Basic and Clinical Pharmacology, 13th Ed.

The Eicosanoids: Prostaglandins, Thromboxanes, Leukotrienes, & Related Compounds

Emer M. Smyth, PhD, & Garret A. FitzGerald, MD

The eicosanoids are oxygenation products of polyunsaturated long-chain fatty acids. They are ubiquitous in the animal kingdom and are also found—together with their precursors—in a variety of plants. They constitute a very large family of compounds that are highly potent and display an extraordinarily wide spectrum of biologic activity. Because of their biologic activity, the eicosanoids, their specific receptor antagonists and enzyme inhibitors, and their plant and fish oil precursors have great therapeutic potential.


Arachidonic acid (AA), or 5,8,11,14-eicosatetraenoic acid, the most abundant of the eicosanoid precursors, is a 20-carbon (C20) fatty acid containing four double bonds (designated C20:4–6). The first double bond in AA occurs at 6 carbons from the methyl end, defining AA as an omega-6 fatty acid. AA must first be released or mobilized from the sn-2 position of membrane phospholipids by one or more lipases of the phospholipase A2 (PLA2) type (Figure 18–1) for eicosanoid synthesis to occur. The phospholipase A2 superfamily consists of 15 groups, with at least three classes of phospholipases contributing to arachidonate release from membrane lipids: (1) cytosolic (c) PLA2, and (2) secretory (s) PLA2, which are calcium-dependent; and (3) calcium-independent (i) PLA2. Chemical and physical stimuli activate the Ca2+-dependent translocation of cPLA2, which has high affinity for AA, to the membrane, where it releases arachidonate. Arachidonate release via iPLA2 and various sPLA2 subtypes also has been examined. Under nonstimulated conditions, AA liberated by iPLA2 is reincorporated into cell membranes, so there is negligible eicosanoid biosynthesis. While cPLA2 dominates in the acute release of AA, inducible sPLA2 contributes under conditions of sustained or intense stimulation of AA production. AA can also be released from phospholipase C-generated diacylglcerol esters by the action of diacylglycerol and monoacylglycerol lipases.


FIGURE 18–1 Pathways of arachidonic acid (AA) release and metabolism.

Following mobilization, AA is oxygenated by four separate routes: enzymatically via the cyclooxygenase (COX), lipoxygenase, and P450 epoxygenase pathways; and nonenzymatically via the isoeicosanoid pathway (Figure 18–1). Among factors determining the type of eicosanoid synthesized are (1) the substrate lipid species, (2) the type of cell, and (3) the manner in which the cell is stimulated. Distinct but related products can be formed from precursors other than AA. For example, homo-γ-linoleic acid (C20:3–6), also an omega-6 fatty acid, or eicosapentaenoic acid (C20:5–3), an omega-3 fatty acid, yield products that differ quantitatively and qualitatively from those derived from AA. This shift in product formation is the basis for dietary manipulation of eicosanoid generation using fatty acids obtained from cold-water fish or from plants as nutritional supplements in humans. For example, thromboxane (TXA2), a powerful vasoconstrictor and platelet agonist, is synthesized from AA via the COX pathway. COX metabolism of eicosapentaenoic acid yields TXA3, which is relatively inactive. 3-Series prostaglandins, such as prostaglandin E3 (PGE3), can also act as partial agonists or antagonists thereby reducing the activity of their AA-derived 2-series counterparts. The hypothesis that dietary eicosapentaenoate substitution for arachidonate could reduce the incidence of cardiovascular disease and cancer remains controversial.




Products of Prostaglandin Endoperoxide Synthases (Cyclooxygenases)

Two unique COX isozymes convert AA into prostaglandin endoperoxides. PGH synthase-1 (COX-1) is expressed constitutively in most cells. In contrast, PGH synthase-2 (COX-2) is more readily inducible; its expression varies depending on the stimulus. COX-2 is an immediate early-response gene product that is markedly up-regulated by shear stress, growth factors, tumor promoters, and cytokines, consistent with the presence of multiple regulatory motifs in the promoter and 3′ untranslated regions of the COX-2 gene. COX-1 generates prostanoids for “housekeeping” functions, such as gastric epithelial cytoprotection, whereas COX-2 is the major source of prostanoids in inflammation and cancer. This distinction is overly simplistic, however; there are both physiologic and pathophysiologic processes in which each enzyme is uniquely involved and others in which they function coordinately. For example, endothelial COX-2 is the primary source of vascular prostacyclin (PGI2), whereas renal COX-2-derived prostanoids are important for normal renal development and maintenance of function. Nonsteroidal anti-inflammatory drugs (NSAIDs; see Chapter 36) exert their therapeutic effects through inhibition of the COXs. Most older NSAIDs, like indomethacin, sulindac, meclofenamate, and ibuprofen nonselectively inhibit both COX-1 and COX-2, whereas the selective COX-2 inhibitors follow the order celecoxib = diclofenac = meloxicam = etodolac < valdecoxib << rofecoxib < lumiracoxib = etoricoxib for increasing COX-2 selectivity. Aspirin acetylates and inhibits both enzymes covalently and hence irreversibly. Low doses (< 100 mg/d) inhibit preferentially, but not exclusively, platelet COX-1, whereas higher doses inhibit both systemic COX-1 and COX-2. Genetic variations in human COX-2 variants have been linked with increased coronary heart disease risk, increases in some cancers, and reduced pain perception.

Both COX-1 and COX-2 function as homodimers inserted into the membrane of the endoplasmic reticulum to promote the uptake of two molecules of oxygen by cyclization of AA to yield a C9–C11endoperoxide C15hydroperoxide (Figure 18–2). This product is PGG2, which is then rapidly modified by the peroxidase moiety of the COX enzyme to add a 15-hydroxyl group that is essential for biologic activity. This product is PGH2. Both endoperoxides are highly unstable. Analogous families—PGH1 and PGH3 and their subsequent 1-series and 3-series products—are derived from homo-γ-linolenic acid and eicosapentaenoic acid, respectively. In both COX-1 and COX-2 homodimers, one protomer acts as the catalytic unit binding AA for oxygenation, while the other acts as an allosteric modifier of catalytic activity.


FIGURE 18–2 Prostanoid biosynthesis. Compound names are enclosed in boxes.

The prostaglandins, thromboxane, and prostacyclin, collectively termed the prostanoids, are generated from PGH2 through the action of downstream isomerases and synthases. These terminal enzymes are expressed in a relatively cell-specific fashion, such that most cells make one or two dominant prostanoids. The prostaglandins differ from each other in two ways: (1) in the substituents of the pentane ring (indicated by the last letter, eg, E and F in PGE and PGF) and (2) in the number of double bonds in the side chains (indicated by the subscript, eg, PGE1, PGE2). PGH2 is metabolized by prostacyclin, thromboxane, and PGF synthases (PGIS, TXAS, and PGFS) to PGI2, TXA2, and PGF, respectively. Two additional enzymes, 9,11-endoperoxide reductase and 9-ketoreductase, provide for PGF synthesis from PGH2 and PGE2, respectively. At least three PGE2 synthases have been identified: microsomal (m) PGES-1, the more readily inducible mPGES-2, and cytosolic PGES. There are two distinct PGDS isoforms, the lipocalin-type PGDS and the hematopoietic PGDS.

Several products of the arachidonate series are of current clinical importance. Alprostadil (PGE1) may be used for its smooth muscle relaxing effects to maintain the ductus arteriosus patent in some neonates awaiting cardiac surgery and in the treatment of impotence. Misoprostol, a PGE1 derivative, is a cytoprotective prostaglandin used in preventing peptic ulcer and in combination with mifepristone (RU-486) for terminating early pregnancies. Dinoprostone (PGE2) and PGF are used in obstetrics to induce labor. Latanoprost and several similar compounds are topically active PGF derivatives used in ophthalmology to reduce intraocular pressure in open-angle glaucoma or ocular hypertension. Prostacyclin (PGI2) is synthesized mainly by the vascular endothelium and is a powerful vasodilator and inhibitor of platelet aggregation. Synthetic PGI2 (epoprostenol) and PGI2 analogs (iloprost, treprostinil) are used to treat pulmonary hypertension and portopulmonary hypertension. In contrast, thromboxane (TXA2) has undesirable properties (aggregation of platelets, vasoconstriction). Therefore TXA2-receptor antagonists and synthesis inhibitors have been developed for cardiovascular indications, although these (except for aspirin) have yet to establish a place in clinical usage, and, in a recent large clinical trial, TXA2 receptor antagonism failed to show superiority over low-dose aspirin for secondary stroke protection.

All the naturally occurring COX products undergo rapid metabolism to inactive products either by hydration (for PGI2 and TXA2) or by oxidation of the key 15-hydroxyl group to the corresponding ketone by prostaglandin 15-hydroxy prostaglandin dehydrogenase (15-PGDH) after cellular uptake via an organic anion transporter polypeptide (OATP 2A1). Further metabolism is by δ13 reduction, β-oxidation, and ω-oxidation. The inactive metabolites are chemically stable and can be quantified in blood and urine by immunoassay or mass spectrometry as a measure of the in vivo synthesis of their parent compounds.

Products of Lipoxygenase

The metabolism of AA by the 5-, 12-, and 15-lipoxygenases (LOX) results in production of hydroperoxyeicosatetraenoic acids (HPETEs), which rapidly convert to hydroxy derivatives (HETEs). 5-LOX, the most actively investigated pathway, gives rise to the leukotrienes (Figure 18–3) and is present in leukocytes (neutrophils, basophils, eosinophils, and monocyte-macrophages) and other inflammatory cells such as mast cells and dendritic cells. This pathway is of great interest since it is associated with asthma, anaphylactic shock, and cardiovascular disease. Stimulation of these cells elevates intracellular Ca2+ and releases arachidonate; incorporation of molecular oxygen by 5-LOX, in association with 5-LOX-activating protein (FLAP), yields 5(S)-HPETE, which is then further converted by 5-LOX to the unstable epoxide leukotriene A4 (LTA4). This intermediate is either converted to the dihydroxy leukotriene B4 (LTB4), via the action of LTA4 hydrolase, or is conjugated with glutathione to yield leukotriene C4 (LTC4), by LTC4 synthase. Sequential degradation of the glutathione moiety by peptidases yields LTD4 and LTE4. These three products, LTC4, D4, and E4, are called cysteinyl leukotrienes. Although leukotrienes are predominantly generated in leukocytes, nonleukocyte cells (eg, endothelial cells) that express enzymes downstream of 5-LOX/FLAP can take up and convert leukocyte-derived LTA4 in a process termed transcellular biosynthesis. Transcellular formation of prostaglandins has also been shown; for example, endothelial cells can use platelet PGH2 to form PGI2.


FIGURE 18–3 Leukotriene (LT) biosynthesis. LTC4, LTD4, and LTE4 are known collectively as the cysteinyl (Cys) LTs. FLAP, 5-LOX-activating protein; GT, glutamyl transpeptidase; GL, glutamyl leukotrienase. *Additional products include 5,6-; 8,9-; and 14,15-EET; and 19-, 18-, 17-, and 16-HETE.

LTC4 and LTD4 are potent bronchoconstrictors and are recognized as the primary components of the slow-reacting substance of anaphylaxis (SRS-A) that is secreted in asthma and anaphylaxis. There are four current approaches to antileukotriene drug development: 5-LOX enzyme inhibitors, cysteinyl leukotriene-receptor antagonists, inhibitors of FLAP, and phospholipase A2 inhibitors. Variants in the human 5-LOX gene (ALOX5) or the cysteinyl receptors (CYSLTR1 or CYSLTR2) have been linked with asthma and with the response to antileukotriene drugs.

LTA4, the primary product of 5-LOX, can be converted with appropriate stimulation via 12-LOX in platelets in vitro to the lipoxins LXA4 and LXB4 in vitro. These mediators can also be generated through 5-LOX metabolism of 15(S)-HETE, the product of 15-LOX-2 metabolism of arachidonic acid. The stereochemical isomer, 15(R)-HETE, may be derived from the action of aspirin-acetylated COX-2 and further transformed in leukocytes by 5-LOX to 15-epi-LXA4 or 15-epi-LXB4, the so-called aspirin-triggered lipoxins. The 15-LOX-1 isoform prefers linoleic acid as a substrate, forming 13(S)-hydroxyoctadecadienoic acid, while the sequential action of 15-LOX-1 and 5-LOX can convert the omega-3 fatty acid docosahexaenoic acid (DHA) to the resolvins, potentially anti-inflammatory, pro-resolving lipids. Synthetic resolvins, lipoxins, and epi-lipoxins exert anti-inflammatory actions when applied in vivo. Although these compounds can be formed from endogenous substrates in vitro and when synthesized may have potent biologic effects, the importance of the endogenous compounds in vivo in human biology remains ill defined. 12-HETE, a product of 12-LOX, can also undergo a catalyzed molecular rearrangement to epoxyhydroxyeicosatrienoic acids called hepoxilins. Proinflammatory effects of synthetic hepoxilins have been reported although their biologic relevance is unclear.

The epidermal LOXs, 12(R)-LOX and LOX-3, are distinct from “conventional” enzymes both in their natural substrates, which appear to not be arachidonic acid and linoleic acid, and in the products formed. Mutations in the genes for 12(R)-LOX (ALOX12B) or LOX-3 (ALOXE3) are commonly associated with autosomal recessive congenital ichthyosis, and epidermal accumulation of 12(R)-HETE is a feature of psoriasis and ichthyosis. Inhibitors of 12(R)-LOX are under investigation for the treatment of these proliferative skin disorders.

Epoxygenase Products

Specific isozymes of microsomal cytochrome P450 monooxygenases convert AA to hydroxy- or epoxyeicosatrienoic acids (Figures 18–1 and 18–3). The products are 20-HETE, generated by the CYP hydroxylases (CYP3A, 4A, 4F) and the 5,6-, 8,9-, 11,12-, and 14,15-epoxyeicosatrienoic acids (EETs), which arise from the CYP epoxygenase (2J, 2C). Their biosynthesis can be altered by pharmacologic, nutritional, and genetic factors that affect P450 expression. The biologic actions of the EETs are reduced by their conversion to the corresponding, and biologically less active, dihydroxyeicosatrienoic acids (DHETs) through the action of soluble epoxide hydrolase (sEH). Unlike the prostaglandins, the EETs can be esterified into phospholipids, which then act as storage sites. Intracellular fatty acid-binding proteins promote EET uptake into cells, incorporation into phospholipids, and availability to sEH. EETs are synthesized in endothelial cells and cause vasodilation in a number of vascular beds by activating the smooth muscle large conductance Ca2+-activated K+ channels. This results in smooth muscle cell hyperpolarization and vasodilation, leading to reduced blood pressure. Substantial evidence indicates that EETs may function as endothelium-derived hyperpolarizing factors, particularly in the coronary circulation. 15(S)-Hydroxy-11,12-EET, which arises from the 15-LOX pathway, is also an endothelium-derived hyperpolarizing factor and a substrate for sEH. Consequently there is interest in inhibitors of soluble sEH as potential antithrombotic and antihypertensive drugs. An exception to the general response to EETs as vasodilators is the pulmonary vasculature where they cause vasoconstriction. It is unclear yet whether this activity of EETs may limit the potential clinical use of sEH inhibitors. Down-regulation of pulmonary sEH may contribute to pulmonary hypertension. Anti-inflammatory, antiapoptotic, and proangiogenic actions of the EETs have also been reported.


The isoeicosanoids, a family of eicosanoid isomers, are formed nonenzymatically by direct free radical-based action on AA and related lipid substrates. The isoprostanes thus formed are prostaglandin stereoisomers. Because prostaglandins have many asymmetric centers, they have a large number of potential stereoisomers. COX is not needed for the formation of the isoprostanes, and its inhibition with aspirin or other NSAIDs should not affect the isoprostane pathway. The primary epimerization mechanism is peroxidation of arachidonate by free radicals. Peroxidation occurs while arachidonic acid is still esterified to the membrane phospholipids. Thus, unlike prostaglandins, these stereoisomers are “stored” as part of the membrane. They are then cleaved by phospholipases, circulate, and are excreted in urine. Isoprostanes are present in relatively large amounts (tenfold greater in blood and urine than the COX-derived prostaglandins). They have potent vasoconstrictor effects when infused into renal and other vascular beds and may activate prostanoid receptors. They also may modulate other aspects of vascular function, including leukocyte and platelet adhesive interactions and angiogenesis. It has been speculated that they may contribute to the pathophysiology of inflammatory responses in a manner insensitive to COX inhibitors. A particular difficulty in assessing the likely biologic functions of isoprostanes—several of which have been shown to serve as incidental ligands at prostaglandin receptors—is that while high concentrations of individual isoprostanes may be necessary to elicit a response, multiple compounds are formed coincidentally in vivo under conditions of oxidant stress. Analogous leukotriene and EET isomers have been described.



Receptor Mechanisms

As a result of their short half-lives, the eicosanoids act mainly in an autocrine and a paracrine fashion, ie, close to the site of their synthesis, and not as circulating hormones. These ligands bind to receptors on the cell surface, and pharmacologic specificity is determined by receptor density and type on different cells (Figure 18–4). A single gene product has been identified for each of the PGI2 (IP), PGF (FP), and TXA2 (TP) receptors, while four distinct PGE2 receptors (EPs 1–4) and two PGD2 receptors (DP1 and DP2) have been cloned. Additional isoforms of the human TP (α and β), FP (A and B), and EP3 (I, II, III, IV, V, VI, e, and f) receptors can arise through differential mRNA splicing. Two receptors exist for LTB4 (BLT1 and BLT2) and for LTC4/LTD4 (cysLT1 and cysLT2). It appears that LTE4 functions through one or more receptors distinct from cysLT1/cysLT2, with some evidence that the orphan receptor GPR99 and the ADP receptor P2Y12 may function as LTE4 receptors. The formyl peptide (fMLP)-1 receptor can be activated by lipoxin A4 and consequently has been termed the ALX receptor. Receptor heterodimerization has been reported for a number of the eicosanoid receptors, providing for additional receptor subtypes from the currently identified gene products. All of these receptors are G protein-coupled; properties of the best-studied receptors are listed in Table 18–1.

TABLE 18–1 Eicosanoid receptors.1



FIGURE 18–4 Prostanoid receptors and their signaling pathways. fMLP, formylated MetLeuPhe, a small peptide receptor; PLC-β, phospholipase C-β. All of the receptors shown are of the seven-transmembrane, G protein-coupled type. The terms “relaxant,” “contractile,” and “inhibitory” refer to the phylogenetic characterization of their primary effects. **, all EP3 isoforms couple through Gi but some can also activate Gs or G12/13 pathways. RhoGEF, rho guanine nucleotide exchange factor. See text for additional details.

EP2, EP4, IP, and DP1 receptors activate adenylyl cyclase via Gs. This leads to increased intracellular cAMP levels, which in turn activate specific protein kinases (see Chapter 2). EP1, FP, and TP activate phosphatidylinositol metabolism, leading to the formation of inositol trisphosphate, with subsequent mobilization of Ca2+ stores and an increase of free intracellular Ca2+. TP also couples to multiple G proteins, including G12/13 and G16, to stimulate small G protein signaling pathways, and may activate or inhibit adenylyl cyclase via Gs (TPα) or Gi (TPβ), respectively. EP3 isoforms can couple to both increased intracellular calcium and to increased or decreased cAMP. The DP2receptor (also known as the chemoattractant receptor-homologous molecule expressed on TH2 cells, or CRTH2), which is unrelated to the other prostanoid receptors, is a member of the fMLP receptor superfamily. This receptor couples through a Gi-type G protein and leads to inhibition of cAMP synthesis and increases in intracellular Ca2+in a variety of cell types.

LTB4 also causes inositol trisphosphate release via the BLT1 receptor, causing activation, degranulation, and superoxide anion generation in leukocytes. The BLT2 receptor, a low-affinity receptor for LTB4, is also bound with reasonable affinity by 12(S)- and 12(R)-HETE, although the biologic relevance of this observation is not clear. CysLT1 and cysLT2 couple to Gq, leading to increased intracellular Ca2+. Studies have also placed Gi downstream of cysLT2. An orphan receptor, GPR17, binds cysLTs and may negatively regulate the function of cysLT1, but its physiologic role remains ill defined. As noted above, the EETs promote vasodilation via paracrine activation of calcium-activated potassium channels on smooth muscle cells leading to hyperpolarization and relaxation. This occurs in a manner consistent with activation of a Gs-coupled receptor, although a specific EET receptor has yet to be identified. EETs may also act in an autocrine manner directly activating endothelial transient receptor potential channels to cause endothelial hyperpolarization, which is then transferred to the smooth muscle cells by gap junctions or potassium ions. Specific receptors for isoprostanes have not been identified, and the biologic importance of their capacity to act as incidental ligands at prostaglandin receptors remains to be established.

Although prostanoids can activate peroxisome proliferator-activated receptors (PPARs) if added in sufficient concentration in vitro, it remains questionable whether these compounds ever attain concentrations sufficient to function as endogenous nuclear-receptor ligands in vivo.

Effects of Prostaglandins & Thromboxanes

The prostaglandins and thromboxanes have major effects on smooth muscle in the vasculature, airways, and gastrointestinal and reproductive tracts. Contraction of smooth muscle is mediated by the release of calcium, while relaxing effects are mediated by the generation of cAMP. Many of the eicosanoids’ contractile effects on smooth muscle can be inhibited by lowering extracellular calcium or by using calcium channel-blocking drugs. Other important targets include platelets and monocytes, kidneys, the central nervous system, autonomic presynaptic nerve terminals, sensory nerve endings, endocrine organs, adipose tissue, and the eye (the effects on the eye may involve smooth muscle).

A. Smooth Muscle

1. VascularTXA2 is a potent vasoconstrictor. It is also a smooth muscle cell mitogen and is the only eicosanoid that has convincingly been shown to have this effect. The mitogenic effect is potentiated by exposure of smooth muscle cells to testosterone, which up-regulates smooth muscle cell TP expression. PGF is also a vasoconstrictor but is not a smooth muscle mitogen. Another vasoconstrictor is the isoprostane 8-iso-PGF, also known as iPFIII, which may act via the TP receptor.

Vasodilator prostaglandins, especially PGI2 and PGE2, promote vasodilation by increasing cAMP and decreasing smooth muscle intracellular calcium, primarily via the IP and EP4 receptors. Vascular PGI2 is synthesized by both smooth muscle and endothelial cells, with the COX-2 isoform in the latter cell type being the major contributor. In the microcirculation, PGE2 is a vasodilator produced by endothelial cells. PGI2 inhibits proliferation of smooth muscle cells, an action that may be particularly relevant in pulmonary hypertension. PGD2 may also function as a vasodilator, in particular as a dominant mediator of flushing induced by the lipid-lowering drug niacin.

2. Gastrointestinal tractMost of the prostaglandins and thromboxanes activate gastrointestinal smooth muscle. Longitudinal muscle is contracted by PGE2 (via EP3) and PGF (via FP), whereas circular muscle is contracted strongly by PGF and weakly by PGI2, and is relaxed by PGE2 (via EP4). Administration of either PGE2 or PGF results in colicky cramps (see Clinical Pharmacology of Eicosanoids, below). The leukotrienes also have powerful contractile effects.

3. AirwaysRespiratory smooth muscle is relaxed by PGE2 and PGI2 and contracted by PGD2, TXA2, and PGF. Studies of DP1 and DP2 receptor knockout mice suggest an important role of this prostanoid in asthma, although the DP2 receptor appears more relevant to allergic airway diseases. The cysteinyl leukotrienes are also bronchoconstrictors. They act principally on smooth muscle in peripheral airways and are a thousand times more potent than histamine, both in vitro and in vivo. They also stimulate bronchial mucus secretion and cause mucosal edema. Bronchospasm occurs in about 10% of people taking NSAIDs, possibly because of a shift in arachidonate metabolism from COX metabolism to leukotriene formation.

4. ReproductiveThe actions of prostaglandins on reproductive smooth muscle are discussed below under section D, Reproductive Organs.

B. Platelets

Platelet aggregation is markedly affected by eicosanoids. Low concentrations of PGE2 enhance (via EP3 receptors), whereas higher concentrations inhibit (via IP receptors), platelet aggregation. Both PGD2and PGI2 inhibit aggregation via, respectively, DP1- and IP receptor-dependent elevation in cAMP generation. Unlike their human counterparts, mouse platelets do not express DP1. TXA2 is the major product of COX-1, the only COX isoform expressed in mature platelets, with COX-1-derived PGD2 found in lesser amounts. Itself a platelet aggregator, TXA2 amplifies the effects of other, more potent, platelet agonists such as thrombin. The TP-Gq signaling pathway elevates intracellular Ca2+ and activates protein kinase C, facilitating platelet aggregation and TXA2 biosynthesis. Activation of G12/G13 induces Rho/Rho-kinase–dependent regulation of myosin light chain phosphorylation leading to platelet shape change. Mutations in the human TP have been associated with mild bleeding disorders. The platelet actions of TXA2 are restrained in vivo by PGI2, which inhibits platelet aggregation by all recognized agonists, and PGD2. Platelet COX-1-derived TXA2 biosynthesis is increased during platelet activation and aggregation and is irreversibly inhibited by chronic administration of aspirin at low doses. Urinary metabolites of TXA2 increase in clinical syndromes of platelet activation such as myocardial infarction and stroke. Macrophage COX-2 appears to contribute roughly 10% of the increment in TXA2 biosynthesis observed in smokers, while the rest is derived from platelet COX-1. A variable contribution, presumably from macrophage COX-2, may be insensitive to the effects of low-dose aspirin. In a single trial comparing low- and high-dose aspirin, no increase in benefit was associated with increased dose; in fact, this study, as well as indirect comparisons across placebo-controlled trials, suggests an inverse dose-response relationship, perhaps reflecting increasing inhibition of PGI2 synthesis at higher doses of aspirin.

C. Kidney

Both the medulla and the cortex of the kidney synthesize prostaglandins, the medulla substantially more than the cortex. COX-1 is expressed mainly in cortical and medullary collecting ducts and mesangial cells, arteriolar endothelium, and epithelial cells of Bowman’s capsule. COX-2 is restricted to the renal medullary interstitial cells, the macula densa, and the cortical thick ascending limb.

The major renal eicosanoid products are PGE2 and PGI2, followed by PGF and TXA2. The kidney also synthesizes several hydroxy-eicosatetraenoic acids, leukotrienes, cytochrome P450 products, and epoxides. Prostaglandins play important roles in maintaining blood pressure and regulating renal function, particularly in marginally functioning kidneys and volume-contracted states. Under these circumstances, renal cortical COX-2-derived PGE2 and PGI2maintain renal blood flow and glomerular filtration rate through their local vasodilating effects. These prostaglandins also modulate systemic blood pressure through regulation of water and sodium excretion. Expression of medullary COX-2 and mPGES-1 is increased under conditions of high salt intake. COX-2-derived prostanoids increase medullary blood flow and inhibit tubular sodium reabsorption, while COX-1-derived products promote salt excretion in the collecting ducts. Increased water clearance probably results from an attenuation of the action of antidiuretic hormone (ADH) on adenylyl cyclase. Loss of these effects may underlie the systemic or salt-sensitive hypertension often associated with COX inhibition. A common misperception—often articulated in discussion of the cardiovascular toxicity of drugs such as rofecoxib—is that hypertension secondary to NSAID administration is somehow independent of the inhibition of prostaglandins. Loop diuretics, eg, furosemide, produce some of their effect by stimulating COX activity. In the normal kidney, this increases the synthesis of the vasodilator prostaglandins. Therefore, patient response to a loop diuretic is diminished if a COX inhibitor is administered concurrently (see Chapter 15).

There is an additional layer of complexity associated with the effects of renal prostaglandins. In contrast to the medullary enzyme, cortical COX-2 expression is increased by low salt intake, leading to increased renin release. This elevates glomerular filtration rate and contributes to enhanced sodium reabsorption and a rise in blood pressure. PGE2 is thought to stimulate renin release through activation of EP4 or EP2. PGI2 can also stimulate renin release and this may be relevant to maintenance of blood pressure in volume-contracted conditions and to the pathogenesis of renovascular hypertension. Inhibition of COX-2 may reduce blood pressure in these settings.

TXA2 causes intrarenal vasoconstriction (and perhaps an ADH-like effect), resulting in a decline in renal function. The normal kidney synthesizes only small amounts of TXA2. However, in renal conditions involving inflammatory cell infiltration (such as glomerulonephritis and renal transplant rejection), the inflammatory cells (monocyte-macrophages) release substantial amounts of TXA2. Theoretically, TXA2synthase inhibitors or receptor antagonists should improve renal function in these patients, but no such drug is clinically available. Hypertension is associated with increased TXA2 and decreased PGE2 and PGI2 synthesis in some animal models, eg, the Goldblatt kidney model. It is not known whether these changes are primary contributing factors or secondary responses. Similarly, increased TXA2 formation has been reported in cyclosporine-induced nephrotoxicity, but no causal relationship has been established. PGF may elevate blood pressure by regulating renin release in the kidney. Although more research is necessary, FP antagonists have potential as novel antihypertensive drugs.

D. Reproductive Organs

1. Female reproductive organsAnimal studies demonstrate a role for PGE2 and PGF in early reproductive processes such as ovulation, luteolysis, and fertilization. Uterine muscle is contracted by PGF, TXA2, and low concentrations of PGE2; PGI2 and high concentrations of PGE2 cause relaxation. PGF, together with oxytocin, is essential for the onset of parturition. The effects of prostaglandins on uterine function are discussed below (see Clinical Pharmacology of Eicosanoids).

2. Male reproductive organsDespite the discovery of prostaglandins in seminal fluid, and their uterotropic effects, the role of prostaglandins in semen is still conjectural. The major source of these prostaglandins is the seminal vesicle; the prostate, despite the name “prostaglandin,” and the testes synthesize only small amounts. The factors that regulate the concentration of prostaglandins in human seminal plasma are not known in detail, but testosterone does promote prostaglandin production. Thromboxane and leukotrienes have not been found in seminal plasma. Men with a low seminal fluid concentration of prostaglandins are relatively infertile.

Smooth muscle-relaxing prostaglandins such as PGE1 enhance penile erection by relaxing the smooth muscle of the corpora cavernosa (see Clinical Pharmacology of Eicosanoids).

E. Central and Peripheral Nervous Systems

1. FeverPGE2 increases body temperature, predominantly via EP3, although EP1 also plays a role, especially when administered directly into the cerebral ventricles. Exogenous PGF and PGI2 induce fever, whereas PGD2 and TXA2 do not. Endogenous pyrogens release interleukin-1, which in turn promotes the synthesis and release of PGE2. This synthesis is blocked by aspirin, other antipyretic NSAIDs, and acetaminophen.

2. SleepWhen infused into the cerebral ventricles, PGD2 induces natural sleep (as determined by electroencephalographic analysis) via activation of DP1 receptors and secondary release of adenosine. PGE2infusion into the posterior hypothalamus causes wakefulness.

3. NeurotransmissionPGE compounds inhibit the release of norepinephrine from postganglionic sympathetic nerve endings. Moreover, NSAIDs increase norepinephrine release in vivo, suggesting that the prostaglandins play a physiologic role in this process. Thus, vasoconstriction observed during treatment with COX inhibitors may be due, in part, to increased release of norepinephrine as well as to inhibition of the endothelial synthesis of the vasodilators PGE2 and PGI2. PGE2 and PGI2 sensitize the peripheral nerve endings to painful stimuli. PGE2 acts via EP1 and EP4 receptors to potentiate excitatory cation channel activity and inhibit hyperpolarizing K+ channel activity, thereby increasing membrane excitability. Prostaglandins also modulate pain centrally. Both COX-1 and COX-2 are expressed in the spinal cord and release prostaglandins in response to peripheral pain stimuli. PGE2, and perhaps also PGD2, PGI2, and PGF, contribute to so-called central sensitization, an increase in excitability of spinal dorsal horn neurons, that augments pain intensity, widens the area of pain perception, and results in pain from normally innocuous stimuli. PGE2 acts on the EP2 receptor to facilitate presynaptic release of excitatory neurotransmitters and block inhibitory glycinergic neurotransmission as well as postsynaptically to enhance excitatory neurotransmitter receptor activity.

F. Inflammation and Immunity

PGE2 and PGI2 are the predominant prostanoids associated with inflammation. Both markedly enhance edema formation and leukocyte infiltration by promoting blood flow in the inflamed region. PGE2 and PGI2, through activation of EP2 and IP, respectively, increase vascular permeability and leukocyte infiltration. Through its action as a platelet agonist, TXA2 can also increase platelet-leukocyte interactions. Although probably not made by lymphocytes, prostaglandins may potently regulate lymphocyte function. PGE2 and TXA2 may play a role in T-lymphocyte development by regulating apoptosis of immature thymocytes. PGI2 contributes to immune suppression by interfering with dendritic cell maturation and antigen uptake for presentation to immune cells. PGE2 suppresses the immunologic response by inhibiting differentiation of B lymphocytes into antibody-secreting plasma cells, thus depressing the humoral antibody response. It also inhibits cytotoxic T-cell function, mitogen-stimulated proliferation of T lymphocytes, and maturation and function of TH1 lymphocytes. PGE2 can modify myeloid cell differentiation, promoting type 2 immune-suppressive macrophage and myeloid suppressor cell phenotypes. These effects likely contribute to immune escape in tumors where infiltrating myeloid-derived cells predominantly display type 2 phenotypes. PGD2, a major product of mast cells, is a potent chemoattractant for eosinophils in which it also induces degranulation and leukotriene biosynthesis. PGD2 also induces chemotaxis and migration of TH2 lymphocytes, mainly via activation of DP2, although a role for DP1 has also been established. It remains unclear how these two receptors coordinate the actions of PGD2 in inflammation and immunity. A degradation product of PGD2, 15d-PGJ2, at concentrations actually formed in vivo, may also activate eosinophils via the DP2 (CRTH2) receptor.

G. Bone Metabolism

Prostaglandins are abundant in skeletal tissue and are produced by osteoblasts and adjacent hematopoietic cells. The major effect of prostaglandins (especially PGE2, acting on EP4) in vivo is to increase bone turnover, ie, stimulation of bone resorption and formation. EP4 receptor deletion in mice results in an imbalance between bone resorption and formation, leading to a negative balance of bone mass and density in older animals. Prostaglandins may mediate the effects of mechanical forces on bones and changes in bone during inflammation. EP4-receptor deletion and inhibition of prostaglandin biosynthesis have both been associated with impaired fracture healing in animal models. COX inhibitors can also slow skeletal muscle healing by interfering with prostaglandin effects on myocyte proliferation, differentiation, and fibrosis in response to injury. Prostaglandins may contribute to the bone loss that occurs at menopause; it has been speculated that NSAIDs may be of therapeutic value in osteoporosis and bone loss prevention in older women. However, controlled evaluation of such therapeutic interventions has not been carried out. NSAIDs, especially those specific for inhibition of COX-2, delay bone healing in experimental models of fracture.

H. Eye

PGE, PGF, and PGD derivatives lower intraocular pressure. The mechanism of this action is unclear but probably involves increased outflow of aqueous humor from the anterior chamber via the uveoscleral pathway (see Clinical Pharmacology of Eicosanoids).

I. Cancer

There has been considerable interest in the role of prostaglandins, and in particular the COX-2 pathway, in the development of malignancies. Pharmacologic inhibition or genetic deletion of COX-2 restrains tumor formation in models of colon, breast, lung, and other cancers. Large human epidemiologic studies have found that the incidental use of NSAIDs is associated with significant reductions in relative risk for developing these and other cancers. Chronic low-dose aspirin does not appear to have a substantial impact on cancer incidence; however, it is associated with reduced cancer death in a number of studies. The anti-cancer efficacy of aspirin in humans may be related to hyperactivity of the PI3 kinase/Akt pathway in tumor cells. In patients with familial polyposis coli, COX inhibitors significantly decrease polyp formation. Polymorphisms in COX-2 have been associated with increased risk of some cancers. Several studies have suggested that COX-2 expression is associated with markers of tumor progression in breast cancer. In mouse mammary tissue, COX-2 is oncogenic whereas NSAID use is associated with a reduced risk of breast cancer in women, especially for hormone receptor-positive tumors. Despite the support for COX-2 as the predominant source of oncogenic prostaglandins, randomized clinical trials have not been performed to determine whether superior anti-oncogenic effects occur with selective inhibition of COX-2, compared with nonselective NSAIDs. Indeed data from animal models and epidemiologic studies in humans are consistent with a role for COX-1 as well as COX-2 in the production of oncogenic prostanoids.

PGE2, which is considered the principal oncogenic prostanoid, facilitates tumor initiation, progression, and metastasis through multiple biologic effects, increasing proliferation and angiogenesis, inhibiting apoptosis, augmenting cellular invasiveness, and modulating immunosuppression. Augmented expression of mPGES-1 is evident in tumors, and preclinical studies support the potential use of mPGES-1 inhibitors in chemoprevention or treatment. In tumors reduced levels of OATP2A1 and 15-PGDH, which mediate cellular uptake and metabolic inactivation of PGE2, respectively, likely contribute to sustained PGE2 activity. The pro- and anti-oncogenic roles of other prostanoids remain under investigation, with TXA2 emerging as another likely procarcinogenic mediator, deriving either from macrophage COX-2 or platelet COX-1. Studies in mice lacking EP1, EP2, or EP4 receptors confirm reduced disease in multiple carcinogenesis models. EP3, in contrast, plays no role or may even play a protective role in some cancers. Transactivation of epidermal growth factor receptor (EGFR) has been linked with the oncogenic activity of PGE2. PGD2, acting on the DP1 receptor, may reduce angiogenesis thereby reducing tumor progression.

Effects of Lipoxygenase & Cytochrome P450-Derived Metabolites

Lipoxygenases generate compounds that can regulate specific cellular responses that are important in inflammation and immunity. Cytochrome P450-derived metabolites affect nephron transport functions either directly or via metabolism to active compounds (see below). The biologic functions of the various forms of hydroxy- and hydroperoxyeicosaenoic acids are largely unknown, but their pharmacologic potency is impressive.

A. Blood Cells and Inflammation

LTB4, acting at the BLT1 receptor, is a potent chemoattractant for T lymphocytes, neutrophils, eosinophils, monocytes, and possibly mast cells. LTB4 also contributes to activation of neutrophils and eosinophils, and to monocyte-endothelial adhesion. The cysteinyl leukotrienes are potent chemoattractants for eosinophils and T lymphocytes. Cysteinyl leukotrienes may also generate distinct sets of cytokines through activation of mast cell cysLT1 and cysLT2. At higher concentrations, these leukotrienes also promote eosinophil adherence, degranulation, cytokine or chemokine release, and oxygen radical formation. Cysteinyl leukotrienes also contribute to inflammation by increasing endothelial permeability, thus promoting migration of inflammatory cells to the site of inflammation. The leukotrienes have been strongly implicated in the pathogenesis of inflammation, especially in chronic diseases such as asthma and inflammatory bowel disease.

Lipoxins have diverse effects on leukocytes, including activation of monocytes and macrophages and inhibition of neutrophil, eosinophil, and lymphocyte activation. Both lipoxin A and lipoxin B inhibit natural killer cell cytotoxicity.

B. Heart and Smooth Muscle

1. Cardiovascular12(S)-HETE promotes vascular smooth muscle cell proliferation and migration at low concentrations; it may play a role in myointimal proliferation that occurs after vascular injury such as that caused by angioplasty. Its stereoisomer, 12(R)-HETE, is not a chemoattractant, but is a potent inhibitor of the Na+/K+-ATPase in the cornea. In vascular smooth muscle LTB4 may cause vasoconstriction as well as smooth muscle cell migration and proliferation, possibly contributing to atherosclerosis and injury-induced neointimal proliferation. LTC4 and LTD4 reduce myocardial contractility and coronary blood flow, leading to depression of cardiac output. Lipoxin A and lipoxin B exert coronary vasoconstrictor effects in vitro. In addition to their vasodilatory action, EETs may reduce cardiac hypertrophy as well as systemic and pulmonary vascular smooth muscle proliferation and migration.

2. GastrointestinalHuman colonic epithelial cells synthesize LTB4, a chemoattractant for neutrophils. The colonic mucosa of patients with inflammatory bowel disease contains substantially increased amounts of LTB4. It appears that activation of the BLT2 receptor, possibly by agonists other than LTB4, is protective in colonic epithelium and contributes to maintenance of barrier function.

3. AirwaysThe cysteinyl leukotrienes, particularly LTC4 and LTD4, are potent bronchoconstrictors and cause increased microvascular permeability, plasma exudation, and mucus secretion in the airways. Controversies exist over whether the pattern and specificity of the leukotriene receptors differ in animal models and humans. LTC4-specific receptors have not been found in human lung tissue, whereas both high- and low-affinity LTD4 receptors are present.

C. Renal System

There is substantial evidence for a role of the epoxygenase products in regulating renal function although their exact role in the human kidney remains unclear. Both 20-HETE and the EETs are generated in renal tissue. 20-HETE, which potently blocks the smooth muscle cell Ca2+-activated K+ channel and leads to vasoconstriction of the renal arteries, has been implicated in the pathogenesis of hypertension. In contrast, studies support an antihypertensive effect of the EETs because of their vasodilating and natriuretic actions. EETs increase renal blood flow and may protect against inflammatory renal damage by limiting glomerular macrophage infiltration. Inhibitors of soluble epoxide hydrolase, which prolong the biologic activities of the EETs, are being developed as potential new antihypertensive drugs. In vitro studies, and work in animal models, support targeting soluble epoxide hydrolase for blood pressure control, although the potential for pulmonary vasoconstriction and tumor promotion through antiapoptotic actions require careful investigation.

D. Miscellaneous

The effects of these products on the reproductive organs have not been elucidated.

Similarly, actions on the nervous system have been suggested but not confirmed. 12-HETE stimulates the release of aldosterone from the adrenal cortex and mediates a portion of the aldosterone release stimulated by angiotensin II but not that by adrenocorticotropic hormone. Very low concentrations of LTC4 increase and higher concentrations of arachidonate-derived epoxides augment luteinizing hormone (LH) and LH-releasing hormone release from isolated rat anterior pituitary cells.


Corticosteroids block all the known pathways of eicosanoid synthesis, perhaps in part by stimulating the synthesis of several inhibitory proteins collectively called annexins or lipocortins. They inhibit phospholipase A2 activity, probably by interfering with phospholipid binding, thus preventing the release of arachidonic acid.

The NSAIDs (eg, indomethacin, ibuprofen; see Chapter 36) block both prostaglandin and thromboxane formation by reversibly inhibiting COX activity. The traditional NSAIDs are not selective for COX-1 or COX-2. The more recent, purposefully designed selective COX-2 inhibitors vary—as do the older drugs—in their degree of selectivity. Indeed, there is considerable variability between (and within) individuals in the selectivity attained by the same dose of the same NSAID. Aspirin is an irreversible COX inhibitor. In platelets, which lack nuclei, COX-1 (the only isoform expressed in mature platelets) cannot be restored via protein biosynthesis, resulting in extended inhibition of TXA2 biosynthesis.

EP-receptor agonists and antagonists are under evaluation in the treatment of bone fracture and osteoporosis, whereas TP-receptor antagonists are being investigated for usefulness in the treatment of cardiovascular syndromes. Direct inhibition of PGE2 biosynthesis through selective inhibition of the inducible mPGES-1 isoform is also under examination for potential therapeutic efficacy in pain and inflammation, cardiovascular disease, and chemoprevention of cancer.

Although they remain less effective than inhaled corticosteroids, a 5-LOX inhibitor (zileuton) and selective antagonists of the CysLT1 receptor for leukotrienes (zafirlukast, montelukast, and pranlukast; see Chapter 20) are used clinically in mild to moderate asthma. Growing evidence for a role of the leukotrienes in cardiovascular disease has expanded the potential clinical applications of leukotriene modifiers. Conflicting data have been reported in animal studies depending on the disease model used and the molecular target (5-LOX versus FLAP). Human genetic studies demonstrate a link between cardiovascular disease and polymorphisms in the leukotriene biosynthetic enzymes, and indicate an interaction between the 5-LOX and COX-2 pathways, in some populations.

NSAIDs usually do not inhibit lipoxygenase activity at concentrations attained clinically that inhibit COX activity. In fact, by preventing arachidonic acid conversion via the COX pathway, NSAIDs may cause more substrate to be metabolized through the lipoxygenase pathways, leading to an increased formation of the inflammatory and proliferative leukotrienes. Even among the COX-dependent pathways, inhibiting the synthesis of one derivative may increase the synthesis of an enzymatically related product. Therefore, drugs that inhibit both COX and lipoxygenase are being developed.


Several approaches have been used in the clinical application of eicosanoids. First, stable oral or parenteral long-acting analogs of the naturally occurring prostaglandins have been developed (Figure 18–5). Second, enzyme inhibitors and receptor antagonists have been developed to interfere with the synthesis or effects of the eicosanoids. The discovery of COX-2 as a major source of inflammatory prostanoids led to the development of selective COX-2 inhibitors in an effort to preserve the gastrointestinal and renal functions directed through COX-1, thereby reducing toxicity. However, it is apparent that the marked decrease in biosynthesis of PGI2 that follows COX-2 inhibition occurring without a concurrent inhibition of platelet COX-1-derived TXA2 removes a protective constraint on endogenous mediators of cardiovascular dysfunction and leads to an increase in cardiovascular events in patients taking selective COX-2 inhibitors. Third, efforts at dietary manipulation—to change the polyunsaturated fatty acid precursors in the cell membrane phospholipids and so change eicosanoid synthesis—is used extensively in over-the-counter products and in diets emphasizing increased consumption of cold-water fish.


FIGURE 18–5 Chemical structures of some prostaglandins and prostaglandin analogs currently in clinical use.

Female Reproductive System

Studies with knockout mice have confirmed a role for prostaglandins in reproduction and parturition. COX-1-derived PGF appears important for luteolysis, consistent with delayed parturition in COX-1-deficient mice. A complex interplay between PGF and oxytocin is critical to the onset of labor. EP2 receptor-deficient mice demonstrate a preimplantation defect, which underlies some of the breeding difficulties seen in COX-2 knockouts.

A. Abortion

PGE2 and PGF have potent oxytocic actions. The ability of the E and F prostaglandins and their analogs to terminate pregnancy at any stage by promoting uterine contractions has been adapted to common clinical use. Many studies worldwide have established that prostaglandin administration efficiently terminates pregnancy. The drugs are used for first- and second-trimester abortion and for priming or ripening the cervix before abortion. These prostaglandins appear to soften the cervix by increasing proteoglycan content and changing the biophysical properties of collagen.

Dinoprostone, a synthetic preparation of PGE2, is administered vaginally for oxytocic use. In the USA, it is approved for inducing abortion in the second trimester of pregnancy, for missed abortion, for benign hydatidiform mole, and for ripening of the cervix for induction of labor in patients at or near term (see below). Dinoprostone stimulates the contraction of the uterus throughout pregnancy. As the pregnancy progresses, the uterus increases its contractile response, and the contractile effect of oxytocin is potentiated as well. Dinoprostone also directly affects the collagenase of the cervix, resulting in softening. Dinoprostone is metabolized in local tissues and on the first pass through the lungs (about 95%). The metabolites are mainly excreted in the urine. The plasma half-life is 2.5–5 minutes.

For abortifacient purposes, the recommended dosage is a 20-mg dinoprostone vaginal suppository repeated at 3- to 5-hour intervals depending on the response of the uterus. The mean time to abortion is 17 hours, but in more than 25% of cases, the abortion is incomplete and requires additional intervention.

Antiprogestins (eg, mifepristone) have been combined with an oral oxytocic synthetic analog of PGE1 (misoprostol) to produce early abortion. This regimen is available in the USA and Europe (see Chapter 40). The ease of use and the effectiveness of the combination have aroused considerable opposition in some quarters. The major toxicities are cramping pain and diarrhea. The oral and vaginal routes of administration are equally effective, but the vaginal route has been associated with an increased incidence of sepsis, so the oral route is now recommended.

An analog of PGF is also used in obstetrics. This drug, carboprost tromethamine (15-methyl-PGF; the 15-methyl group prolongs the duration of action) is used to induce second-trimester abortions and to control postpartum hemorrhage that is not responding to conventional methods of management. The success rate is approximately 80%. It is administered as a single 250-mcg intramuscular injection, repeated if necessary. Vomiting and diarrhea occur commonly, probably because of gastrointestinal smooth muscle stimulation. In some patients transient bronchoconstriction can occur. Transient elevations in temperature are seen in approximately one eighth of patients.

B. Facilitation of Labor

Numerous studies have shown that PGE2, PGF, and their analogs effectively initiate and stimulate labor, but PGF is one tenth as potent as PGE2. There appears to be no difference in the efficacy of PGE2and PGF when they are administered intravenously; however, the most common usage is local application of PGE2 analogs (dinoprostone) to promote labor through ripening of the cervix. These agents and oxytocin have similar success rates and comparable induction-to-delivery intervals. The adverse effects of the prostaglandins are moderate, with a slightly higher incidence of nausea, vomiting, and diarrhea than that produced by oxytocin. PGF has more gastrointestinal toxicity than PGE2. Neither drug has significant maternal cardiovascular toxicity in the recommended doses. In fact, PGE2 must be infused at a rate about 20 times faster than that used for induction of labor to decrease blood pressure and increase heart rate. PGF is a bronchoconstrictor and should be used with caution in women with asthma; however, neither asthma attacks nor bronchoconstriction have been observed during the induction of labor. Although both PGE2 and PGF pass the fetoplacental barrier, fetal toxicity is uncommon.

For the induction of labor or softening of the cervix, dinoprostone is used either as a gel (0.5 mg PGE2 every 6 hours; maximum 24-hour cumulative dose of 1.5 mg) or as a controlled-release vaginal insert (10 mg PGE2) that releases PGE2 over 12 hours. The softening of the cervix for induction of labor substantially shortens the time to onset of labor and the delivery time. An advantage of the controlled-release formulation is a lower incidence of gastrointestinal effects (< 1% versus 5.7%).

The effects of oral PGE2 administration (0.5–1.5 mg/h) have been compared with those of intravenous oxytocin and oral demoxytocin, an oxytocin derivative, in the induction of labor. Oral PGE2 is superior to the oral oxytocin derivative and in most studies is as efficient as intravenous oxytocin. Oral PGF causes too much gastrointestinal toxicity to be useful by this route.

Theoretically, PGE2 and PGF should be superior to oxytocin for inducing labor in women with preeclampsia-eclampsia or cardiac and renal diseases because, unlike oxytocin, they have no antidiuretic effect. In addition, PGE2has natriuretic effects. However, the clinical benefits of these effects have not been documented. In cases of intrauterine fetal death, the prostaglandins alone or with oxytocin seem to cause delivery effectively.

C. Dysmenorrhea

Primary dysmenorrhea is attributable to increased endometrial synthesis of PGE2 and PGF during menstruation, with contractions of the uterus that lead to ischemic pain. NSAIDs successfully inhibit the formation of these prostaglandins (see Chapter 36) and so relieve dysmenorrhea in 75–85% of cases. Some of these drugs are available over the counter. Aspirin is also effective in dysmenorrhea, but because it has low potency and is quickly hydrolyzed, large doses and frequent administration are necessary. In addition, the acetylation of platelet COX, causing irreversible inhibition of platelet TXA2 synthesis, may increase the amount of menstrual bleeding.

Male Reproductive System

Intracavernosal injection or transurethral suppository therapy with alprostadil (PGE1) is a second-line treatment for erectile dysfunction. Injected doses are 2.5–25 mcg; suppositories are recommended to start at 125 mcg or 250 mcg up to 1000 mcg. Penile pain is a frequent side effect, which may be related to the algesic effects of PGE derivatives; however, only a few patients discontinue the use because of pain. Prolonged erection and priapism are side effects that occur in less than 4% of patients and are minimized by careful titration to the minimal effective dose. When given by injection, alprostadil may be used as monotherapy or in combination with either papaverine or phentolamine.

Renal System

Increased biosynthesis of prostaglandins has been associated with one form of Bartter’s syndrome. This is a rare disease characterized by low-to-normal blood pressure, decreased sensitivity to angiotensin, hyperreninemia, hyperaldosteronism, and excessive loss of K+. There also is an increased excretion of prostaglandins, especially PGE metabolites, in the urine. After long-term administration of COX inhibitors, sensitivity to angiotensin, plasma renin values, and the concentration of aldosterone in plasma return to normal. Although plasma K+ rises, it remains low, and urinary wasting of K+ persists. Whether an increase in prostaglandin biosynthesis is the cause of Bartter’s syndrome or a reflection of a more basic physiologic defect is not yet known.

Cardiovascular System

A. Pulmonary Hypertension

PGI2 lowers peripheral, pulmonary, and coronary vascular resistance. It has been used to treat primary pulmonary hypertension as well as secondary pulmonary hypertension, which sometimes occurs after mitral valve surgery. In addition, prostacyclin has been used successfully to treat portopulmonary hypertension, which arises secondary to liver disease. The first commercial preparation of PGI2(epoprostenol) approved for treatment of primary pulmonary hypertension improves symptoms, prolongs survival, and delays or prevents the need for lung or lung-heart transplantation. Side effects include flushing, headache, hypotension, nausea, and diarrhea. The extremely short plasma half-life (3–5 minutes) of epoprostenol necessitates continuous intravenous infusion through a central line for long-term treatment, which is its greatest limitation. Intravenous infusion of epoprostenol is increased in a graded dose-dependent manner, based on recurrence, persistence, or worsening of symptoms. Several prostacyclin analogs with longer half-lives have been developed and used clinically. Iloprost (half-life about 30 minutes) is usually inhaled six to nine times per day (2.5–5 mcg/dose), although it has been delivered by intravenous administration outside the USA. Treprostinil (half-life about 4 hours) may be delivered by subcutaneous or intravenous infusion or by inhalation. Other drugs used in pulmonary hypertension are discussed in Chapter 17.

B. Peripheral Vascular Disease

A number of studies have investigated the use of PGE1 and PGI2 compounds in Raynaud’s phenomenon and peripheral arterial disease. However, these studies are mostly small and uncontrolled, and these therapies do not have an established place in the treatment of peripheral vascular disease.

C. Patent Ductus Arteriosus

Patency of the fetal ductus arteriosus depends on COX-2-derived PGE2 acting on the EP4 receptor. At birth, reduced PGE2 levels, a consequence of increased PGE2 metabolism, allow ductus arteriosus closure. In certain types of congenital heart disease (eg, transposition of the great arteries, pulmonary atresia, pulmonary artery stenosis), it is important to maintain the patency of the neonate’s ductus arteriosus until corrective surgery can be carried out. This can be achieved with alprostadil (PGE1). Like PGE2, PGE1 is a vasodilator and an inhibitor of platelet aggregation, and it contracts uterine and intestinal smooth muscle. Adverse effects include apnea, bradycardia, hypotension, and hyperpyrexia. Because of rapid pulmonary clearance (the half-life is about 5–10 minutes in healthy adults and neonates), the drug must be continuously infused at an initial rate of 0.05–0.1 mcg/kg/min, which may be increased to 0.4 mcg/kg/min. Prolonged treatment has been associated with ductal fragility and rupture.

In delayed closure of the ductus arteriosus, COX inhibitors are often used to inhibit synthesis of PGE2 and so close the ductus. Premature infants in whom respiratory distress develops due to failure of ductus closure can be treated with a high degree of success with indomethacin. This treatment often precludes the need for surgical closure of the ductus.


As noted above, eicosanoids are involved in thrombosis because TXA2 promotes platelet aggregation while PGI2, and perhaps also PGE2 and PGD2, are endogenous platelet antagonists. Chronic administration of low-dose aspirin (81 mg/d) selectively and irreversibly inhibits platelet COX-1, and its dominant product TXA2, without modifying the activity of systemic COX-1 or COX-2 (see Chapter 34). TXA2, in addition to activating platelets, amplifies the response to other platelet agonists; hence inhibition of its synthesis inhibits secondary aggregation of platelets induced by adenosine diphosphate, by low concentrations of thrombin and collagen, and by epinephrine. Because their effects are reversible within the typical dosing interval, nonselective NSAIDs (eg, ibuprofen) do not reproduce this effect, although naproxen, because of its variably prolonged half-life, may provide antiplatelet benefit in some individuals. Not surprisingly, given the absence of COX-2 in platelets, selective COX-2 inhibitors do not alter platelet TXA2 biosynthesis and are not platelet inhibitors. However, COX-2-derived PGI2 generation is substantially suppressed during selective COX-2 inhibition, removing a restraint on the cardiovascular action of TXA2, and other platelet agonists. It is highly likely that selective depression of PGI2 generation explains the increase in vascular events, particularly major coronary events, in humans treated with a coxib or diclofenac. High-dose ibuprofen may confer a similar risk, whereas high-dose naproxen appears to be neutral with respect to thrombotic risk. All NSAIDs appear to increase the risk of heart failure.

Overview analyses have shown that low-dose aspirin reduces the secondary incidence of heart attack and stroke by about 25%. However, low-dose aspirin also elevates the low risk of serious gastrointestinal bleeding about twofold over placebo. Low-dose aspirin also reduces the incidence of first myocardial infarction. However, in this case, the benefit versus risk of gastrointestinal bleeding is less clear. The effects of aspirin on platelet function are discussed in greater detail in Chapter 34.

Respiratory System

PGE2 is a powerful bronchodilator when given in aerosol form. Unfortunately, it also promotes coughing, and an analog that possesses only the bronchodilator properties has been difficult to obtain.

PGF and TXA2 are both strong bronchoconstrictors and were once thought to be primary mediators in asthma. Polymorphisms in the genes for PGD2 synthase, both DP receptors, and the TP receptor have been linked with asthma in humans. DP antagonists, particularly those directed against DP2, are being investigated as potential treatments for allergic diseases including asthma. However, the cysteinyl leukotrienes—LTC4, LTD4, and LTE4—probably dominate during asthmatic constriction of the airways. As described in Chapter 20, leukotriene-receptor inhibitors (eg, zafirlukast, montelukast) are effective in asthma. A lipoxygenase inhibitor (zileuton) has also been used in asthma but is not as popular as the receptor inhibitors. It remains unclear whether leukotrienes are partially responsible for acute respiratory distress syndrome.

Corticosteroids and cromolyn are also useful in asthma. Corticosteroids inhibit eicosanoid synthesis and thus limit the amounts of eicosanoid mediator available for release. Cromolyn appears to inhibit the release of eicosanoids and other mediators such as histamine and platelet-activating factor from mast cells.

Gastrointestinal System

The word “cytoprotection” was coined to signify the remarkable protective effect of the E prostaglandins against peptic ulcers in animals at doses that do not reduce acid secretion. Since then, numerous experimental and clinical investigations have shown that the PGE compounds and their analogs protect against peptic ulcers produced by either steroids or NSAIDs. Misoprostol is an orally active synthetic analog of PGE1. The FDA-approved indication is for prevention of NSAID-induced peptic ulcers. The drug is administered at a dosage of 200 mcg four times daily with food. This and other PGE analogs (eg, enprostil) are cytoprotective at low doses and inhibit gastric acid secretion at higher doses. Because it is also an abortifacient, misoprostol is a pregnancy category X drug. Misoprostol use is low, probably because of its adverse effects including abdominal discomfort and occasional diarrhea. Dose-dependent bone pain and hyperostosis have been described in patients with liver disease who were given long-term PGE treatment.

Selective COX-2 inhibitors were developed in an effort to spare gastric COX-1 so that the natural cytoprotection by locally synthesized PGE2 and PGI2 is undisturbed (see Chapter 36). However, this benefit is seen only with highly selective inhibitors and is offset, at least at a population level, by increased cardiovascular toxicity.

Immune System

Cells of the immune system contribute substantially to eicosanoid biosynthesis during an immune reaction. T and B lymphocytes are not primary synthetic sources; however, they may supply arachidonic acid to monocyte-macrophages for eicosanoid synthesis. In addition, there is evidence for eicosanoid-mediated cell-cell interaction by platelets, erythrocytes, leukocytes, and endothelial cells.

PGE2 and PGI2 limit T-lymphocyte proliferation in vitro, as do corticosteroids. PGE2 also inhibits B-lymphocyte differentiation and the antigen-presenting function of myeloid-derived cells, suppressing the immune response. T-cell clonal expansion is attenuated through inhibition of interleukin-1 and interleukin-2 and class II antigen expression by macrophages or other antigen-presenting cells. The leukotrienes, TXA2, and platelet-activating factor stimulate T-cell clonal expansion. These compounds stimulate the formation of interleukin-1 and interleukin-2 as well as the expression of interleukin-2 receptors. The leukotrienes also promote interferon-γ release and can replace interleukin-2 as a stimulator of interferon-γ. PGD2 induces chemotaxis and migration of TH2 lymphocytes. These in vitro effects of the eicosanoids agree with in vivo findings in animals with acute organ transplant rejection.

A. Inflammation

Aspirin has been used to treat arthritis of all types for approximately 100 years, but its mechanism of action—inhibition of COX activity—was not discovered until 1971. COX-2 appears to be the form of the enzyme most associated with cells involved in the inflammatory process, although, as outlined above, COX-1 also contributes significantly to prostaglandin biosynthesis during inflammation. Aspirin and other anti-inflammatory agents that inhibit COX are discussed in Chapter 36.

B. Rheumatoid Arthritis

In rheumatoid arthritis, immune complexes are deposited in the affected joints, causing an inflammatory response that is amplified by eicosanoids. Lymphocytes and macrophages accumulate in the synovium, whereas leukocytes localize mainly in the synovial fluid. The major eicosanoids produced by leukocytes are leukotrienes, which facilitate T-cell proliferation and act as chemoattractants. Human macrophages synthesize the COX products PGE2 and TXA2 and large amounts of leukotrienes.


Latanoprost, a stable long-acting PGF derivative, was the first prostanoid used for glaucoma. The success of latanoprost has stimulated development of similar prostanoids with ocular hypotensive effects, and bimatoprost, travoprost, and unoprostone are now available. These drugs act at the FP receptor and are administered as drops into the conjunctival sac once or twice daily. Adverse effects include irreversible brown pigmentation of the iris and eyelashes, drying of the eyes, and conjunctivitis.


Dietary intake of linoleic and α-linolenic acids, which are, respectively, omega-6 and omega-3 essential fatty acids, can modify arachidonic acid metabolism and the nature of the eicosanoids produced. Therefore, the effects of dietary manipulation on arachidonic acid metabolism have been extensively studied. Two approaches have been used. The first adds corn, safflower, and sunflower oils, which contain linoleic acid (C18:2), to the diet, allowing for generation of 1-series prostaglandins via dihomo-γ-linoleic acid. The second approach adds oils from cold-water fish that contain the omega-3 fatty acids eicosapentaenoic (C20:5) and docosahexaenoic acids (C22:6). Diets high in fish oils have been shown to impact ex vivo indices of platelet and leukocyte function, blood pressure, and triglycerides with different dose-response relationships. There is an abundance of epidemiologic data relating diets high in fatty fish to a reduction in the incidence of myocardial infarction and sudden cardiac death although there is more ambiguity about stroke. Of course, epidemiologic data may confound such diets with a reduction in saturated fats and other elements of a “healthy” lifestyle and negative overview analyses have raised questions regarding the cardiovascular benefit of dietary omega-3 fatty acids. Some data from prospective randomized trials suggest that such dietary interventions may reduce the incidence of sudden death while experiments in vitro suggest that fish oils protect against experimentally induced arrhythmogenesis, platelet aggregation, vasomotor spasm, and dyslipidemias.





Bäck M et al: International Union of Basic and Clinical Pharmacology. LXXXIV: Leukotriene receptor nomenclature, distribution, and pathophysiological functions. Pharmacol Rev 2011;63:539.

Bhala N et al: Vascular and upper gastrointestinal effects of non-steroidal anti-inflammatory drugs: Meta-analyses of individual participant data from randomised trials. Lancet 2013;382:769.

Christin-Maitre S, Bouchard P, Spitz IM: Medical termination of pregnancy. N Engl J Med 2000;342:946.

Grosser T, Yu Y, Fitzgerald GA: Emotion recollected in tranquility: Lessons learned from the COX-2 saga. Annu Rev Med 2010;61:17.

Hao CM, Breyer MD: Physiological regulation of prostaglandins in the kidney. Annu Rev Physiol 2008;70:357.

Rokach J et al: Nomenclature of isoprostanes: A proposal. Prostaglandins 1997;54:853.

Smith WL, Urade Y, Jakobsson PJ: Enzymes of the cyclooxygenase pathways of prostanoid biosynthesis. Chem Rev 2011;111:5821.

Smyth EM: Thromboxane and the thromboxane receptor in cardiovascular disease. Clin Lipidol 2010;5:209.

Smyth EM et al: Prostanoids in health and disease. J Lipid Res 2009;50:S423.

Wang D, Dubois RN: Eicosanoids and cancer. Nat Rev Cancer 2010;10:181.

Waxman AB, Zamanian RT: Pulmonary arterial hypertension: New insights into the optimal role of current and emerging prostacyclin therapies. Am J Cardiol 2013;111(5 Suppl):1A.

Woodward DF, Jones RL Narumiya S: International Union of Basic and Clinical Pharmacology. LXXXIII: Classification of prostanoid receptors, updating 15 years of progress. Pharmacol Rev 2011;63:471.