Placental Transmission of Antibiotics
Roger E. Bawdon and Michael Hnat
Table Of Contents
Roger E. Bawdon, PhD
Michael Hnat, DO
DRUG CLASSES AND PREGNANCY CATEGORIES|
CLASSIFICATION OF ANTIBIOTICS
During the last 2 decades, a myriad of new antimicrobial and antiviral agents have been developed to combat the new infections and resistant organisms in pregnant women. These agents are used for infections in immunologically competent women and infections related to immunocompromised pregnant women suffering from the complications of human immunodeficiency virus (HIV).
Because of legislation in 1977, it is difficult to study the in vivo transfer of new antimicrobial agents in the pregnant patient. With this change, the only alternatives for studying the placental transmission of antimicrobials are animal models and the ex vivo human placental perfusion model. The only animal model with an identical placental anatomy is the nonhuman primate, the macaque (Macaca nemestrina), which is expensive to purchase and time-consuming to use, considering the amount of time involved to term pregnancy and the amount of data obtained from an individual animal. The remaining alternative, the ex vivo placental perfusion model, does not provide pharmacokinetic data but does provide transfer data and relates to previous studies by the use of standardized procedures and reference compounds.
|DRUG CLASSES AND PREGNANCY CATEGORIES|
Because of ethical considerations, information on the use of drugs, including antimicrobial agents, in pregnant patients comes only from inadvertent use or accidental use in early pregnancy. Other data, although not completely reliable, come from studies in pregnant animals. Because of this, many new drugs are difficult to assign to the U.S. Food and Drug Administration (FDA) categories. However, in this chapter, the FDA pregnancy categories are discussed only briefly. The FDA has established five categories (A, B, C, D, X) to indicate the teratogenic possibilities during the perinatal period to evaluate the short- and long-term effects of drugs on the mother and fetus.1
Category A drugs are those for which well-controlled studies in pregnant women have failed
to show any fetal risk.
The drugs in these categories are discussed further, and the previous criteria can be related to the placental transfer of these compounds. Although the FDA categories may be of some value, other considerations must be included in the use of antimicrobials in pregnancy. Among the most important considerations for the use of any drug is the gestational age of the fetus. The use of any drug, including antimicrobials, in the first trimester of pregnancy should be avoided unless it is positively justified because most drugs have some side effects that could affect the development of the fetus. This has become an unfortunate dilemma because of the increased incidence of genital herpes and HIV infection in pregnant women. Many women are HIV-infected before pregnancy and are on zidovudine therapy during the first trimester, and since the advent of a registration system, no known anomalies have occurred. Similar registers have been developed for the use of acyclovir in first trimester patients.
|CLASSIFICATION OF ANTIBIOTICS|
Antimicrobials may be considered in groups based on their chemical structure. They may have similar antimicrobial activity but differ in their pharmacokinetic properties. Table 1 is a brief list of common antimicrobial agents and their FDA categories.
FDA, U.S. Food and Drug Administration.
With all the new classifications of antimicrobials confronting the physician, the choice of which compound may be used for a specific infection is confusing. Further complicating the situation has been the rapid development of resistant microbial strains. Laboratory involvement in identification and susceptibility is important in the treatment of an infectious agent. Because the central nervous system is still vulnerable to mutogenesis until the last trimester, treatment of viral infections should be delayed until the last trimester. Further complications in the pregnant patient may alter the toxicity of drugs, including hypertension, diabetes mellitus, and preeclampsia. Not every antimicrobial is discussed in this chapter because some do not apply to the pregnant patient or are absolutely contraindicated.
The human placenta is different from all other tissue barriers in the human body. It originates from the chorion supplied by the allantoic vessels. The placenta develops as a unique organ that is partly maternal and partly the independent fetal portion. The human placenta is unique to primates and nonhuman primates in that it is discoidal, has a fetomaternal villous interdigitation, has a monochorial placental barrier, and is multivillous. It is different, for example, from the murine placenta that, although discoidal, has a labyrinthine fetomaternal interdigitation, a hemotrichorial placental barrier, and a countercurrent fetomaternal blood flow interrelation. With these differences, it is difficult to relate placental transfer of drugs from an animal model to the pregnant human. The basic structure of the placenta is vascular, allowing the close contact of the maternal and fetal circulations through the fetal intervillous tree and the maternal intervillous space.
The placenta has the capability to selectively transfer drugs and metabolic products in either direction rapidly and completely or to a limited extent. The transfer is based on molecular weight, pH, protein binding, lipophilicity, and type of transfer. The three mechanisms of transfer are passive diffusion, facilitated diffusion, and active transport.
Simple diffusion may occur with almost any compound with molecular weight less than 5000 daltons. Most low-molecular-weight drugs fall into this category, including the antimicrobial agents. In addition to allowing small molecules to cross into the fetal circulation, a bidirectional flow permits the removal of metabolites from the fetal compartment. Simple diffusion is the rapid equilibrium of a compound from the maternal to the fetal compartment in various amounts of time, depending on the following factors: molecular weight, pKa (pH), structural formula, placental blood flow, lipophilic or lipophobic characteristics, and concentration.
Simple diffusion or passive diffusion is defined as the passage of drugs or metabolites without the use of energy, movement that is described by Fick’s equation. This system may be useful in the physical chemistry laboratory, but in the case of practical drug transfer, many of the chemical-physical parameters (constants) are not available. The patient’s biochemical and physical variability factors include plasma volume, cardiac output, renal function, and plasma protein concentration.
Simple diffusion is a free exchange of small compounds with equilibrium of the maternal fetal concentration of nutrients into the fetal compartment and the opposite for metabolic products from the fetal compartment. This bidirectional equilibrium does not necessarily mean that the concentrations of the compounds are equivalent in the maternal and fetal compartments. However, the fetus plays an important role in the equilibrium of a compound because it swallows amnionic fluid and urinates; there is a sort of drug recirculation within the fetal compartment along with the transfer back to maternal blood. Additional factors that are important in the transfer and equilibrium of compounds in the molecular weight range of up to 5000 daltons are symmetry of the molecule, hydrophilic characteristics, hydrophobic characteristics, and solubility.
Facilitated diffusion differs from simple diffusion in that the compounds are transported across the placenta by a carrier substance within the placenta. These carriers are usually proteins that move freely about the placenta. Examples of compounds that are transported across the placenta by this mechanism are glucose, iron, and ascorbic acid. These compounds may be found in the fetal circulation at two to three times that of the maternal circulation.
Active transport of nutrients is the passage across placental membranes at the expense of energy. The transport of these compounds is against the concentration gradient at the expense of adenosine triphosphate (ATP) or electron transport. Such transport in the placenta involves sodium, calcium, and ATPase.
In addition to the three most common transfer mechanisms, it is possible to transfer drugs and metabolic compounds by pinocytosis, phagocytosis, and discontinuity of placental membranes. Because of the various possible transfer mechanisms, placental transmission of antimicrobials is extremely variable and unpredictable. The remainder of the chapter is devoted to the available data on the placental transfer of these agents.
Ampicillin and Sulbactam
Several studies have been conducted on the placental transfer of ampicillin with and without the inclusion of the β-lactamase inhibitor sulbactam (Table 2). In studies in which 250 mg to 1 g of ampicillin was given by intravenous doses, the cord blood levels exceeded 5 μg/mL for at least 4 hours, and there was accumulation in the amnionic fluid.2,3,4 Maternal concentrations ranged from 3 μg/mL for the 500-mg oral dose and from 4.6 to 60 μg/mL for the intravenous doses.
* Not all the commonly used antibiotics in obstetrics have been studied in the ex vivo human placental model, but they have been included elsewhere in the text and placental transfer has been observed in all cases.
† Clearance index is based on transfer of antipyrine, a freely diffusable small molecule.
In another study, ampicillin was given to three women at term with chorioamnionitis.5,6 The maternal cord serum and placental tissue were assayed for ampicillin at delivery. The cord blood concentration was 4.7 μg/mL, and the cord blood to maternal blood (c/m) ratio was 0.71, which is in agreement with the ratios of noninfected patients. In a third study done in the same laboratory, five women received ampicillin and sulbactam (Unasyn) while in labor, and the c/m ratio for ampicillin was 1.0, and for sulbactam, the c/m ratio was 1.3. These data support the use of ampicillin or ampicillin and sulbactam for use in chorioamnionitis, with the latter agent having good anaerobic coverage.
Ticarcillin and Clavulanic Acid
Ticarcillin and clavulanic acid have been studied in the placental perfusion model with maternal blood, cord blood, and placental tissue samples collected at delivery.7 In five patients at delivery, maternal blood, cord blood, and placental tissue were collected, and the time of infusion to collection was 1 hour. The maternal blood level was 85.3 μg/mL, the cord blood level was 54.9 μg/mL, and the placental tissue level was 23.4 μg/g, suggesting excellent transfer of the drugs.
In the placental perfusion model, the transfer of clavulanic acid did not appear to transfer as readily as other compounds. At maternal therapeutic peak concentrations of clavulanic acid, the drug could not be detected in the fetal circulation. The poor transfer of this compound does not necessarily contraindicate its use for treatment of pregnant patients.
In a study to assess the placental transfer of cefazolin, 2 g were given intravenously as prophylaxis for intravascular intrauterine transfusion for Rh isoimmunization. Samples of maternal and cord blood were collected, as was amnionic fluid for antibiotic assays. Sample times ranged from 6 to 48 minutes, and the concentrations of cefazolin in maternal serum was 121 ± 50 μg/mL. Cord blood levels were 20.0 ± 12 μg/mL, and amnionic fluid concentrations were 0.9 ± 0.4 μg/mL. The c/m ratio was 0.16, and the amnionic to maternal (a/m) ratio was 0.007. These data confirm the rapid passage of cefazolin across the placenta. Because of the short sampling times, the ratios of c/m and a/m are probably low compared with what they would be on long-term steady-state therapy.8
In the placental perfusion model for comparison of cefoperazone and ceftizoxime, it was found that ceftizoxime transferred more readily across the placenta than the cefoperazone. Cefoperazone had a clearance index of 0.037 ± 0.0014, whereas ceftizoxime had a clearance index of 0.126 ± 0.013 (see Table 2). These data suggest that ceftizoxime may be more useful in pregnant women, but cefoperazone also crosses and has been useful in obstetric infections.7
Before delivery, a bolus dose of cefotaxime was given to five women. At delivery, maternal blood, cord blood, and placental tissue were collected for assay. The maternal serum and cord serum had cefotaxime levels of 8.9 μg/mL and 8.6 μg/mL, respectively, and the tissue concentration was 1.4 μg/g. These data suggest excellent placental transfer of cefotaxime.6
Cefoxitin levels after bolus administration to six term pregnant women at delivery were 24.0 ± 26.3 μg/mL in maternal serum, 8.5 ± 7.0 μg/mL in cord blood, and 11.5 ± 8.8 μg/g in placental tissue. These data essentially agree with the findings of another study in which 35 term patients had an average maternal serum level of 25 μg/mL, and the cord blood level was 15 μg/mL.8,9,10,11,12,13 These studies indicate that cefoxitin readily crossed the human placenta and achieved therapeutic levels.
In 20 patients, after a steady-state dose of 2 g administered intravenously every 8 hours, maternal blood, cord blood, and amniotic fluid were collected at delivery. The mean concentrations of ceftizoxime were 11.96± 2.35 μg/mL, 24.5 ± 4.78 μg/mL, and 43.45 ± 4.97 μg/mL, respectively. These data indicate that ceftizoxime is potentially one of the best broad-spectrum cephalosporins used for treatment of intrauterine infection.9,11
In patients administered an intramuscular dose of 750 mg, the maternal blood level was three to five times higher than cord blood. These data suggest an incomplete transfer to the fetal compartment.12,13,14
In a rather extensive controlled study using term pregnant and nonpregnant women as controls, 500 mg of imipenem was infused over 20 minutes. Samples of maternal blood, cord blood, and amnionic fluid were collected from patients 30 minutes after infusion. Therapeutic concentrations were achieved in umbilical blood, with a maternal to fetal blood ratio of 30%, 30 minutes after infusion. Amniotic fluid concentrations were low at this time, but this could readily be explained by the short time of sample collection after infusion.15
A dose of 4 g of piperacillin was given as prophylaxis for intrauterine transfusions in three patients. The sample time ranged from 12 to 42 minutes, with maternal serum concentrations of 138 ± 69 μg/mL. Cord blood serum levels were 22 ± 12 μg/mL, and amnionic fluid levels were 0.6 ± 0.7 μg/mL. The c/m ratio was 0.15, and the a/m ratio was 0.004. These data are similar to those for cefazolin.8
There is little information on the maternal fetal transfer of the frequently used combinations of antimicrobials for the treatment of urinary tract infections. In the placental perfusion model, neither trimethoprim nor sulfamethoxazole crossed the placenta readily, suggesting that there would be no adverse effects. Although trimethoprim is a folate antagonist and its use in pregnancy in the United States is not recommended, its use in England is common, and it appears to be efficacious and safe.16,17
Clindamycin was shown to readily cross the placenta in three term pregnant patients. Although the m/c ratio was only 0.15, the placental tissue apparently has an affinity for clindamycin, because the maternal blood to placental tissue ratio was 1.1. These data suggest that clindamycin, because of its good anaerobic coverage, is useful for treatment of patients with chorioamnionitis.5,18 In another study, 54 women were given a 600-mg intravenous dose, with similar results in maternal and cord blood.
Gentamicin has been studied in the pregnant patient, with two studies reporting conflicting data. In one study, maternal blood levels were 3.5 ± 0.95 μg/mL, and cord blood levels were 2.2 ± 0.87 μg/mL, for a ratio of 0.62. Placental membranes had a concentration of 13.9 ± 10 μg/g, the highest of any drug in the study.8 In another study, cord blood levels were found to be 42% lower than maternal blood levels, with gentamicin not detected in amnionic fluid. The study concluded that cord blood levels were subtherapeutic.19 Amikacin was found to have low concentration in amniotic fluid, suggesting poor placental penetrations.
Because fluoroquinolones are not used to treat chorioamnionitis or for prophylaxis in pregnancy, there is a paucity of information on the maternal fetal transfer of these compounds. In one study of 20 women scheduled for pregnancy termination between 19 to 25 weeks′ gestation, two 200-mg intravenous doses of ciprofloxacin were given at 12- and 24-hour intervals prior to termination. Blood was collected at 4, 8, and 12 hours after the initial dose. Maternal blood levels were 0.28, 0.09, and 0.01 μg/mL, respectively. Amniotic fluid levels were 0.12, 0.13, and 0.10 μg/mL, respectively.20 Trovafloxacin placental transfer was done in the ex vivo human placenta and the clearance index was 0.19 ± 0.13 at trough concentrations with little or no unusual accumulation. This study concluded that trovafloxacin crossed the human placenta by simple diffusion.20,21
Vancomycin is a narrow-spectrum, bactericidal antibiotic with limited and restricted use in medical practice, especially in obstetrics. It is highly efficacious against gram-positive organisms, and its use is reserved for the treatment of Clostridium difficile colitis, methicillin-resistant Staphylococcus aureus, and endocarditis prophylaxis in patients allergic to penicillin. In addition, the Centers for Disease Control and Prevention now recommends 1 g of vancomycin every 12 hours as a last resort in parturients, who are penicillin allergic, for prophylaxis against group B streptococcal (GBS) strains that are resistant to clindamycin and erythromycin.22
Vancomycin has a narrow therapeutic range secondary to its nephrotoxic and ototoxic side effects and is almost solely eliminated by glomerular filtration. In nonpregnant patients with normal renal function, the serum half-life is 6 to 8 hours.23 When 500 mg of vancomycin is slowly administered intravenously, the peak serum levels can reach 50 μg/mL and the trough serum levels are between 6 and 10 μg/mL.23 One hour after administration of vancomycin 1 g, serum peak levels range from 20 to 50 μg/mL and trough levels can be between 5 and 12 μg/mL.23 Bourget and colleagues demonstrated therapeutic serum and normal half-life levels in pregnant woman.24
In the treatment of GBS, it is important for the antibiotic to reach therapeutic or bacterial concentrations in the fetal circulation. Two small studies have demonstrated transplacental passage of vancomycin.24,25 In a case report by Bourget and colleagues, maternal serum and amniotic fluid vancomycin levels were measured in a woman treated for chorioamnionitis from 26 to 28 weeks’ gestation with vancomycin hydrochloride (1 g every 12 hours) and tobramycin sulfate.24 Fetal cord blood vancomycin concentrations were measured at birth. Peak maternal serum levels were between 28.3 and 41.55 μg/mL and amniotic fluid levels ranged from 1.02 μg/mL on day 1, to 3.80 μg/mL on day 3, and 9.20 μg/mL on day 13. At delivery, the fetal and maternal serum levels were 3.65 and 4.80 μg/mL, respectively. After 13 days of maternal antibiotic therapy, the fetal-maternal concentration ratio was 0.76.24
In another study, Reyes and associates also demonstrated transplacental crossing of vancomycin.25 Vancomycin levels were measured in two infants born to mothers who were treated for methicillin-resistant S. aureus infections for at least 1 week prior to delivery. The results are shown in Table 3. Most importantly, Reyes and associates concluded that infants (n = 30) exposed to vancomycin during the second and third trimester did not experience hearing loss or nephrotoxicity.25
*6 hr after the end of infusion.
†2.5 hr after the end of infusion.
Reyes MP, Ostrea EM, Cabinian AE, et al: Vancomycin during pregnancy: does it cause hearing loss or nephrotoxicity in the infant? Am J Obstet Gynecol 161:977, 1989
Of note, in an ex vivo human placental model with one placenta in which 1 U of heparin was present, less than 0.5 μg/mL of vancomycin could be detected in the fetal circulation only when 14.8 μg/mL of vancomycin was present in the maternal circulation (Hnat M, Bawdon R, unpublished data).
Vancomycin has a molecular weight of approximately 1450 daltons and should cross from the maternal to the fetal circulation. Limited data have shown that vancomycin, in vivo, does cross the human placenta at therapeutic concentrations.
Anti–Human Immunodeficiency Syndrome Compounds
Not only is the knowledge of placental transfer of most anti-HIV drugs inadequate, but with the rapid increase in the number of HIV-infected pregnant women, there is a lack of information on the transfer of anti-HIV compounds in pregnant and nonpregnant adult women and on the perinatal transmission of antiviral compounds.
Several anti-HIV agents may reduce the HIV RNA copy number in infected patients, but the potential adverse long-term effects on the fetus are unknown in most cases. In this chapter, the information is restricted to three different classes of agents.
Initially, cases of zidovudine, formerly called AZT, treatment were confined to inadvertent treatment of HIV-infected women who became pregnant while being treated with zidovudine. These individual case studies indicated that there were no adverse effects attributed to the inadvertent treatment of HIV during pregnancy.
Subsequent studies in the placental perfusion model and the pediatric acquired immunodeficiency syndrome (AIDS) clinical trial group 076 were initiated to determine the efficacy and placental transfer of zidovudine treatment after the first trimester of pregnancy. The results of the research studies and the clinical trials determined that zidovudine did cross the human placenta but did not accumulate to a great extent and that it reduced the incidence of HIV transmission from mother to infant from approximately 24% to about 8%. Because of the development of viral resistance to zidovudine alone, the use of nucleoside inhibitors in combination with other nucleoside, nonnucleoside, and protease inhibitors has become important. The anti-HIV drugs used in combination or triple therapy are shown in Table 4.26,27,28
AZI, azidothymidine; ddC, dideoxycytidine; ddT, dideoxyinosine; d4T, stavudine; FDA, U.S. Food and Drug Administration; 3TC, lamivudine.
* Clearance index unknown.
† Same as lopinavir (Keletra).
Examples of nucleoside inhibitors used in combination with zidovudine are stavidine (Zerit [d4T]) and lamivudine (Epivir [3TC]). Both compounds are synergistic with zidovudine and are not used for single-drug therapy. In a South African study, lamivudine (4 mg/kg/day) and zidovudine (8 mg/kg/day) were used in combination at 38 weeks and continued 1 week postpartum in 20 HIV-infected pregnant women. No viral RNA was found in the 20 newborns at birth and at 2 weeks postpartum.29
Because development of resistance to combination therapy is common in both nonpregnant and pregnant patients, it may be necessary to change the two-drug combination and add a protease inhibitor. This becomes complicated because of the maternal side effects and the unknown side effects to the fetus. It may be necessary to add a different protease inhibitor if the viral RNA copy number increases or the CD4 count decreases. The addition of the protease inhibitor should be approached with caution. Although the short-term effects of the additional therapy may be beneficial to the mother and fetus, long-term effects of triple therapy are relatively unknown. For the latest information on the treatment guidelines for the various scenarios of HIV infection during pregnancy, contact the HIV/AIDS Treatment Information Service website (http://www.hivatis.org).30
Other Antiviral Compounds
Ganciclovir has been studied in the placental perfusion model and, like other antiviral nucleoside analogs, readily crosses the placenta. In the perfusion model, the clearance index of ganciclovir was nearly identical to that of acyclovir (see Table 3). Other than these data, little is known about the pharmacology and safety of ganciclovir during pregnancy.
Acyclovir, because of its widespread use in the treatment of genital herpes, has been studied in the placental perfusion model and in numerous cases of exposed pregnancies over a period of at least 8 years. In more than 300 cases of the use of acyclovir in pregnancy, no birth defects were associated with its use.31,32
Although tuberculosis is on the increase, particularly the disease caused by the atypical mycobacteria complicating HIV infections, the disease is still uncommon in pregnant patients. Studies conducted in the 1950s found that isoniazid readily crossed the placental membranes after a 100-mg dose. In other studies to determine the teratology of isoniazid, more than 200,000 pregnant women received isoniazid, and there were no teratogenic effects.34,35,36
No in vivo human studies have been conducted on the placental transfer of rifampin and rifabutin. The drugs have been studied in the ex vivo human placental model. At trough and peak concentrations of rifampin and rifabutin, the clearance index of rifabutin was 0.44 ± 0.11 and 0.37 ± 0.15, respectively, or approximately three times that of rifampin (see Table 2). Placental tissue concentration of rifabutin was also three times higher than that of rifampin. The effect of these differences on the pregnant patient and the fetus is unknown.37
Little information about the placental transfer and teratologic effects of antifungal in humans has been reported. Although amphotericin B is a category B drug and itraconazole, ketoconazole, and fluconazole are all category C drugs, the use of these agents during pregnancy should be evaluated with each case. They should be used only in extreme cases, when the health of the mother requires their use.
Although parasitic infections during pregnancy in permanent residents in the United States are relatively uncommon, the mobility of U.S. citizens and increased numbers of immigrants amplifies the possibility of the combination of pregnancy and parasitic disease. When dealing with a potential parasitic infection, the patient should be asked where she has been and where she is going. The advice given may provide answers to or prevent contracting a serious complication from a parasite that may be endemic to a geographic location.
Because little information is available on the placental transfer of the drugs used for treatment of the parasitic diseases, risk assessment of the pregnant patient must be considered. Most of the drugs used in the treatment of parasites in pregnant women are considered category C drugs. The risk of these drugs causing problems with the fetus are minimal. However, if treatment can be avoided until after delivery, it is probably advisable.
Much of the data on maternal fetal transfer of drugs during pregnancy is limited to a single dose of the drug before delivery. Although a number of patients may be included in a study, single samples from each patient do not represent steady-state drug levels in the maternal and fetal compartments. It is important to interpret these data with some degree of caution because of their variability.
2. MacAulay MA, Abou-Sabe M, Charles D: Placental transfer of ampicillin. Am J Obstet Gynecol 96:943, 1996
3. Kraybill NE, Chaney NE, McCarthy LR: Transplacental ampicillin: Inhibitory concentrations in neonatal serum. Am J Obstet Gynecol 138:793, 1980
4. Blecher TE, Edgar WM, Melville HA, et al: Transplacental passage of ampicillin. BMJ 5480:137, 1966
7. Fortunato SJ, Bawdon RE, Swan KF, et al: Transfer of Timentin (ticarcillin and clavulanic acid) across in vitro perfused human placenta: Comparison with other agents. Am J Obstet Gynecol 167:1595, 1992
12. Bergogne-Berezin E, Pierce J, Rauvillois JL, et al: Placental transfer of cefuroxime.Current Chemotherapy for Infectious Diseases: Proceedings of the11th International Conference on Antimicrobial Agents and Chemotherapy Boston, p 1168, 1980
14. Holt DE, Fisk NM, Spencer JA, et al: Transplacental transfer of cefuroxime in uncomplicated pregnancies and those complicated by hydrops or changes in amniotic fluid volume. Arch Dis Child 68:54, 1993
23. Fekety R: Vancomycin, teicoplanin, and the streptogramins: Quinupristin and dalfopristin. In Mandell GL, Bennett JE, Dolin R (eds) (eds): Principles and Practice of Infectious Disease. pp 382, 387 Philidelphia, Churchill Livingstone, 2000
28. Forestier F, de Renty P, Peytavin G, et al: Maternal-fetal transfer of saquinavir studied in the ex vivo placental perfusion model. Am J Obstet Gynecol 185:178, 2001
30. Public Health Service Task Force: Recommendations for use of antiretroviral drugs inpregnant HIV-1–infected women for maternal health and intervention to reduce perinatal HIV-1transmission in the United States. February 4, 2002
31. Gilstrap LC, Bawdon RE, Roberts SW, et al: The transfer of the synthetic nucleoside analog ganciclovir across the perfused human placenta. Am J Obstet Gynecol 170:967, 1994
35. Miceli JN, Olson WA, Cohen SN: Elimination kinetics of isoniazid in the newborn infant. Dev Pharmacol Ther 2:235, 1981
37. McGee K, Bawdon RE: The ex vivo placental transfer of rifampin and rifabutin. J Infect Dis Obstet Gynecol 4:319, 1996