Prostaglandins in Pregnancy

The chemical name for arachidonic acid is eicosatetraenoic acid; therefore, the metabolites of arachidonate often are referred to as “eicosanoids.” Arachidonate is metabolized by two major enzymatic pathways: cyclooxygenase and lipoxygenase (Fig. 1).^{1, 2} Metabolism by the cyclooxygenase pathway leads to the prostaglandins (PGs) and thromboxanes (TXs); metabolism by the lipoxygenase pathway leads to the hydroperoxyeicosatetraenoic acids (HPETEs), hydroxyeicosatetraenoic acids (HETEs), lipoxins, and leukotrienes (LTs). The HPETEs are hydroperoxidases of arachidonic acid, and the HETEs and LTB₄ are hydroxylated forms. The leukotrienes C₄, D₄, and E₄ are modifications of arachidonate that contain an amino acid attached to the fatty acid chain by a sulfide bond. Thus, they often are referred to as cysteine leukotrienes. The cysteine-containing leukotrienes are the compounds that were previously referred to as the “slow-reacting substance of anaphylaxis” (SRS-A).

There are two isoforms of the cyclooxygenase enzyme: COX-1, which is the constitutive form, and COX-2, which is the inducible form. Both isoforms convert free arachidonic acid into PGG₂ and PGH₂. There are several lipoxygenase enzymes. The 5-, 12- and 15-lipoxygenase enzymes lead to biologically active compounds. Arachidonic acid first is converted to the HPETEs and then to the HETEs, lipoxins, or leukotrienes.^{1, 2} The 5-lipoxygenase is the primary pathway to the leukotrienes. The lipoxins are formed by the actions of 5- and 15-lipoxygenases on arachidonic acid. Competition for arachidonic acid between the cyclooxygenase and lipoxygenase enzymes exists in cells and tissues containing both pathways.

Essentially, all cells produce eicosanoids. In general, they are considered intracellular or local mediators of physiologic responses, rather than circulating hormones. They exert their effects by autocrine mechanisms (within the same cell in which they are produced) or paracrine mechanisms (between cells or tissues in which they are produced). The fact that they exert their effects locally can complicate interpretation of results from studies evaluating blood levels or studies in which they are administered exogenously.

Several investigators have attempted to measure circulating levels of prostaglandins during human pregnancy. Reported concentrations range from 3–10 pg/mL to several hundred picograms per milliliter of plasma, and there is little agreement among investigators as to the absolute plasma concentrations for any one prostaglandin. There are probably at least two reasons for the lack of agreement. First, solvent extraction, purification, and radioimmunoassays for plasma prostaglandins can be problematic, often resulting in nonspecifically high assay values for “blank” samples that do not contain any prostaglandin.³ Second, prostaglandins, like other arachidonic acid metabolites, function primarily as intracellular mediators of physiologic responses rather than as true circulating hormones. The prostaglandins that do leak into the circulation are metabolized rapidly by the lungs, liver, and kidneys.

The information available for plasma levels of PGE₂, PGF_2α, and their metabolites, 13, 14-dihydro-15-keto-PGE₂ and 13,14-dihydro-15-keto-PGF_2α, does not indicate convincing gestationally related changes except during parturition, when the plasma levels of PGF_2α metabolite increase.^{4, 5, 6, 7, 8, 9, 10} Plasma concentrations of prostacyclin's stable metabolite, 6-keto-PGF_1α, are reportedly higher during late pregnancy than during early pregnancy or the nonpregnant state,^{11, 12} and they increase further with labor.^{12, 13} However, there is considerable controversy as to whether prostacyclin and 6-keto-PGF_1α are circulating hormones. Early reports of plasma 6-keto-PGF_1α concentrations were in the range of 100–300 pg/mL, and it was believed that the lung acted as an endocrine organ, releasing prostacyclin into the systemic circulation.^{14, 15} Measurements with gas chromatography/mass spectrometry or with properly validated radioimmunoassays indicate that plasma prostacyclin levels are only approximately 2–40 pg/mL. Several references on both sides of this issue can be found in reviews.^{16, 17} It is probably more meaningful to determine and evaluate production rates from specific tissues rather than circulating levels when assessing the involvement of prostaglandins in physiologic functions during human pregnancy.

Circulating concentrations of thromboxane's stable metabolite, TXB₂, also have been determined in human pregnancy. Plasma concentrations range from 100 to 300 pg/mL, and serum concentrations may reach 300 ng/mL.^{18, 19} The reason serum concentrations are so much higher than plasma concentrations is because platelets release thromboxane during the clotting process. Thromboxane levels are higher in pregnant women than in nonpregnant women, and they increase further with the onset of labor.

Maternal cardiovascular changes

Several cardiovascular changes typically occur during human pregnancy.^{20, 21, 22} Heart rate increases by 10–15%, cardiac output increases by 30–50%, and blood volume increases by up to 48%. Blood pressure changes are small in normotensive women, but generally there is a slight decrease of 2–3 mmHg in systolic pressure and 5–10 mmHg in diastolic pressure. Normal blood pressures during the third trimester average 95/50 mmHg in the lateral position and 106/65 mmHg in the supine position. Women with essential hypertension actually may show a significant decrease in blood pressure during the first two trimesters. These facts indicate that normal pregnancy is associated with an increase in the production of vasodilatory agents.

PROSTAGLANDINS

Prostaglandins are involved in the vasodilatation of pregnancy. Everett and colleagues²³ showed that the dose of angiotensin II (AII) necessary to cause a 20-mmHg increase in diastolic blood pressure in pregnant women was significantly reduced in women treated with indomethacin. Thus, the decreased sensitivity to the vasoconstrictive effects of AII characteristic of healthy pregnant women was altered by prostaglandin synthesis inhibition to a state of increased sensitivity to vasoconstriction.

The prostaglandin most likely responsible for increased vasodilatation in normal pregnancy is prostacyclin (PGI₂) because of its potent effect to relax the smooth muscle of blood vessels and lower systemic arterial blood pressure.^{24, 25, 26} Prostacyclin's actions on vascular tone are most likely exerted locally in the blood vessel wall through a paracrine mechanism between the endothelial cells, in which prostacyclin is primarily produced, and the vascular smooth muscle. Leukocytes also can form prostacyclin.²⁷

If prostacyclin is not a circulating hormone, there must be some other compound that circulates to stimulate prostacyclin synthesis locally in blood vessels. The identity of such a compound is not known, but it seems reasonable to assume that the placenta is the source of a circulating vasodilatory agent. This idea is supported by the facts that blood pressure typically decreases during pregnancy and that simply turning a pregnant patient from the lateral to the supine position causes an increase in blood pressure despite a decrease in cardiac output.^{21, 22} Women in their third trimester (28–32 weeks' gestation) in whom preeclampsia is destined to develop have a greater increase in blood pressure than do healthy pregnant women when turned from the lateral to the supine position.²⁸ One explanation for the increase in blood pressure in the supine position is that the gravid uterus compresses the vena cava and thereby reduces venous return to the heart, leading to decreased cardiac output. This would activate the aortic and carotid sinus baroreceptors to cause reflex neural vasoconstriction. An alternative explanation is that compression of the vena cava by the gravid uterus in the supine position restricts the amount of placentally derived vasodilatory hormone reaching the maternal systemic arterial blood vessels.

THROMBOXANE

The primary source of thromboxane in the adult circulation is the platelets, although leukocytes also produce it. Thromboxane is a potent stimulator of platelet aggregation and vasoconstriction,^{24, 25, 26} and as such, it has been implicated in several pathologic conditions that are associated with thrombosis and vascular congestion,^{25, 29} including preeclampsia, as discussed later in this chapter.

Under normal conditions, prostacyclin released by the vascular endothelial cells acts to stimulate an increase in adenosine 3',5'-cyclic phosphate (cAMP) levels within the platelets, thus inhibiting aggregation.³⁰ Within the platelets, thromboxane can prevent prostacyclin-induced increases in cAMP, so under typical circumstances, thromboxane and prostacyclin counteract each other's actions to maintain homeostasis with respect to platelet function. If, however, the endothelial cells are damaged (e.g. injury, atherosclerosis) or prostacyclin synthesis is impaired, then the actions of thromboxane to produce platelet aggregation and vasoconstriction are manifested. This mechanism serves to protect the body at local sites of injury, but in pathologic states in which thromboxane is allowed to exert its effects throughout an organ or systemically throughout the body, this mechanism can be harmful.

During normal pregnancy, the production of thromboxane is increased. Maternal plasma levels of its stable metabolite, TXB₂, as well as its urinary metabolites, are higher during late pregnancy than during midpregnancy or the nonpregnant state.^{18, 19, 31} The placenta, as well as platelets, is an important source of thromboxane during pregnancy (Fig. 2). Placentas obtained from healthy pregnant women produce approximately 92 μg of thromboxane daily in vitro,³² and trophoblastic cell cultures, uncontaminated with platelets or other blood cells, produce 50–100 pg of thromboxane per microgram of protein daily.³³ After delivery of the placenta, maternal urinary metabolites of thromboxane decrease significantly,³¹ which is a further indication of the placenta as a source of thromboxane in pregnancy.

LIPOXYGENASE COMPOUNDS

Arachidonic acid is metabolized by lipoxygenase enzymes in several tissues, organs, and cells of the body, most notably the lungs and leukocytes (in which they were first discovered and from which the leukotrienes derive their name). The human placenta also produces lipoxygenase compounds. 5-HETE, 12-HETE, 15-HETE, LTB₄,^{16, 17, 34, 35} and LTC₄ all have been identified.

The physiologic functions of lipoxygenase metabolites during pregnancy are not known, but the HETEs, HPETEs, lipoxins, and leukotrienes do exert biologic actions, some of which must be important to placental function and pregnancy. Both 5-HETE and LTB₄ stimulate leukocytic chemotaxis and chemokinesis, increase aggregation of leukocytes, stimulate the uptake of calcium and d-glucose by neutrophils and eosinophils, increase bronchial constriction and mucus secretion, increase vasodilatation, and enhance vascular permeability.^{36, 37, 38}

LTC₄ and LTD₄ exert biologic actions in various tissues of the body as reviewed in several studies.^{2, 37, 38, 39, 40, 41} They produce contraction of a number of smooth muscle tissues, including the small airways of the lung, the ileum, the stomach, the gallbladder, and the uterus.

In the cardiovascular system, leukotrienes often produce opposite effects in different vascular beds and in different species.^{2, 40, 42} For example, LTC₄ and LTD₄ produce vasoconstriction and cause plasma leakage in guinea pig skin and the hamster cheek pouch, but they are potent vasodilators in human skin. LTC₄ and LTD₄ are potent vasoconstrictors in the coronary circulation of several species. Systemically, LTC₄ and LTD₄ lower arterial blood pressure in most adult nonpregnant species,^{2, 40, 42, 43, 44} including humans.⁴⁵ In pregnant and postpartum rhesus monkeys,^{16, 17, 46} injection of LTB₄, LTC₄, or LTD₄ (0.5 μg/kg) into the lower vena cava to mimic the systemic route of placentally produced hormones lowers both systolic and diastolic blood pressures.

The mechanisms of action through which systematically administered leukotrienes lower blood pressure are not known, but it is likely that they decrease cardiac output and stimulate prostacyclin production by endothelial cells in the lungs or locally in the systemic vasculature. LTC₄ promotes prostacyclin synthesis by human endothelial cells,^{47, 48} and both LTC₄ and LTD₄ increase prostacyclin release by bovine aortic endothelial cells.⁴⁹ Endothelial cells metabolize leukotrienes, which indicates that they have receptors for them but do not synthesize leukotrienes.⁴⁸ Although it might be argued that leukotrienes are local mediators of physiologic responses and do not function as circulating hormones, this may not be entirely true during pregnancy because of supplemental production by the placenta.

The lipoxins share some of the same biologic effects of the prostaglandins, HETEs, and leukotrienes, but chemically they are distinct molecules. Lipoxin A induces arteriolar dilation but does not affect microvascular permeability, as does LTC₄, or leukocyte adherence to venular endothelium, as does LTB₄.^{1, 2} Lipoxin A stimulates superoxide anion generation without provoking aggregation, causes chemotaxis, and possesses spasmogenic activities, as do the HETEs and leukotrienes. Lipoxin A elicits long-lasting contractions of guinea pig lung strips, but, unlike LTC₄, it does not stimulate contraction of the guinea pig ileum. Lipoxins A and B inhibit human natural killer cell activity, whereas HETEs and leukotrienes do not affect natural killer cell cytotoxicity. Furthermore, lipoxins may serve as important intracellular mediators. Lipoxin A is a potent activator of protein kinase C, even more potent than diacylglycerol or arachidonic acid. Thus, lipoxins display patterns of physiologic effects that are distinct from prostaglandins, thromboxanes, HETEs, or leukotrienes. The importance of lipoxins in normal and abnormal pregnancy is not known, but the fact that lipoxin A dilates arterioles implies that they may serve to augment vasodilatation during normal pregnancy.

Uterine blood flow

The effects of prostaglandins on uterine blood flow have been evaluated in vivo in a number of species, including pregnant monkeys and pregnant and nonpregnant chronically catheterized sheep. Radioactive microspheres and uterine artery flow probes are the two techniques most commonly used for evaluating uterine blood flow. When assessing the effects on uterine blood flow of uterotonic agents, such as the prostaglandins, their simultaneous effects on uterine contractility also must be determined. A compound that stimulates uterine contractions also reduces the amount of blood perfusing the uterus by myometrial compression of the blood vessels.

The prostaglandins that decrease uterine blood flow are PGE₂, PGF_2α, 6-keto-PGF_1α, TXB₂, and analogs of the prostaglandin endoperoxides PGG₂ and PGH₂.^{50, 51, 52, 53, 54} PGE₂ and PGF_2α are potent stimulators of uterine contractility. The effect of PGE₂ on uterine blood flow is primarily due to its uterotonic action, whereas PGF_2α exerts vasoconstrictive actions in addition to its uterine contractile effect.

The prostaglandins that increase uterine blood flow are PGI₂, PGD₂, PGA₂, PGE₁, and 6-keto-PGE₁. PGE₂ is a vasodilator in nonpregnant sheep, provided it does not stimulate uterine contractions.^{51, 52, 53, 54, 55, 56, 57, 58} PGD₂ and PGA₂ stimulate uterine contractions, whereas PGI₂ inhibits uterine activity.

Placental blood flow

IN VITRO STUDIES

Several in vitro studies have been published concerning the vasoactive effects of thromboxane and prostaglandins on human placental vessels. These studies demonstrate that thromboxane is one of the most potent vasoconstrictors of the human placental vasculature.^{59, 60, 61, 62, 63} Because the active metabolite of thromboxane (TXA₂) is extremely labile, it is necessary to use a stable analog for experimental studies. Most investigators use the thromboxane mimic U46619. In vitro studies demonstrate that thromboxane is 10- to 1000-times more active as a vasoconstrictor than is PGE₂, PGF_2α, AII, 5-hydroxytryptamine (5-HT, or serotonin), norepinephrine (NE), or bradykinin.

With respect to the vasoconstrictive effects of prostaglandins in the human placenta, there is agreement among all investigators that PGF_2α and PGE₂ are vasoconstrictors.^{60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71} PGD₂ reportedly causes vasoconstriction when given alone,^{61, 64} but attenuates vasoconstriction when given with AII.⁶¹ PGA₁ causes vasoconstriction equivalent to PGF_2α.⁷¹ The stable metabolite of thromboxane, TXB₂, is without effect on the placental vasculature.⁷¹

In the human placenta, prostacyclin consistently is the most potent vasodilator.^{61, 62, 63, 67, 68, 71} PGE₁ was found to be a weak vasodilator in two studies,^{61, 62} but a weak vasoconstrictor in another study.⁷¹ One of the stable metabolites of prostacyclin, 6-keto-PGE₁, is a weak vasodilator in the human placenta.⁶² One of prostacyclin's other stable metabolites, 6-keto-PGF_1α, is a weak vasodilator.²⁵

There is little information available regarding the vascular effects of leukotrienes in the placenta. When comparing the vasoconstrictive effect of LTB₄, LTC₄, and LTD₄ with that of thromboxane in the perfused human placental cotyledon,⁵⁹ we found that the leukotrienes cause vasoconstriction, but only at high doses. Their potency is considerably less than that of thromboxane.

Based on in vitro studies using isolated human placental vascular strips or isolated perfused human placental cotyledons, one can rank the vasoactive potencies of thromboxane, prostaglandins, and other vasoactive compounds as follows: vasoconstriction – TX (U46619) greater than PGF_2α equal to PGA₁ greater than PGE₂ greater than AII greater than 5-HT greater than NE; vasodilation – PGI₂ greater than PGE₁ greater than 6-keto-PGE₁. 

IN VIVO STUDIES

Several in vivo studies have been reported concerning the vasoactive effects of prostaglandins in the placenta. The most commonly used animal model is the chronically catheterized sheep, but other species have been used, including nonhuman primates.

Vasoconstrictors

There is agreement among investigators that fetal administration of PGE₂ or PGF_2α causes umbilical–placental vasoconstriction in the sheep.^{72, 73, 74, 75, 76} The thromboxane mimic U46619 is a potent vasoconstrictor in the ovine fetal systemic circulation and in the umbilical–placental circulation.^{77, 78} Furthermore, fetal infusion of thromboxane can result in respiratory acidosis in the fetus, a further indication of its detrimental vasoconstrictive effects on the fetoplacental vasculature. Thus, there is agreement between in vitro and in vivo studies that thromboxane, PGE₂, and PGF_2α are vasoconstrictors in the fetoplacental vascular bed.

Does prostacyclin vasodilate the placental vasculature in vivo?

IN VITRO VERSUS IN VIVO RESULTS

With respect to vasodilatation in the fetoplacental vasculature, there is controversy and disagreement between in vitro and in vivo studies. Prostacyclin is a potent vasodilator,^{24, 25, 26} and it consistently relaxes human placental blood vessels when tested in vitro, as discussed previously. However, when tested in vivo, infusion of prostacyclin into the ovine fetus decreases fetoplacental blood flow and results in an increase or no change in the calculated placental vascular resistance.^{77, 78, 79} Furthermore, fetal infusion of prostacyclin into the systemic circulation in conjunction with either thromboxane or AII reverses systemic and renal vasoconstriction induced by thromboxane or AII but further decreases fetoplacental blood flow.⁷⁷

The in vivo placental vascular results obtained for prostacyclin in the sheep are disturbing for at least two reasons. First, they contradict the consistently potent vasodilatory effect of prostacyclin obtained in human in vitro placental studies. Second, how can a compound that is such a potent vasodilator in all other vascular beds be a vasoconstrictor in the placenta, a vascular bed that must remain dilated if the fetus is to grow and mature properly?

The aberrant effects observed for prostacyclin in vivo can be explained by Poiseuille's law as it applies to the fetal–placental circulation. The placental vascular effects observed for prostacyclin in vivo (or for other vasodilators) are passive changes caused by shunting of blood away from the placenta to other vasodilated vascular beds more proximal to the heart. We have previously described this concept in detail.⁸⁰

The placental vascular actions are due to systemic alterations in blood flow and perfusion pressure in nonplacental vascular beds rather than to direct vasoconstrictive actions of prostacyclin on the placental vasculature. Indirect evidence suggests that prostacyclin is important for in vivo vasodilation in the human placenta. Doppler ultrasonographic measurements of umbilical blood flow in pregnant women are correlated positively with prostacyclin production in vitro by specimens obtained from the umbilical arteries.⁸¹


Umbilical and fetal blood flow

The arachidonic acid metabolites are present in abundance⁸² and exert potent vasoactive effects in the fetus and on the umbilical vessels. Thromboxane, PGE₂, and PGF_2α cause vasoconstriction, whereas prostacyclin causes vasodilatation of the umbilical vessels.^{73, 74, 83, 84, 85, 86} Thromboxane causes vasoconstriction of the ductus arteriosus and pulmonary vasculature, whereas prostacyclin, PGE₂, and PGE₁ cause vasodilatation.^{87, 88, 89} Inhibitors of prostaglandin synthesis, such as indomethacin and sodium salicylate (aspirin), result in vasoconstriction and closure of the ductus arteriosus,^{90, 91} demonstrating that prostaglandins are necessary to maintain patency of the ductus arteriosus during fetal life.

Thromboxane is a potent vasoconstrictor in the fetal systemic circulation, as evidenced by in vivo studies in chronically catheterized sheep fetuses. Thromboxane infusion results in a significant increase in mean fetal aortic blood pressure, renal vasoconstriction, and fetal acidosis.^{77, 78} Prostacyclin, however, vasodilates the ovine fetal systemic circulation, as evidenced by its ability to significantly lower mean fetal aortic blood pressure^{77, 78, 79} and its ability to antagonize the vasoconstrictive effects of thromboxane and AII.^{77, 78} Prostacyclin increases blood flow to the fetal kidneys,⁷⁷ adrenal glands,⁹² and lungs.^{78, 79}

The leukotrienes are potent fetal pulmonary vasoconstrictors and may be the compounds primarily responsible for maintaining the high pulmonary vascular resistance during fetal life. In newborn lambs and piglets, LTD₄ increases pulmonary and systemic vascular resistance,^{93, 94} whereas blockage of the leukotriene receptors with FPL 57231 decreases pulmonary vascular resistance, increases pulmonary blood flow, and decreases pulmonary and systemic arterial pressures in late gestational fetal lambs.⁹⁵ Human newborns with pulmonary hypertension and persistent fetal circulation have LTC₄ and LTD₄ present in their tracheal lavage; healthy newborns do not.⁹⁶ Leukotriene inhibition with the receptor blocker FPL 57231 prevents and reverses hypoxic pulmonary vasoconstriction in newborn lambs.⁹⁷

Preeclampsia is considered one of the most significant health problems in human pregnancy.^{20, 98, 99, 100, 101} It complicates approximately 5–7% of pregnancies and is a leading cause of fetal growth retardation, indicated premature delivery, and maternal death. It is characterized primarily by proteinuria and increased vasoconstriction leading to maternal hypertension and reduced uteroplacental blood flow. Platelet aggregation, thrombocytopenia, edema, and hyperreflexia, occasionally manifested as convulsions (eclampsia), may be associated with the disease process. Because preeclampsia occurs only during pregnancy or in the presence of placental tissue (i.e. hydatidiform mole),²⁰ the causative factors of preeclampsia most logically originate in the placenta and any effective treatment must correct or prevent abnormalities in the placenta, as well as in the mother. The cause of preeclampsia is not known, but in recent years, a considerable amount of evidence has accrued to indicate that it is associated with an imbalance of increased thromboxane and decreased prostacyclin production, as discussed subsequently.

The first evidence that prostaglandins were involved in preeclampsia came from studies involving vascular sensitivity of pregnant women to AII. Although AII levels are not higher in women with preeclampsia than in healthy pregnant women,^{99, 102} the vascular responsiveness to AII is increased greatly in preeclamptic women. Gant and associates¹⁰³ demonstrated that the dose of AII required to elicit a pressor response of 20 mmHg in diastolic blood pressure was significantly less than normal as early as 23–26 weeks of pregnancy in women destined to develop pregnancy-induced hypertension. This study demonstrated the increased vascular responsiveness of preeclamptic women, and it showed that this increased responsiveness was present as early as the second trimester of pregnancy, long before clinical symptoms were manifest.

The mechanism for the increased vascular responsiveness is not known, but defective prostaglandin production or a loss of response to prostaglandins contributes to the development of pregnancy-induced hypertension.^{23, 99, 104} Everett and colleagues²³ showed that the dose of AII necessary to cause a 20-mmHg rise in diastolic blood pressure in pregnant women is reduced significantly in women treated with indomethacin. In other words, by inhibiting prostaglandin synthesis, the investigators mimicked the increased vascular responsiveness of preeclampsia.

Prostacyclin

Prostacyclin is a potent vasodilator, an inhibitor of platelet aggregation,^{24, 25, 26} and an inhibitor of uterine contractility;^{105, 106, 107} thus, its combined effects favor prevention of maternal hypertension, prevention of platelet aggregation, and promotion of increased uteroplacental blood flow. A deficiency in its production during pregnancy would contribute to the clinical manifestations of preeclampsia. The role of prostacyclin in pregnancy has been reviewed extensively.¹⁰⁸ Various tissues of human pregnancy are known to produce prostacyclin, such as the placenta; umbilical, placental, and uterine vessels; ductus arteriosus; amnion, chorion, and decidua; and myometrium.^{16, 17}

A significant amount of data indicate that prostacyclin production is decreased in preeclampsia, as reviewed comprehensively.^{16, 17} The first evidence came in 1980 from studies of vascular tissues.^{109, 110, 111} Additional studies confirmed the decreased vascular production of prostacyclin in hypertensive pregnancies. Prostacyclin production is, therefore, decreased in umbilical arteries, placental veins, uterine vessels, and subcutaneous vessels obtained from preeclamptic women as compared with healthy pregnant women.

Other studies also confirm decreased prostacyclin production in preeclampsia. For example, urinary metabolite concentrations are depressed,¹¹² as are amniotic fluid concentrations. Plasma levels of prostacyclin's stable metabolite 6-keto-PGF_1α in preeclampsia have been reported to be decreased, unchanged, or increased compared with normal pregnancy. Studies suggest that prostacyclin itself does not circulate in concentrations sufficient to produce physiologic effects, but rather that it exerts its effects locally by a paracrine mechanism between the endothelial cells and the vascular smooth muscle to relax the blood vessels in which it is produced.

Placental production of prostacyclin is decreased significantly in preeclampsia (Fig. 3),^{32, 113} as well as in normotensive pregnancies complicated by fetal growth retardation.^{114, 115} Prostacyclin is produced primarily by the endothelial cells of the placental vessels, but the trophoblast also produces some prostacyclin.^{16, 17, 33, 115, 116, 117} Trophoblastic production of prostacyclin might function to prevent platelet clumping in the intervillous space. Although the production rate of prostacyclin is small when expressed per milligram of wet tissue per hour, the large placental mass makes it a formidable endocrine organ during pregnancy. A placental production rate of 6.7–7.2 pg/mg per hour is equivalent to approximately 94–97 μg per placenta per day.^{32, 113}

With less prostacyclin being produced in preeclampsia, the vasoconstrictor effects of AII, thromboxane, and catecholamines would not be opposed efficiently, leading to hypertension. The renin-angiotensin system is suppressed paradoxically in preeclampsia, but AII levels still are elevated over the nonpregnant state.^{99, 102} Prostacyclin stimulates this system,^{118, 119} so defective prostacyclin production also might explain why AII levels are lower in preeclampsia than in normal pregnancy.

Deficiency in prostacyclin is present before the onset of clinical symptoms. Mills et al. reported in a large multicenter prospective study that urinary concentrations of prostacyclin metabolite were significantly reduced as early as 13–16 weeks of gestation in women who went on to develop preeclampsia.¹²⁰ Klockenbusch et al. reported a similar finding based on longitudinal measurements.¹²¹ These findings that prostacyclin changes predate the clinical onset of preeclampsia are consistent with the findings of Gant et al.¹²² and suggest that increased vascular responsiveness early in pregnancy in women who go on to develop preeclampsia may be due to deficient prostacyclin production.

Thromboxane

Thromboxane opposes the actions of prostacyclin. Thromboxane is a potent vasoconstrictor, a stimulator of platelet aggregation,^{24, 25, 26} and a stimulator of uterine contractility.¹⁰⁶ Its combined actions, if unopposed, lead to maternal hypertension, increased platelet aggregation, and decreased uteroplacental blood flow.

The deficiency in prostacyclin production by preeclamptic placentas is associated with enhanced production of thromboxane similar to the imbalances in these eicosanoids that have been suggested for other pathologic states that favor the development of thrombosis, such as arterial thrombosis, venous thrombosis, myocardial infarction, diabetes, and thrombocytopenia purpura.^{25, 29} The preeclamptic placenta produces more than three times as much thromboxane as the normal placenta, but less than half as much prostacyclin (Fig. 4; also see Fig. 3). The placental imbalance of increased thromboxane and decreased prostacyclin in preeclampsia is even more striking when one compares their production rate ratios between normal and preeclamptic pregnancies (Fig. 5). The normal placenta produces approximately equivalent amounts of thromboxane and prostacyclin; thus, their biologic actions on vascular tone, platelet aggregation, and uterine contractility are balanced. The preeclamptic placenta, however, produces more than seven times as much thromboxane as prostacyclin, so the balance of biologic actions would be tipped heavily in favor of thromboxane (Fig. 6).³² This imbalance explains some of the clinical symptoms of preeclampsia.

Simultaneous measurements of prostacyclin and thromboxane in preeclampsia have been reported for amniotic fluid levels; umbilical, uterine venous, and peripheral plasma levels; and placental release.^{16, 17} Although not all investigators have found decreased prostacyclin coincident with increased thromboxane, in each case, the data show an imbalance in the ratio of thromboxane to prostacyclin that favors thromboxane.

Maternal circulating levels of thromboxane are not increased in mild preeclampsia,^{123, 124, 125} but they are increased significantly in severe preeclampsia¹²⁵ and in hypertensive pregnancies with insufficient fetal growth.¹²⁶ Prostacyclin is decreased significantly in both mild and severe preeclampsia, so the concept of an imbalance between thromboxane and prostacyclin is valid for both mild and severe forms of the disorder.

The thromboxane/prostacyclin imbalance of preeclampsia could reduce the blood flow between the placenta and fetus because thromboxane constricts and prostacyclin dilates the umbilical artery.^{83, 84, 85} Additionally, the distribution of blood flow within the fetus could be altered because the increased thromboxane or the decreased prostacyclin could cause vasoconstriction of the ductus arteriosus and pulmonary vasculature.^{87, 88, 89} The thromboxane/prostacyclin imbalance also would cause vasoconstriction in other fetal vessels and vascular beds because thromboxane is a potent vasoconstrictor in the fetal circulation.^{77, 78} Vasoconstriction coupled with altered distribution of blood flow presumably would have an adverse effect on fetal growth and development.

Increased oxidative stress and the imbalance between thromboxane and prostacyclin

The cause of the imbalance of increased thromboxane and decreased prostacyclin production in women with preeclampsia is not known, but it could be related to a second significant biochemical imbalance. Women with preeclampsia have an increase in oxidative stress and lipid peroxidation; simultaneously, they have a deficiency in several important antioxidants.^{127, 128} The abnormally increased levels of lipid peroxides could be responsible for the imbalance of increased thromboxane and decreased prostacyclin because lipid peroxides can increase thromboxane synthesis by stimulating cyclooxygenase,^{129, 130, 131} but at the same time, they can inhibit prostacyclin synthesis by inhibiting prostacyclin synthase.^{131, 132} Prolonged exposure of endothelial cells to lipid peroxides both in vivo and in vitro results in decreased prostacyclin synthesis.^{132, 133} In vitamin E-deficient rats in which lipid peroxide levels are increased, the ratio of thromboxane to prostacyclin is increased many-fold.¹³⁴ Therefore, abnormally elevated levels of lipid peroxides in preeclampsia could result in the stimulation of thromboxane synthesis and the inhibition of prostacyclin synthesis resulting in their imbalance.

Low-dose aspirin: selective inhibition of thromboxane

Low doses of aspirin (60–150 mg/day) preferentially inhibit thromboxane production without significantly affecting prostacyclin production. Maternal serum thromboxane B₂ concentrations are decreased 47–98% by low-dose aspirin and maternal plasma concentrations by 83%.^{124, 135, 136} Consistent with inhibition of maternal platelet thromboxane synthesis, maternal platelet aggregation is reduced significantly.¹³⁶ Maternal plasma or serum concentrations of prostacyclin metabolite either are unchanged or are decreased only slightly by low-dose aspirin; thus, the ratio of thromboxane to prostacyclin is decreased. In one study of women in whom preeclampsia was at risk of developing, the ratio of thromboxane to prostacyclin decreased 35% after 3 weeks of low-dose aspirin therapy but increased 51% during the same period in the placebo-treated control group.¹²⁴

Low doses of aspirin also preferentially inhibit thromboxane production in the placenta. In vitro studies show that placental arterial production of thromboxane is decreased 84% by low-dose aspirin, but production of prostacyclin is not inhibited significantly.¹³⁷ Similarly, incubation of whole placental tissue with aspirin can result in significant inhibition of thromboxane but not prostacyclin production.¹³⁸ Clinical studies with low-dose aspirin suggest that placental thromboxane production is preferentially inhibited in vivo, as indicated by increased umbilical–placental blood flow and perfusion, increased fetal growth, increased neonatal weight, and increased placental weight.^{123, 124, 139, 140, 141, 142, 143, 144, 145, 146}

Analysis of maternal urinary metabolites of thromboxane and prostacyclin gives an indication of renal production (TXB₂ and 6-keto-PGF_1α) or overall production (2,3-dinor-TXB₂ and 2,3-dinor-6-keto-PGF_1α). Low-dose aspirin decreases the urinary metabolites of thromboxane by 61–87% but has no effect on the urinary metabolites of prostacyclin.^{31, 123}

The ratio of thromboxane to prostacyclin is altered by low-dose aspirin in favor of restoring a balance between thromboxane and prostacyclin. In each of the studies cited previously, the ratio of thromboxane to prostacyclin was decreased by low-dose aspirin. Therefore, the adverse effects of unopposed thromboxane are attenuated or eliminated.

Low-dose aspirin therapy for prevention of preeclampsia

Early studies evaluating low-dose aspirin therapy for the prevention of preeclampsia reported dramatic decreases in the incidence of preeclampsia.¹⁴⁷ However, two large multicenter studies, the NICHD Maternal–Fetal Medicine Unit Network trial of low-dose aspirin and the Collaborative Low-dose Aspirin Study in Pregnancy (CLASP), reported only modest decreases in the incidence of preeclampsia.^{148, 149} The reason for the discrepancy between the earlier studies and the large multicenter studies is not known, but compliance could be an important factor. The initial report based on intent to treat of the NICHD Maternal–Fetal Medicine Unit Network trial of low-dose aspirin to prevent preeclampsia in nulliparous women found a modest reduction in the incidence of preeclampsia from 6.3% for the placebo group to 4.6% for the aspirin group (p = 0.05); however, the intent to treat analysis included noncompliant women and women who discontinued the use of aspirin.¹⁴⁹ Reanalysis based on compliance¹⁵⁰ revealed a significant decrease in the incidence of preeclampsia from 5.7% for patients who were less than 25% compliant, to 2.7% for patients who were greater than 75% compliant (Fig. 7).¹⁵¹ Furthermore, there was a significant decrease in the incidence of low-birth weight <2500 g from 8.8% to 4.3%, a significant decrease in preterm birth <37 weeks from 10.3% to 5.6%, and a significant decrease in “bad outcome” (defined as preeclampsia, eclampsia, severe preeclampsia, or perinatal mortality) from 8.8% to 4.3%. Of those women in the aspirin group, only 53% had a compliance rate greater than 75%, which raises a concern for interpretation of studies based solely on intent to treat. If only one-half of the patients in the treatment group are compliant in taking the test compound, the effectiveness of the compound is not being adequately evaluated. Meta-analysis of low-dose aspirin trials demonstrates that aspirin is effective in preventing preeclampsia.^{152, 153} Emphasis on educating the patient regarding the importance of compliance and making it easy for the patient to comply should be used for low-dose aspirin therapy for the prevention of preeclampsia.

Fig. 7.  Incidence of preeclampsia in relation to compliance of women on low-dose aspirin therapy. Pill compliance was significantly associated with a beneficial effect of aspirin for prevention of preeclampsia. Data are from the NICHD Maternal–Fetal Medicine Network trial of low-dose aspirin to prevent preeclampsia in nulliparous women.¹⁵⁰ The intent to treat analysis of these data demonstrated a modest reduction in the incidence of preeclampsia of 6.3% to 4.6%, p = 0.05, however, only 53% of the women in the aspirin group had a compliance rate greater than 75% and the intent to treat analysis included women who discontinued the use of aspirin. Reanalysis of the data based on percentage of pill compliance of the aspirin group revealed a significant decrease in the incidence of preeclampsia from 5.7% for patients who were less than 25% compliant, to 2.7% for patients who were greater than 75% compliant. Modified from Walsh SW.¹⁵¹

Low-dose aspirin given to pregnant women is beneficial with respect to blood pressure and vascular sensitivity to pressors. Women in whom hypertension is destined to develop in pregnancy are more responsive to the vasoconstrictive effects of AII.¹⁰³ Low-dose aspirin therapy decreases the maternal systemic arterial pressor response to infused AII, and most test results that are positive for AII sensitivity (i.e. abnormally increased vascular sensitivity to AII) become negative after low-dose aspirin therapy.^{135, 154, 155, 156} These studies show that the maternal vasculature becomes less sensitive to vasopressors after administration of low-dose aspirin.

The ratio of thromboxane to prostacyclin is positively correlated with maternal mean blood pressure:¹²⁴ the higher the ratio, the higher the blood pressure. Low-dose aspirin therapy to lower the ratio of thromboxane to prostacyclin results in lower blood pressure. In women with preeclampsia, low-dose aspirin therapy decreases systolic, diastolic, and mean blood pressures.^{155, 157}

Prostaglandins (PGE₂, PGF_2α) have been firmly established as important compounds in the parturitional process. Several comprehensive reviews have been published on this subject. These should be consulted for more detailed descriptions of the role of prostaglandins in labor.^{4, 5, 82, 158, 159, 160, 161, 162} This section presents an overview of the most salient points.

 
Evidence for prostaglandins in labor

Three primary lines of evidence indicate the importance of prostaglandins to the process of labor:

1. Inhibition of prostaglandin synthesis with cyclooxygenase inhibitors, such as indomethacin, flufenamic acid, or aspirin, inhibits uterine contractility, prolongs the length of gestation, and prolongs the duration of labor in monkeys and women.^{163, 164, 165}
   
2. Exogenous administration of prostaglandins (PGE₂ or PGF_2α) stimulates uterine contractility in women and monkeys^{50, 166} and is used to augment labor and for therapeutic termination of pregnancy. Unlike the uterine responsiveness to oxytocin that is manifest at term, the human uterus contracts to prostaglandins at any stage of pregnancy and also during the nonfertile cycle.
   
3. Prostaglandin production is increased at the time of labor (but not before). Peripheral plasma levels of PGF_2α metabolite (PGFM) and TXB₂ increase during labor in women and monkeys,^{6, 7, 8, 18, 167} as do amniotic fluid concentrations of PGFM, PGE₂, PGF_2α, 6-keto-PGF_1α, TX, and arachidonate.^{6, 167, 168, 169, 170, 171, 172, 173, 174, 175} PGE₂ production by amnion and PGE and PGF production by decidua are increased significantly in tissues obtained from women at term in spontaneous labor compared with elective cesarean section.¹⁷⁶
   

 
Stimulators of prostaglandin synthesis

The first step in the synthesis of prostaglandins is the release of arachidonic acid from glycerophospholipids in the cell membranes. The amnion is a rich source of phosphatidylethanolamine and phosphatidylinositol,¹⁷⁷ and these phospholipids appear to be the major source of arachidonate for human parturition. Phospholipase A₂ acts on phosphatidylethanolamine to release arachidonate, and phospholipase C acts on phosphatidylinositol to release arachidonate-rich diacylglycerol, which then is further metabolized by diacylglycerol and monoacylglycerol lipases to yield arachidonic acid.

The specific activities of phospholipase A₂ and phospholipase C are increased in amnion at term compared with during midgestation.¹⁷⁸ The activity of diacylglycerol lipase does not change during the latter part of gestation, but prostaglandin synthase (cyclooxygenase) activity in amnion is greater in tissue collected after spontaneous labor at term than in tissue collected after cesarean section without labor at term.¹⁷⁹ Free intracellular calcium is necessary for activation of phospholipase A₂ and C and for cyclooxygenase.

In sheep, the fetus is the initiator of parturition. In a beautiful set of classic ablation and endocrine experiments, Liggins and colleagues¹⁶² established the sequence of events initiated by the ovine fetus that results in parturition. A summary diagram is presented in Fig. 8. Briefly, the sequence of events is as follows: (1) corticotropin-releasing factor (CRF) is released from the fetal hypothalamus; (2) CRF stimulates adrenocorticotropic hormone (ACTH) release from the fetal pituitary gland; (3) ACTH stimulates cortisol release from the fetal adrenal glands; (4) increased circulating cortisol levels stimulate fetal lung maturation and induce development of 17α-hydroxylase in the placenta; (5) 17α-hydroxylase converts placental progesterone to estradiol, resulting in a decline in progesterone levels and a rise in estradiol levels; (6) the increase in the ratio of estradiol to progesterone results in progesterone withdrawal concomitant with estradiol stimulation of prostaglandin synthesis; (7) increased prostaglandin synthesis stimulates uterine contractility and the onset of parturition. Fetal hypophysectomy or adrenalectomy results in a prolongation of gestation in sheep, whereas fetal ACTH or glucocorticoid infusion results in premature labor.

In primates, the role of the fetus in parturition is less clear. Human anencephalic fetuses¹⁸⁰ and monkey experimental anencephalic fetuses¹⁸¹ are born at an average gestational length that is not significantly different from healthy intact fetuses. There is, however, a significant disruption in the timing of birth, with more fetuses born prematurely or postmaturely. Glucocorticoid administration to pregnant women or monkeys does not induce premature labor as it does in sheep.^{161, 182} Quite the contrary, dexamethasone administered to late gestational pregnant monkeys increases the percentage of animals delivering postmaturely.

The fetus of primates does not play the key role in initiating labor as it does in sheep. However, the primate fetus does play an important role in influencing the timing of birth to ensure that it is not born too early or too late. The mechanisms through which the primate fetus influences the timing of birth are not known, but they may involve nocturnal activity of the fetal adrenal glands.

In the monkey, fetal adrenal activity increases at night because of the release of negative feedback inhibition from transplacental passage of maternal cortisol.¹⁸³ The nocturnal increase in fetal adrenal activity correlates with a nocturnal increase in uterine contractility. Dexamethasone administration not only suppresses fetal adrenal activity, but also abolishes the nocturnal increase in uterine activity, as does death of the fetus. Women, if allowed to deliver naturally, also have a nocturnal rhythm in uterine activity, with the onset of labor contractions being most frequent at night.¹⁸⁴

The fetal adrenal glands appear to control the nocturnal increase in uterine contractility by supplying precursors for estrogen biosynthesis. Studies in nonhuman primates demonstrate that the increase in fetal adrenal activity at night results in an increase in the circulating levels of dehydroepiandrosterone sulfate (DHEAS)¹⁸⁵ and androstenedione, which then are converted to estrogens by the placenta.¹⁸⁶ A nocturnal rhythm in uterine contractility resulting in premature labor can be induced by the infusion of these estrogen precursors.¹⁸⁷ Estrogen stimulates several processes that lead to contraction of the uterine muscle. For example, estrogen stimulates the maternal secretion of oxytocin and increases the number of receptors for oxytocin in the myometrium. Estrogen also stimulates the production of prostaglandins and the formation of gap junctions in the myometrium. Uterine muscle becomes more irritable under the influence of estrogen. The nocturnal rhythm in uterine contractility is under the influence of a nocturnal increase in the maternal circulating levels of oxytocin.¹⁸⁸ Although there are no marked changes in the levels of oxytocin during the day, there is a progressive increase in the nocturnal rise of maternal oxytocin before the onset of labor.

Several other stimulators of prostaglandin synthesis have been identified. Epidermal growth factor (EGF) stimulates PGE₂ output by amniotic cells,¹⁸⁹ increases the rate of synthesis of cyclooxygenase in amnion,¹⁹⁰ and stimulates uterine contractility in estrogen-primed rats.¹⁹¹ Human platelet-derived growth factor also stimulates prostaglandin synthesis,¹⁹² as does platelet activating factor (PAF).^{177, 193} The amnion can synthesize and metabolize PAF, and PAF is present in human amniotic fluid, fetal membranes, and fetal urine.

Neutrophils are also a source of prostaglandins at the time of labor. Labor is associated with extensive infiltration of neutrophils into the decidua and myometrium.¹⁹⁴ Statistical modeling of gene expression at the time of labor found that the most likely initiating event was inflammation. Interleukin-8, a potent neutrophil chemokine, and COX-2, the inducible form of cyclooxygenase, were two key inflammatory genes with increased expression early in the labor process.¹⁹⁵ The early increase in interleukin-8 would activate and attract neutrophils to infiltrate reproductive tissues and increased activity of COX-2 would increase prostaglandin production.

Inhibitors of prostaglandin synthesis

An alternative hypothesis to parturition being initiated by stimulators of prostaglandin synthesis is that pregnancy is maintained until the time for delivery by inhibitors of prostaglandin synthesis. At the time of delivery, the inhibitors are removed or overwhelmed by the appearance of prostaglandin stimulators.

Inhibitors of prostaglandin synthesis have been identified in human amniotic fluid.^{196, 197, 198} Their inhibitory activity is present at midgestation, but the activity decreases significantly by term and declines further with labor. This corresponds to a simultaneous increase in the activity of stimulators of prostaglandin synthesis in the amniotic fluid. The inhibitory actions appear to be exerted either at the cyclooxygenase enzyme or at the phospholipase A₂ enzyme. The chemical identities of the inhibitors are not known, but at least one appears to be similar to lipomodulin (lipocortin, macrocortin), the inhibitory protein active against phospholipase A₂ that is stimulated by the glucocorticoids. The pregnancy factors responsible for regulating the prostaglandin synthesis inhibitors are not known.

 
Steroids

In many nonprimate species (e.g. sheep, rabbits, goats), there is a significant decrease in maternal circulating levels of progesterone coincident with a significant increase in estrogen levels that precedes the onset of parturition. This results in “progesterone withdrawal,” allowing the uterus to contract, and simultaneously results in estrogen stimulation of prostaglandin production, causing the uterus to contract. Overt changes such as these in maternal peripheral blood do not occur in human or nonhuman primate pregnancy, so the concept that progesterone withdrawal precedes labor has been questioned for primates.

Progesterone withdrawal may, however, occur locally within the intrauterine tissues.^{4, 158} Progesterone inhibits PGF_2α production, whereas estrogen stimulates PGE and PGF_2α production by endometrial tissue. Progesterone is formed from pregnenolone by the action of the 3β-hydroxysteroid dehydrogenase/isomerase enzyme system (3βHSD) in amnion, chorion, decidua, and placenta. Estrogens also are formed in these tissues, and they inhibit this enzyme system. At the time of labor, there is a decrease in the formation of progesterone from pregnenolone within the decidua and chorion. At the same time, there is an increase in the activity of the estrone sulfatase enzyme in decidua and chorion. Therefore, the possibility exists that there is a local progesterone withdrawal within the intrauterine tissues brought about by inhibition of the activity of 3βHSD induced by rising intracellular concentrations of unconjugated estrogens.

Another possibility for progesterone withdrawal involves an alteration in progesterone receptors in the myometrium. Progesterone maintains uterine quiescence by binding to the progesterone receptor (PR)-B isoform. Presence of the PR-A isoform antagonizes the relaxing action of progesterone. At the time of parturition there is an increase in the ratio of PR-A to PR-B.^{199, 200, 201, 202} The ability of progesterone to maintain uterine quiescence is thus reduced resulting in a functional progesterone withdrawal despite no decrease in production or plasma levels of progesterone. Evidence suggests that the change in progesterone receptors is mediated by prostaglandins. PGE₂ and PGF_2α acting via the PKC pathway increase the myometrial PR-A/PR-B expression ratio.²⁰³

 
Evidence for lipoxygenase compounds (HETEs, LTs) in labor

Most of the work relating arachidonic acid metabolites to parturition has focused on the cyclooxygenase metabolites, the prostaglandins. Arachidonic acid also can be metabolized by the lipoxygenase enzymes in the human amnion, chorion, decidua, and placenta to form HETEs and leukotrienes.^{16, 17, 34, 35, 82} There is evidence that these compounds also may be involved in the parturitional process. Walsh and colleagues²⁰⁴ presented evidence for this in 1986 by demonstrating that 5-HETE was present in significantly higher concentrations than PGF_2α in the amniotic fluid of chronically catheterized rhesus monkeys (Fig. 9). Furthermore, serial measurements of amniotic fluid 5-HETE demonstrated that its concentrations were altered in association with uterine contractions and labor. In 1987, Romero and others²⁰⁵ reported that 12-HETE, 15-HETE, and LTB₄ all were present in human amniotic fluid at term and that all were significantly higher in amniotic fluid collected from women in labor compared with women not in labor at term. These investigators also reported that 15-HETE and LTB₄ were increased in association with intra-amniotic infection and preterm labor.²⁰⁶ Also in 1987, Bennett and associates²⁰⁷ reported that group B streptococcus stimulated the release of PGE₂ and of dihydroxylated and monohydroxylated HETEs from term human amnion cells in culture.

The concentrations of LTC₄ in amniotic fluid of monkeys also increase significantly with the onset of labor (Fig. 8).²⁰⁸ As was true for 5-HETE, in this study, the concentrations of LTC₄ were significantly higher than those of PGF_2α. Furthermore, one animal went into premature labor and delivered despite treatment with indomethacin and suppressed PGF_2α levels. Amniotic fluid levels of 5-HETE and LTC₄, however, increased progressively until the onset of labor and rupture of the amniotic fluid sac. This demonstrates that labor can occur in the absence of prostaglandins with the presence of HETEs and leukotrienes.

There are so few data available for the lipoxygenase compounds in association with pregnancy and labor that one can only speculate as to their physiologic functions. The HETEs are best known for their chemokinetic and chemotaxic actions on leukocytes.^{1, 2} Mononuclear phagocytes, especially the immature ones, increase in human peripheral blood in association with labor,²⁰⁹ and the cytotoxic activity of human lymphocytes increases significantly at the time of labor.²¹⁰ The intrauterine production of HETEs can act as a signal to recruit leukocytes to the uterus and to activate them, there to augment prostaglandin and leukotriene production and act as a first line of defense against any infection that might enter the uterus from the vagina during or after delivery.

LTC₄ is best known for its potent smooth-muscle stimulating activity. LTC₄ is equal in potency to PGF_2α, but weaker than PGE₂, in stimulating contractions of the guinea pig uterus.²¹¹ The uterus contains receptors for LTC₄.²¹² It is likely that both 5-HETE and LTC₄ act in concert with PGF and PGE to stimulate uterine contractility and parturition because 5-HETE also stimulates human myometrial contractility.

Liggins²¹³ has likened cervical ripening at the time of labor to an inflammatory reaction. Considerable evidence now exists that the entire parturitional process is an inflammatory reaction.^{214, 215, 216, 217}

The role of prostaglandins in pregnancy, including the roles of thromboxane and the lipoxygenase metabolites, the HETEs and leukotrienes, has been discussed in this chapter. The arachidonic acid metabolites are important compounds in pregnancy. They affect maternal blood pressure and blood flow to the uterus, placenta and umbilical–fetal circulation. In preeclampsia, there is an imbalance of increased thromboxane and decreased prostacyclin production that helps explain many of the clinical symptoms of this disorder, such as hypertension, platelet aggregation, and reduced uteroplacental blood flow. Low doses of aspirin selectively inhibit thromboxane without affecting prostacyclin, and all clinical trials with low-dose aspirin for the prevention of preeclampsia show a decreased incidence. The arachidonic acid metabolites also are important compounds for parturition. Prostaglandins stimulate uterine contractility, and there is considerable evidence about their role in the process of labor. The lipoxygenase metabolites, the HETEs and leukotrienes, also play an important role in the parturitional process.