbilical cord, droperidol (0.03 to 0.06 mg/kg IV) and fentanyl (1.4 to 2.8 iug/kg IV) were injected and anesthesia was maintained with a mixture of nitrous oxide and oxygen. During the operation, each patient received acetated Ringer's lactate solution in 5% glucose as an intravenous infusionabout 10 ml/kg during the operation (43 to 62 min) and thereafter about 5 ml/kg, up to 1000 to 1500 ml. The mothers left the recovery room after 1 to 2 hr and received 1000 ml of a Normosol type of infusion in 5% glucose solution on the ward. Group 1 consisted of 13 patients who were 20 to 40 yr old (mean 31), weighed 54 to 85 kg (mean 67), and were 154 to 177 cm tall (mean 162). Each group 1 patient was given 15 mg oral midazolam at between 9 and 10 P.M. the evening before the elective cesarean section. Maternal and umbilical venous serum concentrations of midazolam together with concentrations of midazolam in the umbilical artery and amniotic fluid were determined the next morning, between 10.5 and 12.4 hr (mean 11.4) after the single dose of midazolam the previous evening. At the same time, patients were asked to give their subjective assessment of the value of midazolam as a sleep inducer (very good, good, moderately good, or poor), and the time at which they fell asleep and the duration of sleep were recorded. Group 2 consisted of 11 patients who were 18 to 40 yr old (mean 28), weighed 45 to 98 kg (mean 69), and were 144 to 172 cm tall (mean 165). Each patient received 15 mg oral midazolam as sedative anxiolytic premedication 15 to 60 min (mean 34.3) before elective cesarean section (Table I). In eight patients serum concentrations of midazolam were determined for up to 6 hr after drug (Fig. 1). Group 3 consisted of six patients who were 23 to 36 yr old (mean 28), weighed 54 to 76 kg (mean 67) and were 155 to 168 cm tall (mean 160). Each patient received 0.05 mg/kg IM

Maternal midazolam kinetics

787

HCCOOH HCCOOH

Cl

Midazolam

midazolam as sedative anxiolytic premedication 18 to 45 min (mean 30.5) before elective cesarean section (Table II). In five cases, maternal serum levels of midazolam were determined for up to 6 hr after drug (Fig. 1). Patients in groups 2 and 3 were asked to give their subjective assessment of the value of midazolam as a premedicant in a simple questionnaire covering: sedation (marked, moderate, slight, absent); apprehension (marked, moderate, slight, absent); excitement (marked, moderate, slight, absent); dizziness (marked, moderate, slight, absent); emetic effect (none, slight nausea, severe nausea, vomiting); and headache (none, slight, moderate, severe). Group 4 consisted of eight patients who were 20 to 32 yr old (mean 29), weighed 56 to 98 kg (mean 72), and were 150 to 167 cm tall (mean 161). Each patient was given 0.075 mg/kg IV midazolam to deepen anesthesia after delivery of the child. Serum levels of midazolam were determined for up to 6 hr (Fig. 1), and midazolam kinetics were calculated assuming a twocompartment open mode1.5 Data were evaluated by the STRIPACT9 and nonlinear least squares methods.4 A Radio Shack TRS-80 microcomputer was used for calculations. Concentrations of midazolam and its main metabolite were determined in maternal venous serum, in umbilical venous and umbilical arterial serum, and in amniotic fluid by a gas chromatographic method.6 The results were subjected to Student's t test. Results

Levels of midazolam were measurable in the morning, 11.4 hr after dosing, in only one of the 13 patients in group 1 (maternal venous serum 12 ng/ml, umbilical venous serum and

788

OM. Pharmacol. Ther. June 1983

Kanto et al.

250 C200 3 150

0075 mg/kg tv.,n, 8

0 10

E

0,050mg/kg tm.n=5 15mg p.o.,n= 8

25-

:3

Q5

2

1

5hours

4

3

6

Fig. 1. Maternal venous serum concentrations of midazolam after a single 15-mg oral dose or a 0.05-mg/kg IM dose given as premedication in elective cesarean section. Serum levels were also determined after a single dose of 0.075 mg/kg IV midazolam during cesarean section given to deepen anesthesia once the baby was delivered.

Table I. Maternal venous (MV) serum, umbilical venous (UV) serum, umbilical arterial (UA) serum, and amniotic fluid midazolam concentrations after a single 15-mg oral dose given as premedication in elective cesarean section Time

MV

(min)

(ng 1ml)

UV (ng 1ml)

UV/MV

ratio

UA (ng 1ml)

UA IMV

ratio

15

20 20 20 27 30 40 45 50 50 60 Mean SD

0.86 0.69 0.67

5

0.71

2

0.15

7

6

13

9

6 0

4 0

12

0 29

0.63

27

7 19

0.64 0.70

21 5 5

18

18

1.00

13

46 11

12.7 13.8

8.4 9.7

NM

Amniotic fluid (ng 1ml) 0 0 0 0

0 0

0

0.74 0.14

5.7 7.1

0.46 0.46 0.19 0.72 0.45 0.25

0

0 1.5 0

NM = not measured.

umbilical arterial serum 7 ng/ml). No midazolam was detectable in the amniotic fluid in this instance. Group 1 patients assessed the quality of sleep subjectively as follows: very good, two patients; good, six patients; moderately good, five patients. No patient assessed the quality of sleep as poor. In general patients fell asleep about 30 min after taking midazolam. The duration of sleep was 3 to 8 hr (mean 5.7 hr).

Maternal, umbilical, and amniotic fluid concentrations of midazolam in group 2 patients are listed in Table I. Differences between maternal venous concentrations and umbilical venous and arterial concentrations were not significant. Maternal serum levels in group 2 patients are shown in Fig. 1. The results obtained in group 3 patients are shown in Table II and Fig. 1. Both the umbilical

Maternal midazolam kinetics 789

Volume 33 Number 6

Table II. Maternal venous (MV) serum, umbilical venous (UV) serum, umbilical arterial (UA) serum, and amniotic fluid midazolam concentrations after a single dose of 0.05 mg /kg IM given as premedication in elective cesarean section Time (mm)

MV (ng I ml)

18

34

25 31 32 32 45

41

Mean SD

34 55 48 28

40.0 10.1

UV (ng Iml)

28 22 18

21 24 17

21.7 4.0

UV/MV

UA

UA IMV

ratio

(ng Iml)

ratio

7

0.21

0.82 0.54 0.53 0.38 0.50 0.61 0.56 0.15

Amniotic fluid (ng Iml) 2

NM

NM 7 18 18

0.21 0.33 0.38

0

14

0.50 0.32 0.12

0

12.8 5.5

0 2

NM = not measured.

venous and the umbilical arterial serum concentrations were lower (P < 0.01; P < 0.001) than maternal venous concentrations. The umbilical vein concentrations were significantly higher than the umbilical arterial levels (P < 0.01). Absorption of midazolam after intramuscular dosing (group 3 patients, dose 0.05 mg/kg) was clearly superior to and less variable than absorption after oral doses (group 2 patients, dose 15 mg = 0.21 mg/kg in the case of patient weighing 70 kg). The variability in absorption after oral doses is also evident from the maximum concentration, time to the maximum concentration, and AUC values (Table IIIA), but the differences were not significant. In general, midazolam, whether given orally or intramuscularly, proved to be a useful premedicant for the conduct of elective cesarean section (group 2 and 3 patients). Although it induced only slight-to-moderate sedation, only three patients in group 2 experienced moderate apprehension or excitement. Seven other patients in group 2 and three patients in group 3 experienced no apprehension or excitement. One patient in group 2 and three patients in group 3 experienced only slight apprehension or excitement. No patient reported dizziness, nausea, vomiting, or headache. The Apgar scores of the newborn infants at 1 and 5 min in all four groups were in the range generally recorded in our department, and no score was lower than seven.

Maternal venous serum levels of midazolam and the kinetic parameters derived from these concentrations are listed in Table IIIB. in group 4 patients are shown in Fig.

1

Discussion On the basis of the dynamic and the kinetic data gathered in this study, midazolam would seem to have a rapid onset of action and a short duration of action. These properties are of great importance in relation to elective cesarean section in that the drug was virtually undetectable in the fetomaternal entity the morning after a single dose the previous evening, Metabolites of midazolam were also undetectable, since their t1/2s are about the same or shorter than that of the parent drug. Their pharmacologic activity is, in any case, negligible.5. 7 In contrast to "older" benzodiazepines,7 midazolam can therefore be recommended for routine oral sedative-anxiolytic medication the evening before elective cesarean section. All the older benzodiazepine derivatives result in very appreciable maternal and fetal levels, and may affect the child after use the evening before cesarean section.' A recent study in insomniac patients has shown midazolam to be an effective, short-acting benzodiazepine that does not induce hangover.10 The degree of transfer of midazolam to the placenta has been found to be clearly lower than that of diazepam (for references, see Kanto'),

790

Kanto et al.

CIO?.

Table IIIA. Midazolam maximum concentration (C max), time to C max after the intramuscular and oral doses (mean ± SD) C max (nglml)

Dose

mg/kg IM (n = 5) 15 mg orally (n = 8) 0.05

28.2

(range

-±

37.9

±

max),

and AUC AUC

'max

(nglml

(min)

7.1

20.4-35.9)

(range

(1

27.1

it 7.8 16.5-36.0)

(range

21.1

27.2

-±

(range

9.1-74.0)

Pharmacol. Ther. June 1983

12.0

min)

3172

±

1426

2229

±

1028

9-45.1)

Table IIIB. Midazolam kinetics derived from serum concentrations after a single dose of 0.075 mg /kg IV, given in conjunction with elective cesarean section

Mean

SD

Pafient

t1/20,

t1/2/3

Vdo

Cltot

(mm)

(mm)

V, (1/kg)

Vd

No.

(1/kg)

(1/kg)

(mIlminIkg)

1

12.5

126.6

0.31

0.80

2.08

6.84

10972

2

11.6

95.8

0.98

2.88

11.59

3

6.9

91.6

0.44 0.34

1.20

3.81

15.67

4

15.1

136.5

0.40

1.01

6.0

6 7

11.2

103.9 119.1

0.21 0.36

0.86 0.85

5.69 2.02

11.23

5

5.9

73.5

1.11

2.29

8

11.9

81.2

0.40 0.67

8.66 6.83 14.78

6474 4785 6679 8674

1.24

3.28

16.50

10.1

103.5

1.01

2.97

3.4

22.3

0.39 0.14

0.17

1.31

11.50 3.90

1.70

AUC (nglml min)

10975 5073

4544 7272 2641

t1/20 = distribution t1/2; t1/2/3 = elimination t1/2; V, = volume of the central compartment; Vdss = volume of distribution at steady state; Vds = volume of distribution (hiring elimination phase; C1101 = total serum clearance.

and transfer appears to take place even more slowly than for oxazepam and lorazepam, both of which have been recommended as alternatives to diazepam because they transfer slowly to the placenta and their metabolism is simple (glucuronidation).7 Dynamic and kinetic studies in pregnant ewes support the use of midazolam rather than diazepam.' However, so far the ability of the newborn child to metabolize midazolam is not known, and its place as a premedicant in obstetrics remains to be established. In the small groups of patients studied by us no deleterious effects on the newborn children were found. Midazolam kinetics in pregnant women undergoing cesarean section were generally similar to those in healthy subjects.3'11.13 The total plasma clearance value for midazolam in healthy men has been shown to be higher in patients in the supine position (616 -± 157[ SD] ml/min) than in ambulant patients (317 ± 110 ml/min)

Although no direct comparison with our female patients is possible, mean total clearance in our patients in the supine position was fairly high (11.5 ml/min/kg, or 805 ml/min in a woman weighing 70 kg). Adoption of the supine position has been shown to increase hepatic blood flow by up to 40% to 60%. Consequently, drugs such as midazolam, which have a high total clearance value, should be associated with variations in hepatic elimination as position changes .8 The greater variation in absorption of midazolam after oral than after intramuscular dosing in our patients may have depended primarily on changes in gastrointestinal motility during pregnancy. The maximum concentration in our patients after oral doses varied between 9.1 and 74.0 ng/min (intramuscular, 20.4 to 35.9 ng/ ml), and time to maximum concentration varied between 9 and 45.1 min (intramuscularly, 16.5 to 36 min). To ensure a predictable response in

Volume 33 Number 6

premedication before elective cesarean section, midazolam should, therefore, on the basis of its kinetics, be injected intramuscularly, at the low dose of 0.05 mg/kg. Further dynamic studies are, however, needed to establish the place of midazolam as a premedicant in elective cesarean section. We conclude that our kinetic data support the use of midazolam in obstetrics, but that more clinical experience is needed to assess whether it has advantages over other benzodiazepines.

References Conklin KA, Graham CW, Murad S, Randall FM, Katz RL, Cabalum T, Lieb SM, Brinkman CR: Midazolam and diazepam: maternal and fetal effects in the pregnant ewe. Obstet Gynecol 56:471-474, 1980. Crevoisier C, Eckert M, Heizmann P. Thurneysen DJ, Ziegler WH: Relation entre l'effect clinique et la pharmacocinetique du midazolam apres administration iv. et i.m. Arzneimittelforsch 31:2211-2215, 1981. Daneshmend TK, Jackson L, Roberts CJC: Physiological and pharmacological variability in estimated hepatic blood flow in man. Br J Clin Pharmacol 11:491-496, 1981. Digoxor WJ, Brown MB, editors: BMDP-77 Biomedical Computer Programs, P-Series. Berkeley, 1977, University of California Press. Greenblatt DJ, Locniskar A, Ochs HR, Lauwen

Maternal midazolam kinetics

791

PM: Automated gas chromatography for studies of midazolam pharmacokinetics. Anesthesiology 55:176-179, 1981. Himberg J-J, Mattila MAK: Midazolam, a new short half-life benzodiazepine: Gas-chromatographic assay and pharmacokinetics after in-

tramuscular administration. Finnish Pharmacological Society Meeting, Helsinki, March 12th, 1982. (Abstr.) Kanto JH: Use of benzodiazepines during pregnancy, labour and lactation, with particular reference to pharmacokinetic considerations. Drugs 23:354-380, 1982. Klotz U, Ziegler G: Physiologic and temporal variation in hepatic elimination of midazolam. CLIN PHARMACOL THER 32:107-112, 1982. Leferink JG, Maes RAA: STRIPACT, an interactive curve fit programme for pharmacokinetic analyses. Arzneimittelforsch 29:1894-1898, 1979. Monti JM, Debellis J, Gratadoux E, Alterwair P, Altier H, D'Angelo L: Sleep laboratory study of the effects of midazolam in insomnic patients. Eur J Clin Pharmacol 21:479-484, 1982. Smith MT, Eadie MJ, O'Rourke Brophy T: The pharmacokinetics of midazolam in man. Eur J Clin Pharmacol 19:271-278, 1981. Ziegler WH, Thurneysen JD, Crevoisier C, Eckert M, Amrein R, Dubuis R: Relations entre l'effect clinique et la pharmacocinetique du midazolam apres administration i.m. et i.v. chez des volontaires. Arzneimittelforsch 31:22062210, 1981.