Clinical Pharmacokinetics 4: 297-309 (1979) 0312-5963/79/0500-0297/$03.25/0 © ADIS Press Australasia Pty Ltd. All rights reserved ..
Pharmacokinetics of Antibiotics in Pregnancy and labour Agneta Philipson Department of Infectious Diseases, Danderyd Hospital, S-182 03 Danderyd
Summary
Few of the articles published on antibiotics and pregnancy are concerned with pharmacokinetics. It is particularly difficult to evaluate possible alterations in pharmacokinetic parameters that may be due to pregnancy. Most data available have been obtained in connection with abortion or delivery. Such data may not be representative for pregnancy as such. Marked changes in most organ systems, particularly in renal function, but in composition and amounts of body fluids as well, make it likely that several pharmacokinetic parameters change, possibly gradually as pregnancy progresses. Accumulated data for several ~-lactam antibiotics, and also for aminoglycosides indicate that antibiotics eliminated mainly by renal excretion will produce lower levels in serum or plasma in pregnant women than in other individuals. A Iso, the half-life of certain antibiotics in serum is shorter during pregnancy. Transplacental passage occurs for all antibiotics according to the physicochemical properties of the drug. Bolus injections to a pregnant woman are more efficient than continuous infusion in producing high levels of antibiotic in fetal serum and amniotic fluid. Fetal tissue levels are higher following multiple doses than after a single dose. Lower serum levels of antibiotics in pregnant women than in other individuals following the same dosage will be unsatisfactory as micro-organisms are less likely to be affected.
Infections which require antibiotic treatment are far from uncommon during pregnancy. When antibiotics are prescribed for pregnant women the dosage used by most physicians is the same as that prescribed for men and non-pregnant women. If any change is considered, it is usually towards a lower dosage out of fear for adverse effects. However, it is unlikely that dosage requirements are the same for pregnant and non-pregnant women. The physiological alterations
within most organ systems are vast during pregnancy (Hytten and Leitch, 1971). Therefore pharmacokinetics of many antibiotics may change during pregnancy, the reasons being the same as for other drugs (Krauer and Krauer, 1977; Eadie et aI., 1977). Altered pharmacokinetics may call for a change in choice of antibiotic, in dosage, or in route of administration, in order to obtain optimal effect of antiinfectious treatment. If the fetus is affected, as in the
Pharmacokinetics of Antibiotics in Pregnancy and Labour
case of syphilis, it is mandatory that sufficient levels of antibiotic are reached and maintained in fetal tissues, to effectively cure the disease. In that case the need for knowledge of transplacental passage and fetal levels of the antibiotic to be used, is obvious. If, on the other hand, there is a threat of a rising infection which will reach the amniotic fluid, it is essential that the antibiotic administered to the pregnant woman will reach the amniotic fluid in adequate levels. There is an obvious need for knowledge about the behaviour of antibiotics in pregnancy. In spite of this, little has been published on this subject compared with the knowledge of pharmacokinetics of antibiotics in man generally. The most extensive encyclopaedia on antibiotics, the Antibiotika-Fibel by Walter and Heilmeyer (J 975), provides very little help to the physician who wants to treat an infection in a pregnant patient. The lack of pharmacokinetic data in pregnancy is of course due to the technical as well as the ethical problems related to such studies.
I. Problems Associated with Pharmacokinetic Studies in Pregnancy 1.1 Technical Problems Knowledge of the pharmacokinetics of a drug is derived from data from several individuals who should either be healthy and in the same condition, or suffer from the same disorder, the possible influence of which on one or more of the pharmacokinetic parameters must be known or possible to evaluate. Meaningful information can only be obtained for antibiotics which may be accurately assayed and are not metabolised or have metabolites which can be assayed. For studies of antibiotic levels in maternal plasma or serum, standards should preferably be prepared in plasma or serum from untreated pregnant women. Because of differences in protein content, a minimum requirement is to check that a standard prepared in such serum does not perform differently
298
in the assay from the actual standard to be used. Studies involving determination of fetal plasma levels of an antibiotic appropriately call for standards prepared in fetal plasma, as drugs may bind differently in fetal and adult plasma (Ehrnebo et a\., 1971; Pruitt and Dayton, 1971). As plasma or serum levels of antibiotic are determined as total levels of drug, a lower degree of protein binding in the fetus will give false high values for fetal levels. The technical problems involved in fetal tissue level monitoring are immense; freeing the tissue from blood, extracting the antibiotic, and selection of proper standards being examples of such problems. For adequate comparison of the pharmacokinetics ofa drug in pregnant and in non-pregnant individuals a control group is essential as variations in antibiotic serum levels between pregnant women may be marked. Controls must be matched for age, weight, sex, and if possible for medical condition as well. Samples to be assayed must be treated in an identical way as certain antibiotics will deteriorate within weeks even when stored at - 20° (Dornbusch, personal communication).
1.2 Ethical Problems Over the last 10 years ethical aspects have increasingly been brought into focus. It is regarded as unethical to give to a pregnant woman, who wants to proceed with her pregnancy any antibiotic solely for study purposes. Medication has to be medically justified. Certain accurate methods for determination of excretory capacity, metabolism, or volumes of distribution, are carried out with radioactive substances and obviously may not be employed in pregnant women. However, there is another ethical aspect of drug studies during pregnancy. Is it ethical not to perform such studies? Is it ethical to give a pregnant woman any drug for which the pharmacokinetics during pregnancy are not known? Many drugs which may be beneficial in treating pregnant women are not considered because of lack of knowledge of their activity
299
Pharmacokinetics of Antibiotics in Pregnancy and Labour
during pregnancy, thus depriving the pregnant woman of the normal choice of therapy. Because the pharmacodynamics of a drug most likely will have changed, the pregnant patient may be either undertreated or overtreated if a dosage is used that has been calculated from studies performed on men or non-pregnant women.
2. Pharmacokinetics in Pregnancy
Several reasons for alterations in drug kinetics in pregnancy have been discussed (Krauer and Krauer, 1977). Dose requirements of certain drugs, particularly antiepileptic drugs, are known to change when the patient becomes pregnant and as pregnancy proceeds (Eadie et ai., 1977). Even for antibiotics, pharmacokinetics will most likely change as pregnancy advances and changes may successively become more pronounced. The gradual increase in renal function will probably increase the elimination rate of antibiotics which are renally excreted. Altered drug metabolism (Feuer and Kardish, 1975; Gut et ai., 1976) may well influence pharmacokinetics of those antibiotics which are metabolised. Serum levels may also be altered in pregnant patients, the extent of change depending on the distribution properties of each antibiotic. Serum levels of antibiotics with a low volume of distribution, may be affected differently from those of antibiotics with a large volume of distribution. Little specific data are available on pharmacokinetics of antibiotics during pregnancy. Most studies include investigation of the transplacental passage and have been carried out at the time of abortion, labour and delivery, and caesarean section. When evaluating the results derived from such studies it must be borne in mind that the data may not be representative for any stage of pregnancy except precisely abortion, labour or caesarean section. Abortion (whether surgical or not), labour and caesarean section will most likely affect renal excretion, blood flow to various organs, and other parameters by which drug pharmacokinetics is regulated.
The stress involved for the patient may be an important factor in itself. Gastrointestinal absorption of antibiotics administered in such situations may be impaired. The influence of other pharmaceutical compounds, administered during labour or surgery, on the pharmacokinetics of an antibiotic is difficult to evaluate. Antibiotic levels in amniotic fluid and fetal serum, plasma, or tissues following administration of the drug to the mother during labour must be evaluated with precaution. Pronounced changes in the haemodynamics due to contractions of the uterus will most likely impair placental transfer of the drug.
3. Present Knowledge of Antibiotic Kinetics in Pregnancy
In spite of the problems involved investigations of antibiotics in pregnancy have been carried out, according to various protocols. Maternal levels of antibiotic in plasma or serum have been studied following single or repeated doses, mostly prior to abortion, delivery, or caesarean section. In some studies serial serum levels have been studied in the same patients whereas in other studies only the serum level at the time of delivery has been reported. In the latter type of study, the area under the serum concentration versus time curves have been constructed from single values for a great number of patients through variations of the dose-delivery interval. Also, the presentation of data is diverse, thus making comparisons between results obtained in different studies difficult. Evaluation of whether or not serum levels, renal excretion, or other parameters fora particular antibiotic alter during pregnancy is often impossible. However, several antibiotics have been studied and some preliminary information has accumulated. 3.1
~-Lactam
Antibiotics
3.1.1 Isoxazo[ylPenicillins Serum levels of methicillin following a single intravenous dose have been studied in women 32 to 40
Pharmacokinetics of Antibiotics .in Pregnancy and Labour
weeks pregnant or in labour (Depp et aI., 1970; MacAulay et aI., 1973). Similar studies have been published for dicloxacillin where a dose of 500mg was administered to women in labour (MacAulay et aI., 1968; Depp et aI., 1970). According to MacAulay et aI., the biological half-life of the elimination phase (tI/2~) for dicloxacillin was 2h. The maternal serum levels reported in these studies of dicloxacillin are quite different. It is not possible to determine on the basis of these studies if pharmacokinetics for methicillin and dicloxacillin are altered in pregnancy, but serial serum levels of methicillin administered to 3 women in active labour were reported to be generally lower than those of 1 man and I non-pregnant woman (MacAulay et aI., 1973). Pharmacokinetic parameters for oxacillin have been studied in 12 women 38 to 41 weeks pregnant at the time of amniocentesis (Bastert et aI., 1975). Following a single intravenous dose of 2g the calculated serum level at time zero (Co) was 79)lg/ml. The elimination constant was 1.044h- ' , tl/2~ was 40 min, the apparent volume of distribution (Vd) 25.3L, and the total clearance of drug was 441 mIl min. The authors point out that in this study the values for V d and clearance are higher than those in other studies performed in non-pregnant individuals. Calculations were based on observations during the first 2h following dose administration. 3.1.2 Ampicillin Ampicillin has been the most widely studied antibiotic in pregnant women, most often with regard to its transplacental passage. In comparing studies where the same dose has been used it is obvious that mean serum levels differ from one report to another (table The actual figures for serum levels are not always given but have then been deduced from graphs. The elimination half-life calculated from the graph in one of these studies is 46 minutes (Perry and Leblanc, 1967). It has also been reported to be 30 minutes at caesarean section (Boreus, 197 I), 40 minutes at delivery (Bastert et aI., 1973), and 40 minutes in the 8th month of gestation as opposed to
n.
300
Table I. Mean values reported in various publications for levels of ampicillin in serum and plasma (fl9/ml) in pregnant women at 1, 2 and 3h following a single intravenous dose of 500mg. Figurtls within brackets represent the nurnber of patients for which the mean was calculated. Standard deviations (SD) are given when available Time after dosage 1h 6.3(4)
SD 2.2
7.9(13) 15.0(8) 6.2(26)
2.14
Reference
2h
SD
3h
1.0(2)
0.1
SD
0.6(1)
Bray et al. (1966)
4.5(13)
3.4(13)
MacAulay et al. (1966)
6.0(4)
2.8 ( 3)
Perry and Leblanc (1967)
1.9(26)
0.9
0.7(26)
0.4
Philipson (1977)
58 minutes in women in labour (von Voigt et aI., 1978). The discrepancies may in part be due to a difference in the study groups, but also to different protocols and methods for calculations of t I / 2~' V d for ampicillin during labour or surgery in late pregnancy was found to be 0.23L/kg (Bastert et aI., 1973). Alterations in pharmacokinetics of ampicillin in pregnancy have been demonstrated in a crossover study in 26 women, 9 to 36 weeks pregnant, with lower urinary tract infections (Philipson, 1977). Ampicillin was administered intravenously as well as orally. Each woman served as her own non-pregnant control receiving identical doses after pregnancy. Plasma levels were found to be 50 % lower in pregnant women than in the same women when not pregnant. Some other relevant data are shown in table II. Further studies showed that subsequent doses of ampicillin during the course of oral therapy give similar levels in plasma and urine as the initial oral dose and that no accumulation of drug occurs (Philipson, 1978). By doubling the dose of ampicillin given to pregnant women their plasma levels were increased to 'normal'. Pivampicillin has been found to produce plasma levels during pregnancy of the same magnitude as those produced by oral ampicillin with twice the equimolar dose (Philipson, 1978). Following the onset of labour, however, blood levels could not be
Pharmacokinetics of Antibiotics in Pregnancy and Labour
301
Table II. Mean values and SD for apparent volume of distribution (Vd), elimination half-life (t 1/2~), bioavailability, renal clearance (Cl r), and recovery in urine of ampicillin in 26 women during and after pregnancy, following identical doses
1968; Paterson et al., 1970) produced serum levels that are considerably lower than one would expect (Walter and Heilmeyer, 1975). Studies of the pharmacokinetics of cephalothin in women at term or at the time of delivery have been carried out by MacAulay and Charles (I 968) and by Bastert et al. (I 974). In the first study an intravenous dose of Ig was administered to 54 women. Maternal serum was sampled at delivery which took place 9 to 810 minutes later. A serum concentration-time curve was constructed from the 54 individual observations. The elimination constant was found to be 86 x 10- 4 which gives a t 1/ 2~ of I.3h. This value is different from the value of about 40 minutes which is found in non-pregnant populations (Walter and Heilmeyer, 1975). Co was 30.2pg/rnl. Calculations of clearance from plasma give a value of 285ml/min. In the study by Bastert et al. (I 974) an intravenous dose of 2g was given to 25 women 38 to 41 weeks pregnant at the time of amniotomy. A regression line for serum levels was based only on the first 2h of observation. Co was 63.8pg/ml, tl/2~ 31 minutes, Vd 0.3 1L/ kg, and total clearance from serum 687ml/min. Even in these studies mean serum levels of cephalothin in pregnant women appear lower than those which would be expected (Walter and Heilmeyer, 1975). Cephazolin was given intramuscularly in a dose of 14mg/kg to 40 women 8 to 20 weeks pregnant at the time of interruption of pregnancy. Serum levels were measured and found to be 23 to 55 % lower than those which had earlier been reported for non-pregnant adults following a similar dose (Bernard et aI., 1977b). In this study t1 /2~ was l.5h which can be compared to a t I / 2~ of 2h found in non-pregnant individuals (Walter and Heilmeyer, 1975). In a study of cefatrizine, a single oral dose of I g was administered prior to abortion and was found to produce a wide range of serum levels at I h in women 8 to 20 weeks pregnant. Less variation was found at 2, 4 and 8h (Bernard et aI., I 977d). The t1 /2~ was 2Ah. The authors state that this is Ih longer than in healthy adult males but similar to that found in infected hospitalised patients. The authors believe that the
tl/2~ Vd (L/kg) (min)
Bioavailability
Cl r (ml/min)
Rec'overy in urine (% of IV dose)
(%)
During pregnancy SD
0.55
39
45.6
330
79.4
0.28
8.1
20.2
(137-535)0
24.7
After pregnancy SD
0.41
44
48.1
238
79.5
0.20
6.8
19.3
(85-425)a
30.6
Pvalue
a
< 0.05 < 0.05
Not significant
Pharmacokinetics of Antibiotics in Pregnancy and labour Agneta Philipson Department of Infectious Diseases, Danderyd Hospital, S-182 03 Danderyd
Summary
Few of the articles published on antibiotics and pregnancy are concerned with pharmacokinetics. It is particularly difficult to evaluate possible alterations in pharmacokinetic parameters that may be due to pregnancy. Most data available have been obtained in connection with abortion or delivery. Such data may not be representative for pregnancy as such. Marked changes in most organ systems, particularly in renal function, but in composition and amounts of body fluids as well, make it likely that several pharmacokinetic parameters change, possibly gradually as pregnancy progresses. Accumulated data for several ~-lactam antibiotics, and also for aminoglycosides indicate that antibiotics eliminated mainly by renal excretion will produce lower levels in serum or plasma in pregnant women than in other individuals. A Iso, the half-life of certain antibiotics in serum is shorter during pregnancy. Transplacental passage occurs for all antibiotics according to the physicochemical properties of the drug. Bolus injections to a pregnant woman are more efficient than continuous infusion in producing high levels of antibiotic in fetal serum and amniotic fluid. Fetal tissue levels are higher following multiple doses than after a single dose. Lower serum levels of antibiotics in pregnant women than in other individuals following the same dosage will be unsatisfactory as micro-organisms are less likely to be affected.
Infections which require antibiotic treatment are far from uncommon during pregnancy. When antibiotics are prescribed for pregnant women the dosage used by most physicians is the same as that prescribed for men and non-pregnant women. If any change is considered, it is usually towards a lower dosage out of fear for adverse effects. However, it is unlikely that dosage requirements are the same for pregnant and non-pregnant women. The physiological alterations
within most organ systems are vast during pregnancy (Hytten and Leitch, 1971). Therefore pharmacokinetics of many antibiotics may change during pregnancy, the reasons being the same as for other drugs (Krauer and Krauer, 1977; Eadie et aI., 1977). Altered pharmacokinetics may call for a change in choice of antibiotic, in dosage, or in route of administration, in order to obtain optimal effect of antiinfectious treatment. If the fetus is affected, as in the
Pharmacokinetics of Antibiotics in Pregnancy and Labour
case of syphilis, it is mandatory that sufficient levels of antibiotic are reached and maintained in fetal tissues, to effectively cure the disease. In that case the need for knowledge of transplacental passage and fetal levels of the antibiotic to be used, is obvious. If, on the other hand, there is a threat of a rising infection which will reach the amniotic fluid, it is essential that the antibiotic administered to the pregnant woman will reach the amniotic fluid in adequate levels. There is an obvious need for knowledge about the behaviour of antibiotics in pregnancy. In spite of this, little has been published on this subject compared with the knowledge of pharmacokinetics of antibiotics in man generally. The most extensive encyclopaedia on antibiotics, the Antibiotika-Fibel by Walter and Heilmeyer (J 975), provides very little help to the physician who wants to treat an infection in a pregnant patient. The lack of pharmacokinetic data in pregnancy is of course due to the technical as well as the ethical problems related to such studies.
I. Problems Associated with Pharmacokinetic Studies in Pregnancy 1.1 Technical Problems Knowledge of the pharmacokinetics of a drug is derived from data from several individuals who should either be healthy and in the same condition, or suffer from the same disorder, the possible influence of which on one or more of the pharmacokinetic parameters must be known or possible to evaluate. Meaningful information can only be obtained for antibiotics which may be accurately assayed and are not metabolised or have metabolites which can be assayed. For studies of antibiotic levels in maternal plasma or serum, standards should preferably be prepared in plasma or serum from untreated pregnant women. Because of differences in protein content, a minimum requirement is to check that a standard prepared in such serum does not perform differently
298
in the assay from the actual standard to be used. Studies involving determination of fetal plasma levels of an antibiotic appropriately call for standards prepared in fetal plasma, as drugs may bind differently in fetal and adult plasma (Ehrnebo et a\., 1971; Pruitt and Dayton, 1971). As plasma or serum levels of antibiotic are determined as total levels of drug, a lower degree of protein binding in the fetus will give false high values for fetal levels. The technical problems involved in fetal tissue level monitoring are immense; freeing the tissue from blood, extracting the antibiotic, and selection of proper standards being examples of such problems. For adequate comparison of the pharmacokinetics ofa drug in pregnant and in non-pregnant individuals a control group is essential as variations in antibiotic serum levels between pregnant women may be marked. Controls must be matched for age, weight, sex, and if possible for medical condition as well. Samples to be assayed must be treated in an identical way as certain antibiotics will deteriorate within weeks even when stored at - 20° (Dornbusch, personal communication).
1.2 Ethical Problems Over the last 10 years ethical aspects have increasingly been brought into focus. It is regarded as unethical to give to a pregnant woman, who wants to proceed with her pregnancy any antibiotic solely for study purposes. Medication has to be medically justified. Certain accurate methods for determination of excretory capacity, metabolism, or volumes of distribution, are carried out with radioactive substances and obviously may not be employed in pregnant women. However, there is another ethical aspect of drug studies during pregnancy. Is it ethical not to perform such studies? Is it ethical to give a pregnant woman any drug for which the pharmacokinetics during pregnancy are not known? Many drugs which may be beneficial in treating pregnant women are not considered because of lack of knowledge of their activity
299
Pharmacokinetics of Antibiotics in Pregnancy and Labour
during pregnancy, thus depriving the pregnant woman of the normal choice of therapy. Because the pharmacodynamics of a drug most likely will have changed, the pregnant patient may be either undertreated or overtreated if a dosage is used that has been calculated from studies performed on men or non-pregnant women.
2. Pharmacokinetics in Pregnancy
Several reasons for alterations in drug kinetics in pregnancy have been discussed (Krauer and Krauer, 1977). Dose requirements of certain drugs, particularly antiepileptic drugs, are known to change when the patient becomes pregnant and as pregnancy proceeds (Eadie et ai., 1977). Even for antibiotics, pharmacokinetics will most likely change as pregnancy advances and changes may successively become more pronounced. The gradual increase in renal function will probably increase the elimination rate of antibiotics which are renally excreted. Altered drug metabolism (Feuer and Kardish, 1975; Gut et ai., 1976) may well influence pharmacokinetics of those antibiotics which are metabolised. Serum levels may also be altered in pregnant patients, the extent of change depending on the distribution properties of each antibiotic. Serum levels of antibiotics with a low volume of distribution, may be affected differently from those of antibiotics with a large volume of distribution. Little specific data are available on pharmacokinetics of antibiotics during pregnancy. Most studies include investigation of the transplacental passage and have been carried out at the time of abortion, labour and delivery, and caesarean section. When evaluating the results derived from such studies it must be borne in mind that the data may not be representative for any stage of pregnancy except precisely abortion, labour or caesarean section. Abortion (whether surgical or not), labour and caesarean section will most likely affect renal excretion, blood flow to various organs, and other parameters by which drug pharmacokinetics is regulated.
The stress involved for the patient may be an important factor in itself. Gastrointestinal absorption of antibiotics administered in such situations may be impaired. The influence of other pharmaceutical compounds, administered during labour or surgery, on the pharmacokinetics of an antibiotic is difficult to evaluate. Antibiotic levels in amniotic fluid and fetal serum, plasma, or tissues following administration of the drug to the mother during labour must be evaluated with precaution. Pronounced changes in the haemodynamics due to contractions of the uterus will most likely impair placental transfer of the drug.
3. Present Knowledge of Antibiotic Kinetics in Pregnancy
In spite of the problems involved investigations of antibiotics in pregnancy have been carried out, according to various protocols. Maternal levels of antibiotic in plasma or serum have been studied following single or repeated doses, mostly prior to abortion, delivery, or caesarean section. In some studies serial serum levels have been studied in the same patients whereas in other studies only the serum level at the time of delivery has been reported. In the latter type of study, the area under the serum concentration versus time curves have been constructed from single values for a great number of patients through variations of the dose-delivery interval. Also, the presentation of data is diverse, thus making comparisons between results obtained in different studies difficult. Evaluation of whether or not serum levels, renal excretion, or other parameters fora particular antibiotic alter during pregnancy is often impossible. However, several antibiotics have been studied and some preliminary information has accumulated. 3.1
~-Lactam
Antibiotics
3.1.1 Isoxazo[ylPenicillins Serum levels of methicillin following a single intravenous dose have been studied in women 32 to 40
Pharmacokinetics of Antibiotics .in Pregnancy and Labour
weeks pregnant or in labour (Depp et aI., 1970; MacAulay et aI., 1973). Similar studies have been published for dicloxacillin where a dose of 500mg was administered to women in labour (MacAulay et aI., 1968; Depp et aI., 1970). According to MacAulay et aI., the biological half-life of the elimination phase (tI/2~) for dicloxacillin was 2h. The maternal serum levels reported in these studies of dicloxacillin are quite different. It is not possible to determine on the basis of these studies if pharmacokinetics for methicillin and dicloxacillin are altered in pregnancy, but serial serum levels of methicillin administered to 3 women in active labour were reported to be generally lower than those of 1 man and I non-pregnant woman (MacAulay et aI., 1973). Pharmacokinetic parameters for oxacillin have been studied in 12 women 38 to 41 weeks pregnant at the time of amniocentesis (Bastert et aI., 1975). Following a single intravenous dose of 2g the calculated serum level at time zero (Co) was 79)lg/ml. The elimination constant was 1.044h- ' , tl/2~ was 40 min, the apparent volume of distribution (Vd) 25.3L, and the total clearance of drug was 441 mIl min. The authors point out that in this study the values for V d and clearance are higher than those in other studies performed in non-pregnant individuals. Calculations were based on observations during the first 2h following dose administration. 3.1.2 Ampicillin Ampicillin has been the most widely studied antibiotic in pregnant women, most often with regard to its transplacental passage. In comparing studies where the same dose has been used it is obvious that mean serum levels differ from one report to another (table The actual figures for serum levels are not always given but have then been deduced from graphs. The elimination half-life calculated from the graph in one of these studies is 46 minutes (Perry and Leblanc, 1967). It has also been reported to be 30 minutes at caesarean section (Boreus, 197 I), 40 minutes at delivery (Bastert et aI., 1973), and 40 minutes in the 8th month of gestation as opposed to
n.
300
Table I. Mean values reported in various publications for levels of ampicillin in serum and plasma (fl9/ml) in pregnant women at 1, 2 and 3h following a single intravenous dose of 500mg. Figurtls within brackets represent the nurnber of patients for which the mean was calculated. Standard deviations (SD) are given when available Time after dosage 1h 6.3(4)
SD 2.2
7.9(13) 15.0(8) 6.2(26)
2.14
Reference
2h
SD
3h
1.0(2)
0.1
SD
0.6(1)
Bray et al. (1966)
4.5(13)
3.4(13)
MacAulay et al. (1966)
6.0(4)
2.8 ( 3)
Perry and Leblanc (1967)
1.9(26)
0.9
0.7(26)
0.4
Philipson (1977)
58 minutes in women in labour (von Voigt et aI., 1978). The discrepancies may in part be due to a difference in the study groups, but also to different protocols and methods for calculations of t I / 2~' V d for ampicillin during labour or surgery in late pregnancy was found to be 0.23L/kg (Bastert et aI., 1973). Alterations in pharmacokinetics of ampicillin in pregnancy have been demonstrated in a crossover study in 26 women, 9 to 36 weeks pregnant, with lower urinary tract infections (Philipson, 1977). Ampicillin was administered intravenously as well as orally. Each woman served as her own non-pregnant control receiving identical doses after pregnancy. Plasma levels were found to be 50 % lower in pregnant women than in the same women when not pregnant. Some other relevant data are shown in table II. Further studies showed that subsequent doses of ampicillin during the course of oral therapy give similar levels in plasma and urine as the initial oral dose and that no accumulation of drug occurs (Philipson, 1978). By doubling the dose of ampicillin given to pregnant women their plasma levels were increased to 'normal'. Pivampicillin has been found to produce plasma levels during pregnancy of the same magnitude as those produced by oral ampicillin with twice the equimolar dose (Philipson, 1978). Following the onset of labour, however, blood levels could not be
Pharmacokinetics of Antibiotics in Pregnancy and Labour
301
Table II. Mean values and SD for apparent volume of distribution (Vd), elimination half-life (t 1/2~), bioavailability, renal clearance (Cl r), and recovery in urine of ampicillin in 26 women during and after pregnancy, following identical doses
1968; Paterson et al., 1970) produced serum levels that are considerably lower than one would expect (Walter and Heilmeyer, 1975). Studies of the pharmacokinetics of cephalothin in women at term or at the time of delivery have been carried out by MacAulay and Charles (I 968) and by Bastert et al. (I 974). In the first study an intravenous dose of Ig was administered to 54 women. Maternal serum was sampled at delivery which took place 9 to 810 minutes later. A serum concentration-time curve was constructed from the 54 individual observations. The elimination constant was found to be 86 x 10- 4 which gives a t 1/ 2~ of I.3h. This value is different from the value of about 40 minutes which is found in non-pregnant populations (Walter and Heilmeyer, 1975). Co was 30.2pg/rnl. Calculations of clearance from plasma give a value of 285ml/min. In the study by Bastert et al. (I 974) an intravenous dose of 2g was given to 25 women 38 to 41 weeks pregnant at the time of amniotomy. A regression line for serum levels was based only on the first 2h of observation. Co was 63.8pg/ml, tl/2~ 31 minutes, Vd 0.3 1L/ kg, and total clearance from serum 687ml/min. Even in these studies mean serum levels of cephalothin in pregnant women appear lower than those which would be expected (Walter and Heilmeyer, 1975). Cephazolin was given intramuscularly in a dose of 14mg/kg to 40 women 8 to 20 weeks pregnant at the time of interruption of pregnancy. Serum levels were measured and found to be 23 to 55 % lower than those which had earlier been reported for non-pregnant adults following a similar dose (Bernard et aI., 1977b). In this study t1 /2~ was l.5h which can be compared to a t I / 2~ of 2h found in non-pregnant individuals (Walter and Heilmeyer, 1975). In a study of cefatrizine, a single oral dose of I g was administered prior to abortion and was found to produce a wide range of serum levels at I h in women 8 to 20 weeks pregnant. Less variation was found at 2, 4 and 8h (Bernard et aI., I 977d). The t1 /2~ was 2Ah. The authors state that this is Ih longer than in healthy adult males but similar to that found in infected hospitalised patients. The authors believe that the
tl/2~ Vd (L/kg) (min)
Bioavailability
Cl r (ml/min)
Rec'overy in urine (% of IV dose)
(%)
During pregnancy SD
0.55
39
45.6
330
79.4
0.28
8.1
20.2
(137-535)0
24.7
After pregnancy SD
0.41
44
48.1
238
79.5
0.20
6.8
19.3
(85-425)a
30.6
Pvalue
a
< 0.05 < 0.05
Not significant
