THE JOUR~AL OF I~FECTIOUS DISEASES. VOL. 136, ~O. 3 • SEPTEMBER 1977 1977 by the University of Chicago. All rights reserved.
©
Pharmacokinetics of Ampicillin during Pregnancy From the Departments of Infectious Diseases and of Obstetrics and Gynecology, Danderyd Hospital, Danderyd; and the Department of Clinical Pharmacology, Karolinska Hospital, Stockholm, Sweden
Agneta Philipson
A clinical observation of treatment failure with ampicillin during pregnancy initiated the present study. A pregnant woman, suffering from an upper respiratory tract infection caused by a strain of Haemophilus influenzae sensitive to ampicillin, was given ampicillin orally (0.5 g four times daily), but her infection failed to respond. The level of ampicillin in serum was measured 2 hr after the administration of a dose and was found to be below the lowest measurable level. The woman was cured when ampicillin in the same dosage was given im. For investigation of whether the unexpectedly low level following oral administration was due to pregnancy, the same woman was given a single oral test dose of 0.5 g of ampicillin a few months after delivery. The serum level was measured after 2 hr and was found to be adequate. The aim of the present study was to investigate whether a difference in plasma levels of ampicillin between pregnancy and nonpregnancy is a general phenomenon. The study was designed as a comparison between the levels of ampicillin
in plasma and urine produced by a single iv and a single oral dose given to pregnant women and the levels produced by the same doses given to the same women after pregnancy.
Materials and Methods Twenty-six pregnant women volunteered for this study. All were attending out-patient maternity health clinics in the Stockholm, Sweden, area. Their pregnancies were normal. The patients qualified for the study by having asymptomatic bacteriuria caused by microorganisms sensitive to ampicillin (MIC, ~8 iLgjml) that were discovered by routine cultures. The women were otherwise healthy as judged by their histories, physical examinations, and routine laboratory tests performed on blood and urine. Sequence of the study. Each pregnant woman was prescribed oral treatment with ampicillin for 10 days (0.5 g four times daily). However, the first dose of 0.5 g (dose A) was given iv and comprised the first of four test doses (i.e., doses after which concentrations in plasma and urine were measured). The second test dose of 0.5 g (dose B) was given orally not less than one week (washout period) after completion of therapy (usually three to four weeks after dose A). After delivery, when normal menstruation had reappeared and breast-feeding had ceased (three to 12 months after delivery), the third test dose
Received for publication November 3, 1976, and in revised form April 5, 1977. This work was supported in part by research funds from the Karolinska Institute. The author thanks Mr. Anders Lindman for technical assistance. Please address requests for reprints to Dr. Agneta Philipson, Department of Infectious Diseases, Danderyd Hospital, S-18203 Danderyd, Sweden.
370
Downloaded from http://jid.oxfordjournals.org/ at Brunel University on July 18, 2015
Levels of ampicillin in plasma and urine were studied in 26 pregnant women during the 8 hr following an oral as well as an intravenous dose. Identical doses were given to the same women after pregnancy, and the resulting plasma and urine levels were compared. Assays were performed by means of a disk agar diffusion method. Renal plasma clearance was calculated. Plasma levels were significantly lower during pregnancy than after. The difference was unrelated to body weight and length of gestation but was associated with a significant increase in distribution volume and renal plasma clearance in pregnant women. Recovery of ampicillin in urine during 8 hr was identical during pregnancy and after. It is suggested that pregnant women require higher doses of ampicillin per kg of body weight than nonpregnant women in order to achieve similar concentrations in plasma. The experimental design used in the present study is suitable for evaluation of pharmacokinetics in pregnancy.
371
A mpicillin Pharmacokinetics in Pregnancy
per disks (Ford Mill 428; Birger Gustafsson AB, Stockholm, Sweden), 5.5 mm in diameter. The disks were left to dry at room temperature (about 24 C) for 1 hr and frozen dry at -20 C until assayed. Reference standards were prepared from sodium ampicillin dissolved in distilled water. For assays of plasma the ampicillin was diluted with fresh human plasma to concentrations of 100, 30, 10, 3, 1, 0.3, and 0.1 JLgjml (1 JLgjml 2.87 mmoljliter), and for assays of urine it was diluted with potassium phosphate buffer, pH 6.0, to the same concentrations and also to 300 JLgjml. Sets of standard disks of all concentrations were frozen dry at -20 C in separate vials. The test strain for assays of ampicillin in plasma was Sarcina lutea (strain ATCC 9341; American Type Culture Collection, Rockville, Md.), and for assays of drug in urine was Staphylococcus aureus (strain ATCC 6538P). Large square plates (235 X 235 mm) were filled with agar (Bacto Penassay Seed Agar; Difco, Detroit, Mich.) containing the test organism. Standard disks with different amounts of ampicillin, along with the disks to be assayed, were placed on the agar surface at random. The plates were incubated overnight at 37 C, and the diameter of each zone of inhibition was read. A standard curve (log concentration against diameter of the inhibition zone) was made up for each assay, and the results of the "unknown" disks were read against this curve, which was always linear with a correlation coefficient of at least 0.99. The lowest measurable concentration of ampicillin in plasma varied between 0.02 and 0.1 flgjml, and in urine between 0.5 and 3.6 flg/ml (diluted 1:20). All samples were assayed in duplicate, and a mean value of the two readings was calculated. The resulting values of ampicillin in plasma from each patient and dose were plotted against time on an arithmetic as well as on a semilogarithmic scale. The area under the curve (AVC) (hr X JLgj ml, but expressed arbitrarily in cm 2) was calculated for each patient and dose from the arithmetic curve by the trapezoidal rule. The AVC was used as a rough measurement of the total amount of absorbed, circulating, and unexcreted drug. The bioavailability of orally administered ampicillin was calculated from the arithmetic
=
Downloaded from http://jid.oxfordjournals.org/ at Brunel University on July 18, 2015
of 0.5 g of ampicillin (dose C) was given as a single iv injection. The fourth test dose (dose D) was given as a single oral dose of 0.5 g not less than one week after dose C. In this way each woman served as her own nonpregnant control. The mean body weights (± so) of the patients at the time of test doses A, B, C, and D were 64.7 ± 10.5 kg, 66.6 ± 11.0 kg, 63.0 ± 12.0 kg, and 62.9 ± 12.0 kg, respectively. The mean gestational age was 19.5 weeks (range, nine to 33 weeks) and 23.2 weeks (range, 13-36 weeks) at the time of doses A and B, respectively. Sampling schedule. Immediately prior to administration of the test doses, each woman was instructed to empty her bladder, and blood was drawn from a cubital vein through an indwelling catheter. Blood (4 ml) was also collected in this manner in heparinized tubes 10, 20, 30, and 45 min and 1, 2, 3, 5, and 8 hr after administration of doses A and C and 1, 2, 3, 5, and 8 hr after doses Band D. Blood samples were centrifuged, and the plasma was removed. V rine was always collected 3 hr and 8 hr after each dose. The volumes were measured. Prior to administration of doses Band D, the patients had been fasting for at least 8 hr. Food and drink were allowed 3 hr after administration of a dose. None of the patients had taken any antibiotic for at least one month prior to the study, and at the time of the investigation, none were taking any other medication, except for iron supplements. No iron was taken on the days that the oral test doses were given. None of the patients were taking oral contraceptives when doses C and D were given. Ampicillin for oral use was administered as commercial tablets (0.5 g; Doktacillin,® Astra, Sodertiilje, Sweden) and for iv use as sodium ampicillin (corresponding to 0.5 g of Doktacillin®), which was dissolved in 2 ml of sterile water. The iv injection was completed in less than 1 min. The tablets were ingested with --30 ml of water. Assays for levels of ampicillin in plasma and urine were performed by means of a disk agar diffusion method [1]. To bring urine levels down to concentrations that could be measured with more accuracy, we diluted the urine samples 1:20 with potassium phosphate buffer, pH 6.0. SampIes of fresh plasma or urine, diluted and undiluted, in 10-JLI volumes were applied to pa-
372
Philipson
Statistical analysis was performed by Student's t-test for paired observations when these appeared to be normally distributed. For clearly skewed observations, statistical analysis was performed by sign test for paired observations.
jJ9/ ml 300.0
100.0
0 0
a Z 0
30.0
i=
Results
©
Pharmacokinetics of Ampicillin during Pregnancy From the Departments of Infectious Diseases and of Obstetrics and Gynecology, Danderyd Hospital, Danderyd; and the Department of Clinical Pharmacology, Karolinska Hospital, Stockholm, Sweden
Agneta Philipson
A clinical observation of treatment failure with ampicillin during pregnancy initiated the present study. A pregnant woman, suffering from an upper respiratory tract infection caused by a strain of Haemophilus influenzae sensitive to ampicillin, was given ampicillin orally (0.5 g four times daily), but her infection failed to respond. The level of ampicillin in serum was measured 2 hr after the administration of a dose and was found to be below the lowest measurable level. The woman was cured when ampicillin in the same dosage was given im. For investigation of whether the unexpectedly low level following oral administration was due to pregnancy, the same woman was given a single oral test dose of 0.5 g of ampicillin a few months after delivery. The serum level was measured after 2 hr and was found to be adequate. The aim of the present study was to investigate whether a difference in plasma levels of ampicillin between pregnancy and nonpregnancy is a general phenomenon. The study was designed as a comparison between the levels of ampicillin
in plasma and urine produced by a single iv and a single oral dose given to pregnant women and the levels produced by the same doses given to the same women after pregnancy.
Materials and Methods Twenty-six pregnant women volunteered for this study. All were attending out-patient maternity health clinics in the Stockholm, Sweden, area. Their pregnancies were normal. The patients qualified for the study by having asymptomatic bacteriuria caused by microorganisms sensitive to ampicillin (MIC, ~8 iLgjml) that were discovered by routine cultures. The women were otherwise healthy as judged by their histories, physical examinations, and routine laboratory tests performed on blood and urine. Sequence of the study. Each pregnant woman was prescribed oral treatment with ampicillin for 10 days (0.5 g four times daily). However, the first dose of 0.5 g (dose A) was given iv and comprised the first of four test doses (i.e., doses after which concentrations in plasma and urine were measured). The second test dose of 0.5 g (dose B) was given orally not less than one week (washout period) after completion of therapy (usually three to four weeks after dose A). After delivery, when normal menstruation had reappeared and breast-feeding had ceased (three to 12 months after delivery), the third test dose
Received for publication November 3, 1976, and in revised form April 5, 1977. This work was supported in part by research funds from the Karolinska Institute. The author thanks Mr. Anders Lindman for technical assistance. Please address requests for reprints to Dr. Agneta Philipson, Department of Infectious Diseases, Danderyd Hospital, S-18203 Danderyd, Sweden.
370
Downloaded from http://jid.oxfordjournals.org/ at Brunel University on July 18, 2015
Levels of ampicillin in plasma and urine were studied in 26 pregnant women during the 8 hr following an oral as well as an intravenous dose. Identical doses were given to the same women after pregnancy, and the resulting plasma and urine levels were compared. Assays were performed by means of a disk agar diffusion method. Renal plasma clearance was calculated. Plasma levels were significantly lower during pregnancy than after. The difference was unrelated to body weight and length of gestation but was associated with a significant increase in distribution volume and renal plasma clearance in pregnant women. Recovery of ampicillin in urine during 8 hr was identical during pregnancy and after. It is suggested that pregnant women require higher doses of ampicillin per kg of body weight than nonpregnant women in order to achieve similar concentrations in plasma. The experimental design used in the present study is suitable for evaluation of pharmacokinetics in pregnancy.
371
A mpicillin Pharmacokinetics in Pregnancy
per disks (Ford Mill 428; Birger Gustafsson AB, Stockholm, Sweden), 5.5 mm in diameter. The disks were left to dry at room temperature (about 24 C) for 1 hr and frozen dry at -20 C until assayed. Reference standards were prepared from sodium ampicillin dissolved in distilled water. For assays of plasma the ampicillin was diluted with fresh human plasma to concentrations of 100, 30, 10, 3, 1, 0.3, and 0.1 JLgjml (1 JLgjml 2.87 mmoljliter), and for assays of urine it was diluted with potassium phosphate buffer, pH 6.0, to the same concentrations and also to 300 JLgjml. Sets of standard disks of all concentrations were frozen dry at -20 C in separate vials. The test strain for assays of ampicillin in plasma was Sarcina lutea (strain ATCC 9341; American Type Culture Collection, Rockville, Md.), and for assays of drug in urine was Staphylococcus aureus (strain ATCC 6538P). Large square plates (235 X 235 mm) were filled with agar (Bacto Penassay Seed Agar; Difco, Detroit, Mich.) containing the test organism. Standard disks with different amounts of ampicillin, along with the disks to be assayed, were placed on the agar surface at random. The plates were incubated overnight at 37 C, and the diameter of each zone of inhibition was read. A standard curve (log concentration against diameter of the inhibition zone) was made up for each assay, and the results of the "unknown" disks were read against this curve, which was always linear with a correlation coefficient of at least 0.99. The lowest measurable concentration of ampicillin in plasma varied between 0.02 and 0.1 flgjml, and in urine between 0.5 and 3.6 flg/ml (diluted 1:20). All samples were assayed in duplicate, and a mean value of the two readings was calculated. The resulting values of ampicillin in plasma from each patient and dose were plotted against time on an arithmetic as well as on a semilogarithmic scale. The area under the curve (AVC) (hr X JLgj ml, but expressed arbitrarily in cm 2) was calculated for each patient and dose from the arithmetic curve by the trapezoidal rule. The AVC was used as a rough measurement of the total amount of absorbed, circulating, and unexcreted drug. The bioavailability of orally administered ampicillin was calculated from the arithmetic
=
Downloaded from http://jid.oxfordjournals.org/ at Brunel University on July 18, 2015
of 0.5 g of ampicillin (dose C) was given as a single iv injection. The fourth test dose (dose D) was given as a single oral dose of 0.5 g not less than one week after dose C. In this way each woman served as her own nonpregnant control. The mean body weights (± so) of the patients at the time of test doses A, B, C, and D were 64.7 ± 10.5 kg, 66.6 ± 11.0 kg, 63.0 ± 12.0 kg, and 62.9 ± 12.0 kg, respectively. The mean gestational age was 19.5 weeks (range, nine to 33 weeks) and 23.2 weeks (range, 13-36 weeks) at the time of doses A and B, respectively. Sampling schedule. Immediately prior to administration of the test doses, each woman was instructed to empty her bladder, and blood was drawn from a cubital vein through an indwelling catheter. Blood (4 ml) was also collected in this manner in heparinized tubes 10, 20, 30, and 45 min and 1, 2, 3, 5, and 8 hr after administration of doses A and C and 1, 2, 3, 5, and 8 hr after doses Band D. Blood samples were centrifuged, and the plasma was removed. V rine was always collected 3 hr and 8 hr after each dose. The volumes were measured. Prior to administration of doses Band D, the patients had been fasting for at least 8 hr. Food and drink were allowed 3 hr after administration of a dose. None of the patients had taken any antibiotic for at least one month prior to the study, and at the time of the investigation, none were taking any other medication, except for iron supplements. No iron was taken on the days that the oral test doses were given. None of the patients were taking oral contraceptives when doses C and D were given. Ampicillin for oral use was administered as commercial tablets (0.5 g; Doktacillin,® Astra, Sodertiilje, Sweden) and for iv use as sodium ampicillin (corresponding to 0.5 g of Doktacillin®), which was dissolved in 2 ml of sterile water. The iv injection was completed in less than 1 min. The tablets were ingested with --30 ml of water. Assays for levels of ampicillin in plasma and urine were performed by means of a disk agar diffusion method [1]. To bring urine levels down to concentrations that could be measured with more accuracy, we diluted the urine samples 1:20 with potassium phosphate buffer, pH 6.0. SampIes of fresh plasma or urine, diluted and undiluted, in 10-JLI volumes were applied to pa-
372
Philipson
Statistical analysis was performed by Student's t-test for paired observations when these appeared to be normally distributed. For clearly skewed observations, statistical analysis was performed by sign test for paired observations.
jJ9/ ml 300.0
100.0
0 0
a Z 0
30.0
i=
Results
