Designing Pregnancy Centered Medications: Drugs Which C. Gedeon and G. Koren

ARTICLE IN PRESS
Placenta (2005), jj, jjjejjj
doi:10.1016/j.placenta.2005.09.001
Designing Pregnancy Centered Medications: Drugs Which
Do Not Cross the Human Placenta*
C. Gedeon1 and G. Koren2,*
Division of Clinical Pharmacology Toxicology, Hospital for Sick Children and University of Toronto, Canada
Paper accepted 10 September 2005
This review considers and evaluates the role of placental transporters (multidrug resistance proteins, P-glycoprotein and breast
cancer resistance protein) in the uptake and efflux of drugs used in pregnancy. The effect of placental transporters in effluxing
drugs such as glyburide and numerous protease inhibitors from the fetal circulation offers the potential to manipulate the passage
of drugs across the placenta. The discovery of the interactions of these drugs with placental transporters may provide a novel
framework for future drug development in which medications can be designed to control the degree of fetal exposure and thus
prevent fetal risk.
Placenta (2005), jj, jjjejjj
Ó 2005 Elsevier Ltd. All rights reserved.
Keywords: Human placenta; ABC transporters; Glyburide; Protease and reverse transcriptase inhibitors
A large number of pregnant women suffer from medical
conditions that require ongoing or episodic drug treatment
such as asthma, epilepsy or hypertension. Moreover, pregnancy
can induce conditions such as nausea and vomiting which may
need to be treated. Pharmacologically, while the woman’s well
being is at the heart of any medical treatment, placental transfer
of drugs leading to potential toxicity to the fetus is a major
concern in the pharmacological management of the pregnant
patient. Hence, when managing a pregnant patient with medication, the treatment of 2 individual patients, mother and fetus
should be considered independently, and the decision must be
based on the risk/benefit assessment of both. While the use of
prescription drugs has sometimes increased fetal risk of teratogenicity, some medical conditions such as gestational diabetes,
hyperthyroidism or hypertension may require drug therapy in
order to ensure optimal health of the mother and fetus.
Of the thousands of available drugs, relatively few have been
shown to adversely affect the fetus. For example, the use of
most traditional anti-convulsants, while necessary for the treatment of mother, may cause structural defects as well as impaired neurocognitive development [27,28]. Thus, a potential
maternalefetal conflict brings to light the need to identify
*
Supported by a grant by the Canadian Institute for Health research.
* Corresponding author. Hospital for Sick Children, 555 University
Ave., Toronto, Ontario M5G 1X8, Canada. Tel.: þ1 416 813 5180.
E-mail address: [email protected] (G. Koren).
1
C. Gedeon is supported by a studentship by the Ontario Graduate
Scholarship.
2
G. Koren holds the Ivey Chair in Molecular Toxicology at the University of Western Ontario.
0143e4004/$esee front matter
drugs that can effectively treat the mother without adversely
affecting the fetus.
Placental passage of a drug is a function of multiple factors
such as protein binding, lipid solubility and ionization constant
(pKa), but fetal exposure to drugs also depends on maternal
pharmacokinetics including the volume of distribution, the
rate of metabolism and excretion by the placenta, the pH difference between maternal and fetal fluids, and the effect of haemodynamic changes in the mother during pregnancy (Table 1).
Ideally, one may wish to identify drugs that will not cross the
placental barrier. However, with the exception of drugs with
large molecular weights, such as heparin or insulin, most drugs
appear to cross the placenta and are associated with varying degrees of fetal exposure.
Recently we have been witnessing a pharmacological breakthrough: glibenclamide (GlyburideÒ), a drug used to treat gestational diabetes, does not appear to demonstrate significant
maternal to fetal transfer. In vitro, studies by Elliott et al. [13]
have demonstrated negligible levels of glibenclamide in the fetal
compartment even when maternal concentrations were 8 times
greater than therapeutic concentrations. Glibenclamide’s poor
maternal to fetal transfer may be due to its high protein binding
(99.8%), its short elimination half-life (6 h) and importantly the
role of specific placental transporters such as P-glycoprotein
(P-gP), multidrug resistance protein 1 (MRP1), multidrug
resistance protein 2 (MRP2) or multidrug resistance protein
3 (MRP3) [24,25]. Indeed, in a recent placental perfusion study,
we have documented that glibenclamide is transferred from the
fetal to the maternal circulation against its concentration gradient. This may be the first generation of drugs on the journey
Ó 2005 Elsevier Ltd. All rights reserved.
ARTICLE IN PRESS
Placenta (2005), Vol. jj
2
Table 1. Physiological changes in pregnancy
Function
Change
Cardiac output
Tidal volume
Pulmonary blood flow
Gastric pH
Glomerular filtration rate
Renal drug elimination
Hepatic drug elimination
Clearance
Total body water
Volume of distribution
Steady state plasma concentration
Peak serum concentration
Intestinal motility
Protein binding capacity
Increased
Increased
Increased
Increased
Increased
Increased
Increase, decrease
or unchanged
Increased
Increased
Increased
Decreased
Decreased
Decreased
Decreased
raised areas denoted as maternal lobes or cotyledons. Each maternal cotyledon is associated with several fetal cotyledons allowing for maternal circulation to be nearly superimposed on
to fetal circulation. Considering the movement of materials
from the mother to the fetus, molecules are carried by the maternal bloodstream in 1 of 3 ways: in free form dissolved in
plasma, bound to carrier proteins, or bound to red blood cells.
As a molecule crosses from the maternal to the fetal bloodstream, the solutes must cross the syncytiotrophoblast either
by passing through the cytoplasm of the trophoblast or via a network of transporters (Figure 1). Thus, transport across the placenta involves the movement of molecules between 3
compartments: the maternal blood, the cytoplasm of the syncytiotrophoblast and the fetal blood. Solute levels in each of these
compartments will play a key role in controlling the rate by
which substances cross the placenta and are largely dependent
on the initial absorption of the drug into maternal bloodstream.
Source [21,28].
PHYSIOLOGICAL CHANGES IN PREGNANCY
towards an ideal drug specifically designed for pregnancy. The
objective of this conceptual paper is to identify the characteristics that render drugs capable of staying in maternal circulation
while obviating fetal exposure and risk, making them optimal
for use in pregnancy.
TRANSPLACENTAL TRANSPORT
MECHANISMS
Typically discoid, the human placenta consists of the chorionic
villi having attendant blood vessels and connective tissues.
Viewed from the basal surface, the placenta displays slightly
Maternal
compartment
D
P
B
R
U
S
H
B
O
R
D
E
R
During pregnancy, the maternal gastrointestinal absorption of
drugs may be altered because of changes in gastric secretion
and motility. Changes in gastric pH influence the degree of ionization and solubility of many drugs, where their absorption rate
is modified thereby altering drug bioavailability. Once absorbed, maternal drug metabolism may be altered due to elevation of endogenous hormones such as progesterone. This can
stimulate the hepatic microsomal oxidase system where elevated
rates of hepatic metabolism may result in increased transformation of drugs such as phenytoin [27,28]. Conversely, theophyline and caffeine experience reduced hepatic elimination as
Fetal compartment
Facilitated or passive Diffusion
MRP 1
D
MRP2
Active transporters
RBC
MRP 3
D
BCRP
P-gP
B
A
S
O
L
A
T
E
R
A
L
Figure 1. Transport across the placental barrier.
Drug bound to protein
Drug bound to RBC
Free drug
ARTICLE IN PRESS
Gedeon, Koren: Designing Pregnancy Centred Medications
3
a consequence of elevated estradiol, which may inhibit microsomal drug metabolizing enzymes. Therapeutic concentrations
of the active drug may also be affected by haemodynamic
changes that take place throughout pregnancy. Cardiac output
and blood volume increase by 40% primarily due to an increase
in plasma volume [17]. Total body water also increases by
5e8%, due to the expansion of the extracellular fluid space
and the growth of new tissue [27,28]. Body water also accumulates in the fetus, placenta and amniotic fluid, which collectively
contributes to an increase in volume of distribution and may
lower the concentration of drugs and increase their elimination
half-life (Table 2).
PROTEIN BINDING IN PREGNANCY
Protein binding alterations may also occur in pregnancy as a result of changes in the concentration of specific proteins as well
as changes in protein binding affinity. Decreases in maternal
serum albumin may lead to corresponding increases in the
free fraction of drug [4,7]. Increased levels of free fatty acids
and total lipids together with the hormonal changes in pregnancy can also decrease the binding capacity of drugs to protein. This may have important implications to maternalefetal
drug transfer, since binding changes may influence the maternal plasma concentration of free drug available to partition
across the placental barrier. Furthermore, there is a lower concentration of specific binding proteins and altered binding affinity of drugs in the fetus: albumin concentration in the cord
blood of neonates is markedly lower than the maternal levels
[22]. In addition, there is a 3 fold lower level of a1-acid glycoprotein in the fetus compared to the mother [17]. Should free
drug cross the placenta, the fetus will experience larger
concentrations of the free drug, due to its diminished protein
binding capacity.
also influence the passage of the drug across the placenta. In
general, uncharged, un-ionized molecules with high lipid solubility and lower molecular weight penetrate the cell membranes more readily than hydrophillic ionized drug
molecules [21].
The transfer of weak acids and bases is influenced by the pH
gradient between the maternal and fetal circulations and by the
pKa of the drug. Since fetal plasma is slightly more acidic, 0.1
lower than maternal plasma pH, for weak bases, the un-ionized
free drug crossing the placenta becomes ionized and is
‘trapped’ in the more acidic fetal circulation [22,27]. The pH
of the amniotic fluid is also lower than that of maternal plasma.
Therefore, the more acidic environment of both fetal blood
and amniotic fluid favor ionization and accumulation of basic
drug due to ‘‘ion trapping’’ often resulting in fetal drug concentrations that may exceed maternal plasma concentrations
and lead to toxicity.
Finally, at the placental cytoplasmic barrier, free drug can
cross either by simple or facilitated diffusion or by active transport. Diffusion, simple or facilitated, depends upon the transmembrane concentration gradient, where it moves from
a higher concentration compartment to an area of lower concentration. Active transport depends not only on solute concentration but also on a steady supply of ATP. In addition,
the placenta itself contributes to maternalefetal drug transfer
since it possesses metabolizing enzymes capable of oxidation,
reduction, hydrolysis and conjugation which can potentially
convert an inactive drug into an active metabolite and vice
versa.
Clearly, the transfer of drugs across the placental membrane
will vary, depending on the apparent volume of distribution,
the degree of protein binding, acidebase equilibrium, metabolic and excretory mechanisms in the placenta and fetus as
well as haemodynamic alterations.
GESTATIONAL pH CHANGES
GLUCOSE, HYPERGLYCEMIA AND
GESTATIONAL DIABETES
Lipid solubility, the pH of the maternal and fetal fluids, the
ionization constant of the drug (pKa) and the molecular weight
Many active transporters mediate the absorption, metabolism
and excretion of drugs in the mammalian body, and are largely
Table 2. Pregnancy induced pharmacokinetic changes for selected drugs
T½
Drugs
Ampicillin
Cefuroxime
Imipenem
Piperacillin
Azlocillin
Nifedipine
Labetolol
Sotalol
Phenytoin
Vd (l)
CL (ml/min)
Pregnant
Non-pregnant (min)
Pregnant
Non-pregnant (min)
Pregnant
Non-pregnant
54.2 3.9
44 5
36 8
46.5 10
65.4
81 18
102 16
396 36
900 314
69.6 6.1
58 8
41 16
53.7 4.6
72
360
160
558 42
32.8 2.5
17.8 1.9
47.1 14.8
67.6 11.8
15.4
34.5 2.7
16.3 2.1
18.9 5.8
41.9 6.2
24.7
450 31
282 34
973 47
1538 362
126.1
266 105
1704 531
196 24
370 30
198 27
338 85
540 75
195.7
27 15
1430
109 7
Source: [32,34e37].
106.4 8.1
87.3 7.2
ARTICLE IN PRESS
4
Placenta (2005), Vol. jj
GLIBENCLAMIDE: BREAKTHROUGH
TREATMENT FOR GESTATIONAL DIABETES
dependent on ATP, rendering glucose and its metabolism of
particular importance to transplacental transport. In particular,
glucose uptake and transfer show a marked dependence on maternal plasma glucose levels. Forty percent of total glucose taken up by the placenta is subsequently transferred to the fetal
circulation [18]. Particularly, insulin is an important regulator
of membrane glucose transport; its action on responsive cells
alters the distribution of placental glucose transporters and
may activate inactive ones. Dually perfused human placental
lobule experiments found that maternal glucose levels exhibit
a marked effect upon both glucose placental transfer and consumption. Specifically, when maternal glucose levels are above
normal there is a corresponding increase in fetal blood glucose
concentrations due to the increased supply transferred from
the mother.
Since maternal blood glucose concentration influences fetal
glucose levels, both placental transport and consumption of
glucose increase in response to maternal hyperglycemia [18].
This is the hallmark of gestational diabetes. The response
to hyperglycemia is often 2 fold. Firstly, glucose is stored
as glycogen by the placenta. Although glycogen is present
in the placenta throughout gestation, the trophoblast does
not contain the rate limiting enzyme, glucose-6-phosphatase,
needed to convert glycogen to glucose [1]. As well, the increased glucose supply leads to an increase in its consumption
[11]. Elevated levels of insulin in the fetus become necessary
to maintain homeostatic blood glucose concentration in lieu of
increased glucose supply. Thus, maternal hyperglycemia will
eventually lead to fetal hyperinsulimia, over nutrition and ultimately macrosomia in many unborn children.
Despite being pharmacokinetically unaltered during pregnancy
[25,26], insulin still presents a risk of transplacental transfer,
and more importantly, non-compliance. Sulfonylureas, oral alternatives to insulin, are a class of drugs that appear to act by
inhibiting potassium efflux via ATP dependent potassium
channels in pancreatic b cells. This action leads to cellular depolarization and calcium-stimulated release of insulin in the
pancreas. In other words, they enhance the release of endogenous insulin.
Glibenclamide is a second generation sulfonylurea that has
been shown not to cross the human placenta. A randomized
control trial of 404 pregnant women suffering from gestational
diabetes conducted by Langer et al. [26,27] found no detectable levels of glibenclamide in cord serum. Glibenclamide’s exceptionally high protein binding, above 99.8%, allows for less
than 0.2% of free drug to circulate and cross the placenta. Additionally, its high protein binding is coupled to a short elimination half-life made possible by its low volume of
distribution (0.2 l/kg) and rapid clearance (1.3 ml/kg/min).
Simply put, glibenclamide has only a brief opportunity to cross
the placenta [15,25].
Moreover, unlike insulin, glibenclamide does not elicit an
immune response in either the baby or the mother. The
unique interplay of pharmacokinetic parameters, such as
high protein binding and short elimination half-life, is suspected to prevent it from crossing the placenta, and make it
suitable for treatment in pregnancy.
TREATMENT OF GESTATIONAL DIABETES
PLACENTAL ABC TRANSPORTERS
AND PREGNANCY
The ultimate goal of gestational diabetes treatment is to restore normoglycemia in the mother thereby ensuring a euglycemic environment for the fetus. The gold standards for the
management of gestational diabetes have been diet and insulin.
For the gestational diabetic (mother), insulin promotes glucose
uptake and oxidative metabolism resulting in a lowering of
blood sugar, controlling glycosuria and polysuria as well as
the disappearance of ketone bodies from the blood and urine
[10]. Insulin is the preferred pharmacological treatment in
pregnancy because it is unable to cross the placenta due to
its large molecular weight (6000 Da). However, insulin is
not without its shortcomings. Elliott et al. [12,13] observed
anti-insulin antibodies in response to human insulin in women
with gestational diabetes. Once bound to these IgG antibodies,
insulin can cross the placental barrier and initiate the cascade
of events leading to neonatal macrosomia. Moreover, most
women who suffer from gestational diabetes were not diabetic
prior to pregnancy. Adequate normoglycemia in these women
may require several daily injections; pain, discomfort as well
as the increased cost and training required to administer
these injections make compliance with the therapy a critical
issue.
The trophoblast is the interface of exchange between the maternal and fetal circulations. As a barrier, the trophoblast has
abundant expression of ABC transporters such as P-glycoprotein (P-gP) and multidrug resistance proteins 1, 2 and 3
(MRP1, MRP2, MRP3, respectively). Powered by ATP, these
transporters actively extrude substrates from the placenta [23]
(Table 3). Particularly, P-gP and MRP2 have been shown to be
expressed on the brush border (maternal-side) of the human
placental trophoblast [42], while MRP1 and MRP3 are on
the basal membrane (fetal side) [19,34]. Multidrug resistance
transporters may result in a lower cellular concentration of
drug via an efflux mechanism, thus creating pharmacological
sanctuaries. For example, MRP1 and MRP3 preferentially
transport organic anions, promote the excretion of glutathione/glucoronide metabolites and thus prevent their entry
into fetal blood.
P-gP is also an active drug transporter of the ATP binding
cassette transporter family with a wide range of substrates [19].
Abundant in the apical membrane of the placental trophoblast,
P-gP transports its substrates in an outward (extracellular) direction. Since it can be detected in placental trophoblast from
the first trimester of pregnancy [14,42] it is likely that it
ARTICLE IN PRESS
Gedeon, Koren: Designing Pregnancy Centred Medications
5
Table 3. Drug efflux transporters and substrates
Substrates
Actinomycine D
Doxorubincin
Irinotecan
Mitroxantrone
Teniposide
Vinblastine
Morphine
Dexamethasone
Beta-acetyldigoxin
Transporter
Etoposide
Daunorubcin
Mitomycin C
Paclitaxel
Topotecan
Vincristine
Loperamide
Ketoconazole
Alpha
methyldigoxin
Quinidine
Indinavir
Saquinavir
Cyclosporine A
Erythromycin
Sparfloxacin
Lovastatin
Simvastatin
Mibefradil
D617, D620
Bisanterine
Citalopram
Corticosterone
Loperamide
Paclitaxel
Prednisolone
Rifampin
(99 m)-Tctetrofosfine
P-glycoprotein
(MDR1)
Adiamycin
Glutathione
Methotrexate
Aflatoxin B1
Paclitaxel
Vincristine
Dehydroepiandrosterone
Etoposide
17-beta glucoronyl
estradiol
S-glutathionyl
Prostaglandin A2
Methotrexate
Etoposide
Mithoxantrone
Organic anions
Cisplatinum
Topotecan
Colchicine
MRP1
multidrug
resistance
protein
Methotrezate
Vinblastine
Ceftriaxone
Rifampicin
Ritonavir
Estradiol
Temocaprilate
Pravastatin
Heavy metals
Food carcinogens
Doxorubicin
Epirubicin
Phenytoin
Cisplatin
Vincristine
Grepafloxacin
Saquinavir
Indinavir
Octreotide
Furosemide
Flavonoids
Arsenite
Acetaminophen
Etoposide
Methotrexate
Glucuronide
Leukotriene C4
Teniposide
Digitosin
Amprenavir
Nelfinavir
Ritonavir
Tacrolimus
Levofloxacin
Atorvastatin
Cerivastatin
Diltiazem
Verapamil
Aldosterone
Calcein-M
Colchicin
FK 560
Ondasetron
Phenytoin
Quinidine
Terfenadine
Table 3 (continued)
Substrates
Topotecan
Doxorubicin
Danurubicin
Zidovudine
Lamivudine
Mitoxantrone
Methotrezate
MTX-glu
Adriamycin
Transporter
9 Aminocamptotecin
Irinotecan
Epirubicin
Idrubicinol
Flavopiridol
Prazocin
Indolocarbazol
Pheophorbide
Etoposide
BCRP
breast
cancer
resistance
protein
protects the fetus from amphipathic xenotoxins. Thus, fetal
tolerance to maternal concentrations of drugs which are considered to be good P-gP substrates will be higher than that
for drugs considered to be poor P-gP substrates. Thus, it
may be preferable to treat pregnant women with drugs that
are good P-gP substrates such as during antineoplastic chemotherapy where drugs, such as PaclitaxelÒ may be preferred
over other antineoplastic agents that cross the placenta more
easily.
ANTIRETROVIRALS IN PREGNANCY
PSC833
Fluorscein
MRP2
multidrug
resistance
protein
Sulphinpyrazone
Etoposide
MRP3
multidrug
resistance
protein
In contrast, there are clinical situations where it is desirable to
increase the penetration of drugs in order to treat the unborn
child. An important example is antiretroviral treatment of
HIV-infected pregnant women. Data suggest that nucleoside
reverse transcriptase inhibitors rapidly cross the placenta
(Table 4). Cord blood concentrations of ZidovudineÒ and
LamivudineÒ tend to equal maternal concentrations at the
time of delivery, whereas cord blood concentrations of didanosine and zalcitabine are approximately 50% that of maternal
concentrations [38,39]. Non-nucleotide reverse transcriptase
inhibitors, such as nevirapine, have also been shown to cross
the placenta [31,38]. Several factors may account for their
passage including the drug’s lower protein binding (60%),
low molecular weights, favorable degree of lipophilicity (log
octanol/water coefficient 1.81) [45]. As well, the majority are
not P-gP substrates. Hence, these drugs passively diffuse
across the placenta and are administered in sufficient doses
to cross the placenta and prevent maternalefetal transmission
of HIV during labor.
Conversely, protease inhibitors (PI) do not cross the placenta
to a clinically appreciable extent [30]. Nelfinavir, ritonavir,
saquinavir and lopinavir undergo incomplete transplacental
transfer [33]. Low PI placental transfer can be attributed to
high protein binding (98%) and that these drugs are substrates
for placental P-gP transporter [23]. For example, Saquinavir
is a P-gP substrate with a high molecular weight (767 g/mol),
high protein binding (98%) and partition coefficient
(octanol/water 4.1 log10) which may contribute to the small
amount that crosses the placenta [16]. Other protease inhibitors
which follow similar placental kinetics include indinavir, ritonavir, lopinavir and nelfinavir. P-gP’s relative affinity to
ARTICLE IN PRESS
Placenta (2005), Vol. jj
6
Table 4. Pharmacokinetic changes in HIV drugs during pregnancy
Drug class
Drug
Nucleoside reverse Zidovudine
transcriptase
inhibitors
Lamivudine
Didanosine
Stavudine
Half-life
No change
(1.1 h)
No change
(6 h)
No change
(1.5 h)
No change
Protein
binding (%) Bioavailability
Clearance
<25
Increases
No change
10e50
Increases
Non-nucleoside
Nevirapine
reverse transcriptase
inhibitors
Unchanged
Protease
inhibitors
Unchanged (3e5 h)
98
Lopinavir
Unchanged (5e6 h)
Amprenavir Unchanged (6 h)
Saquinavir
Hydroxicarbamide: not recommended in pregnancy
99
98
98
Nelfinavir
Ritonavir
Prone to induction
60
Placental transfer
63%
Cross by diffusion
55%
Cross by diffusion
45%
Cross by diffusion
86%
Cross by diffusion
90%
Cross by diffusion
Neural tube defect: not used in pregnancy
Not detectable in cord blood
5.3% maternal [ ] detected in
cord blood
6e8% of maternal [ ] in cord blood
Not detectable in cord blood
Source: [29,31,33,39e42].
different PI drugs varies 3e4 fold [24]. Nelfinavir has become
a commonly used protease inhibitor during pregnancy because
of its tolerability and potency; however, it has also been demonstrated to exhibit a more rapid development of viral resistance
and less durable suppression of viral replication leading to
a greater likelihood of HIV transmission to the fetus.
Saquinavir monotherapy is not adequate and often requires
‘‘boosting’’ with low dose ritonavir. Ritonavir is rarely used
in pregnancy owing to its frequent and severe gastrointestinal
adverse effects [30]. Only 5.3% of maternal plasma
concentrations of ritonavir are detected in fetal cord blood.
Ritonavir’s well documented interaction and perhaps inhibition
of P-gP and MRP1 may account for its exclusion from the placenta [3]. However, ritonavir’s inhibition of these transporters
is disadvantageous since it gives rise to drug exclusions, altered
bioavailability and changes in drug distribution resulting in
a decreased efficacy of treatment with time. This requires the
initiation of combination therapies to achieve clinically beneficial and sustained viral suppression. Consequently, poly-therapy is usually accompanied by decreased adherence. However, if
certain PI are effluxed by the placenta, and not reaching the fetus, future use of the appropriate ABC inhibitor may allow for
greater concentrations in the fetal compartment and the prevention of maternalefetal HIV transmission.
As the use of combination antiretroviral regimens becomes
increasingly common among HIV-infected pregnant women
and the rate of transmission decreases, the safety, toxicity
and teratogenicity of these agents become paramount. For example, a common toxicity with zidovudine is bone marrow
suppression, and decreases in hemoglobin in infants exposed
to zidovudine at 3 weeks of age [9]. As a class, nucleoside reverse transcriptase inhibitors have been suggested to cause fetal hepatic microsomal damage in infants exposed perinatally
[5]. Additionally, combination therapy has been suggested to
trigger glucose intolerance. In particular, zidovudine in combination with lamivudine and nelfinavir fosters a reduction in
neonatal insulin similar to that observed in Type I diabetes
[9]. These data may indicate an early damaging effect on fetal
pancreas such as inhibition of proinsulin conversion to insulin
due to the activity of these protease inhibitors. Cumulative
damaging effects on pancreatic b cells may culminate in the
death of these cells and consequent insulinopenia followed
by clinical diabetes.
GLIBENCLAMIDE’S INTERACTION WITH
THE ABC-TRANSPORTER FAMILY
Recently, we have shown, using the placental perfusion model,
that glibenclamide is actively transferred from the fetal to the
maternal circulations against a concentration gradient. Overall,
there was a net decrease (63%) of glibenclamide in fetal to maternal concentration ratios despite equal initial concentrations
(200 ng/ml) in both maternal and fetal circulation [26]. Undoubtedly, placental active transporters such as the ABCtransporter family should be considered in the movement of
glibenclamide from fetal to maternal circulation.
From an in vitro perspective, glibenclamide is well known to
interact with P-gP. However, the emerging thought is that glibenclamide may represent a general inhibitor for ABC transporters, both P-gP and MRP’s alike and other placental
ARTICLE IN PRESS
Gedeon, Koren: Designing Pregnancy Centred Medications
homologue, in that it may bind to some conserved motif [19].
Yet, the belief that glibenclamide is simply an inhibitor of these
transporters is contradictory to in vivo findings. Elliott et al.’s
[13] dually perfused human placental model showed virtually
no appearance of glibenclamide in the fetal circulation. Even
when maternal concentrations were 8 fold greater than therapeutic peak levels (1000 ng/ml), a relatively large concentration
gradient facilitating the possibility for diffusion, only minimal
concentrations of glibenclamide were seen in fetal blood.
Diffusion, simple or facilitated, cannot explain the negligible presence of glibenclamide in the fetal compartment. As
a substrate of certain members of the ABC-transporter family,
glibenclamide may be actively pumped back to maternal circulation, and can both maintain maternal steady state concentrations and normoglycemia while decreasing fetal exposure and
risk of hypoglycemia. Perhaps the possibility should be considered that glibenclamide is not an inhibitor of the ABC transporters but rather a substrate.
GLIBENCLAMIDE: THE PROTOTYPE OF
AN OPTIMAL DRUG FOR PREGNANCY
The choice of therapy in pregnancy is a culmination of many
factors: efficacy of the drug in treating the maternal condition,
possibility of transplacental passage, risk of teratogenicity and
abnormality due to this transplacental passage as well as factors
influencing tolerability and adherence. In pregnancy, tolerability may be even more important especially if therapy exacerbates common complications of pregnancy such as morning
sickness or glucose intolerance. Glibenclamide appears to fulfill all these characteristics. While it is well tolerated and fails
REFERENCES
[1] Agarwal M, Punnose J. Recent advances in the treatment of gestational
diabetes. Expert Opin Investig Drugs 2004 Sept;13(9):1103e11.
[3] Bawdon RE. The ex vivo human placental transfer of the anti-HIV antinucleoside inhibitor abacabir and the protease inhibitor amprenavir. Infect Dis Obstet Gynecol 1998;6(6):244e6.
[4] Beck F. Comparative placental morphology and function. Dev Toxicol
1981:35e54.
[5] Blanche S, Tardieu M, Rustin P, Slama A, Barret B, Firtion G. Persistent mitochondrial dysfunction and perinatal exposure to antiretroviral
nucleoside analogues. Lancet 1999;354:1084e99.
[7] Boyd J, Hamilton W. The human placenta. 1st ed. Cambridge: Heffer; 1970.
[9] Connor E, Sperling R, Gelber R. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment.
N Engl J Med 1994;331:1173e80.
[10] Conway D, Gonzales O, Skiver D. Use of glibenclamide for the treatment
of gestational diabetes: San Antonio experience. J Matern Fetal Neonatal
Med 2004;15:51e5.
[11] De Lange E. Potential role of ABC transporter as a detoxification system
at the bloodeCSF barrier. Adv Drug Deliv Rev 2004;56:1793e809.
[12] Elliott B, Schenker S, Langer O, Jonhson R, Prihoda T. Comparative
placental transport of oral hypoglycemic agents in humans: a model
of human placental drug transfer. Am J Obstet Gynecol 1994;171:
653e60.
[13] Elliott B, Langer O, Schenker S, Johnson RF. Insignificant transfer of
glyburide occurs across the human placenta. Am J Obstet Gynecol
1991 Oct;165(4 Pt 1):807e12.
7
to illicit an immune response it also effectively treats maternal
hyperglycemia without the risks to the fetus associated with
transplacental passage.
LIMITATION AND CHALLENGES
The present conceptual paper is based on knowledge existing
to date. It is important to acknowledge that data of most human placental transport stems from studies in term placenta
and that our current knowledge of ABC transporters in the
first trimester, during embryopathy, is very limited. Moreover,
even in human term placenta, the understanding of the activity
of different ABC transporters is very preliminary.
In summary, for pregnancy, future drugs should be sought
which can effectively treat the maternal condition, all the while
minimizing fetal exposure. Ideally, the drug should possess
a high protein binding, a short elimination half-life and a small
volume of distribution. Identification of drugs which interact
with placental efflux transporters will allow the possibility of
minimizing fetal exposure (e.g. glibenclamide) in the treatment
of maternal conditions. Conversely, these transporters also allow the possibility for treating fetal conditions (e.g. protease
inhibitors) through their inhibition and modulation. As for future research, the endeavor of understanding the exact protein
motif to which drugs such as glibenclamide interact with placental transporters should include techniques such as vector
construction and site-directed mutagenesis. Replicating the
binding sites and incorporating them into already existing
drugs used in pregnancy, allows for the possibility of altering
transporter binding affinity. Ultimately, this may allow for
drugs specifically designed for pregnancy to come into being.
[14] Faber J, Thornburg K. Placental physiology. 1st ed. New York: Raven
Press; 1983.
[15] Feig D, Kraemer J, Klein J, Koren G. Transfer of glyburide into breast
milk [abstract]. Clin Pharmacol Ther 2004;75:28.
[16] Forestier F, De Renty P, Peytavin G, Dohin E, Farinotti R, Mandelbrot L.
Maternalefetal transfer of saquinavir studied in the ex vivo placental perfusion model. Am J Obstet Gynecol 2001;185(1):178e81.
[17] Bournissen G, Feig D, Koren G. Maternalefetal transport of hypoglycemic drugs. Clin Pharmacokinet 2003;42(4):303e13.
[18] Gluek C, Goldenberg N, Streicher P, Wang P. The contentious nature of
gestational diabetes: diet, insulin, glyburide and metformin. Expert Opin
Pharmacother 2002 Nov;3(11):1557e68.
[19] Gottesman MM, Pastan I, Ambudkar S. P-glycoprotein and multidrug
resistance. Curr Opin Genet Dev 1996;6:610e7.
[21] Heikkila A, Renkonen OV, Erkkola R. Pharmacokinetics and placental
passage of imipenem during pregnancy. Antimicrob Agents Chemother
1992;36:2652e5.
[22] Holcberg G, Tsadkin Tamir M, Sapir O, Huleihel M, Mazor M, Ben
Zvi Z. New aspects in placental drug transfer. Placental Drug Transfer
2003;5:873e6.
[23] Huisman M, Smit J, Schinkel A. Significance of P-glycoprotein for the
pharmacology and clinical use of protease inhibitors. AIDS 2000;14:
237e42.
[24] Jones K, Hoggard PG, Sales SK, Khoo S, Davey R, Back DJ. Differences
in the intracellular accumulation of HIV protease inhibitors in vitro and
the effect of active transport. AIDS 2001;15:675e81.
[25] Koren G. Glyburide and fetal safety; trans-placental pharmacokinetic
considerations. Reprod Toxicol 2001;15:227e9.
ARTICLE IN PRESS
Placenta (2005), Vol. jj
8
[26] Kramer J, Feig D, Klein J, Koren G. The transport of glyburide in the
human placenta: implication for treatment of gestational diabetes [submitted for publication, August, 2005].
[27] Langer O, Conway D, Berkus M, Xenakis E, Gonzales O. A comparison
of glyburide in insulin in women with gestational diabetes mellitus.
N Engl J Med 2000 Oct;343(16):1134e8.
[28] Lobstein R, Lalkin A, Koren G. Pharmacokinetic changes during pregnancy and their clinical relevance. Maternal fetal toxicology a clinician
guide, 3rd ed. Marcel Dekker Inc.; 2001. p. 1e21.
[29] Marsolini C, Rudin C, Decosterd LA, Telenti A, Schreyer A, Biollaz J,
et al. Swiss mother þ child HIV cohort study. Transplacental passage
of protease inhibitors at delivery. AIDS 2002;16(6):889e93.
[30] Maarten T, Huisman J, Smit J, Wiltshire H, Richard M, Hotelmans J, et al.
P-glycoprotein limits oral available brain and fetal penetration of saquinavir
even with high doses of ritonavir. Mol Pharmacol 2001;59(4):806e13.
[31] Mirochnick M. Antiretroviral pharmacology in pregnant women and
their newborns. Ann N Y Acad Sci 2000;918:287e97.
[32] O’Hara MF, Leahey W, Murraghan GA. Pharmacokinetics of sotalol
during pregnancy. Eur J Clin Pharmacol 1983;24(4):521e4.
[33] Owen A, Chandler B, Back DJ. The implication of P-glycoprotein in
HIV: friend or foe? Fundam Clin Pharmacol 2005:283e96.
[34] Pacifi GM, Nottoli R. Placental transfer of drugs administrated to the
mother. Clin Pharmacokinet 1995;28:235e69.
[35] Page K. The physiology of the human placenta. University College
London: UCL Press Limited; 1993.
[36] Perucca E, Richens A, Ruprah M. Serum protein binding of phenytoin in
pregnant women. Proc Br Pharmacol Soc 1981;11:409e10.
[37] Philipson A, Stiernstede G. Pharmacokinetics of cefuroxime in pregnancy.
Am J Obstet Gynecol 1982;142:823e8.
[38] Rogers RC, Sibai BM, Whybrew WD. Labetolo pharmacokinetics in
pregnancy induced hypertension. Am J Obstet Gynecol 1990;162(2):
362e6.
[39] Sanberg JA, Slikker W. Developmental pharmacology and toxicology of
anti-HIV therapeutic agents: dideozynuclosides. FASEB J 1995;9:
1157e63.
[40] Shelley H. Transfer of carbohydrate. Placental transfer,Tunbridge Wells:
Pitman Medical; 1979. p. 118e41.
[41] Sibley C, Boyd R. Control of transfer across the mature placenta. Oxford
reviews of reproductive biology, Oxford: Oxford University Press; 1988.
p. 382e435.
[42] Smit JW, Huisman MT, Cantellingen O, Wiltshire HR, Schinkel AH.
Absence or pharmacological blocking of placental P-glycoprotein
profoundly increases fetal drug exposure. J Clin Invest 1999;104:
1441e7.
[45] Voit R, Schroder S, Peiker G. Pharmacokinetics studies of
azlocillin and piperacillin during late pregnancy. Chemotherapy 1985;
31:417e24.
FURTHER READING
[2] Barton JR, Prevost RR, Wilson DA. Nefedipine pharmacokinetics and
pharmacodynamics during immediate postpartum period in patients
with pre-eclampsia. Am J Obstet Gynecol 1991;165:951e4.
[6] Bloomgardern A. New therapeutic approaches to non-insulin dependent
diabetes mellitus. Endocr Pract 1997;3(5):307e12.
[8] Chamberlain A, White S, Bawdon R. Pharmacokinetics of ampicillin and
salmactam in pregnancy. Am J Obstet Gynecol 1993;168:667e73.
[20] Hay W. Placental control of fetal metabolism. R Coll Obstet Gynecol
1989:33e52.
[43] St-Pierre M, Serrano M, Marcias R. Expression of members of the multidrug resistance protein family in human placenta. Am J Physiol Regul
Integr Comp Physiol 2000;27:R1495e503.
[44] Tamas G, Kerenyi Z. Current controversies in the mechanisms and
treatment of gestational diabetes. Curr Diabetes Rep 2002 Aug;2(4):
337e46.
[46] Yazdanian M. Bloodebrain barrier properties of human immunodeficiency
virus antiretrovirals. J Pharm Sci 1999;88:950e4.
`