871

functions while COX-2 expression is increased during inflammation and other pathologic situations. Inhibition of COX-2 by NSAIDs blocks PG production at sites of inflam-mation while inhibition of COX-1 in certain other tissues, most importantly platelets and the gastroduodenal mucosa, can lead to common adverse effects of NSAIDs such as bleeding, bruising, and gastrointestinal (GI) ulceration.

In addition to their use in rheumatoid arthritis and osteoarthritis, NSAIDs are widely used in the symptomatic management of other rheumatic diseases characterized by chronic musculoskeletal pain and diverse forms of acute pain. Aspirin, which has unique properties among NSAIDs, is used by millions more for primary and secondary prevention of cardiovascular thrombosis. In light of the widespread use of these drugs for common diseases, which are likely to increase in prevalence with the aging of the population, it is critically important to appreciate the potential adverse events and drug interactions associated with NSAIDs.

This chapter analyzes aspirin and other NSAIDs on the basis of chemical structure, pharmacologic properties, and relative inhibition of COX-1 and COX-2. Particular atten-tion to potential adverse events of specific NSAIDs in indi-vidual patients will facilitate use of these drugs in the safest possible manner. Acetaminophen (known as paracetamol outside the United States), an antipyretic and analgesic drug without anti-inflammatory activity, inhibits COX enzymes by a different mechanism than NSAIDs and is also discussed. Colchicine possesses anti-inflammatory charac-teristics similar to NSAIDs in some situations and is dis-cussed in this chapter, although this drug differs in its mechanism of action and profile of adverse effects.

HISTORY

Botanicals containing salicylates have been used since antiquity to treat pain, inflammation, and fever. The Egyp-tian Ebers papyrus recommended use of a decoction of dried myrtle leaves to be applied to the abdomen and back for relief of rheumatic pains about 3500 years ago. A thousand years later, Hippocrates recommended poplar tree juices for eye disease treatment and willow bark to alleviate fever and the pain of childbirth. Throughout Roman times, the use of

KEY POINTS

Nonsteroidal anti-inflammatory drugs (NSAIDs) are effective anti-inflammatory, antipyretic, and analgesic compounds.

There is little difference in the efficacy of the various NSAIDs, but the pharmacologic characteristics of individual drugs including potency, half-life, and relative inhibition of cyclooxygenase (COX)-1 and COX-2 play important roles in toxicity.

Aspirin is an NSAID used in low doses to prevent cardiovascular disease. Aspirin and NSAIDs taken together are associated with increased toxicity in the gastrointestinal tract, and concomitant use of some NSAIDs with aspirin may be associated with aspirin resistance.

NSAIDs are associated with risk for gastrointestinal ulceration and bleeding. Patient-specific risk factors for gastrointestinal toxicity should be recognized in order to implement risk-reduction strategies.

NSAIDs are associated with an elevated risk for cardiovascular disease. Awareness of cardiovascular risk factors and either avoiding NSAIDs or using intermittent, low-dose, short half-life drugs are advisable.

Periodic assessment of blood pressure, hemoglobin, electrolytes, renal function tests, and liver function tests is advisable, particularly in elderly patients.

The use of nonsteroidal anti-inflammatory drugs (NSAIDs) is ubiquitous in the practice of medicine because of their effectiveness as anti-inflammatory, analgesic, and anti-pyretic agents. NSAIDs differ widely in their chemical class but share the property of blocking production of prostaglan-dins (PGs). This is accomplished by inhibiting the activity of the enzyme prostaglandin G/H synthase (PGHS), also called cyclooxygenase (COX).

The clinical effects of NSAIDs are evaluated not only by their specific pharmacologic properties but also in terms of their effects on the different COX isoforms, COX-1 and COX-2. These isoforms serve different biologic functions in that COX-1 is expressed under basal conditions and is involved in the biosynthesis of PGs serving homeostatic

PART

8PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

59 Prostanoid Biology and Its Therapeutic TargetingLESLIE J. CROFFORD

872 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

botanical treatments including willow bark for pain and inflammation was widespread. The medicinal use of salicylate-containing plants occurred in China and other parts of Asia. In addition, the curative effects of other botanicals were known to the indigenous populations of North America. Colchicine-containing extracts of the autumn crocus plant were used for treatment of acute gout as early as the sixth century ad.1

The first modern report of the therapeutic application of salicylate-containing plants was reported to the Royal Society of London by the Reverend Edward Stone, who provided an account of the success of the dried bark of the willow for fever.1 In this first clinical trial, a pound of bark was dried, pulverized, and put into the tea, beer, or water of 50 people with fever. He found that one dose (1 dram = 1.8 g) cured their fever. In 1763 Stone wrote, I have no other motives for publishing this valuable specific, than that it may have a fair and full trial in all its variety of circum-stances and situations, and that the world may reap the benefits accruing from it.

In 1860 salicylic acid was chemically synthesized, which led to its widespread use as an external antiseptic, anti-pyretic, and analgesic.1 The bitter taste of salicylic acid prompted the chemist Felix Hoffman to synthesize the more palatable acetylsalicylic acid (ASA). After demonstration of its anti-inflammatory effects, Dr. Heinrich Dreser of Bayer introduced this compound into medicine in 1899 as aspirin and it remains the most widely used drug in the world.1 Salicylate was identified as the active ingredient of willow bark in 1929.

Phenylbutazone came into clinical practice in 1949 and was followed by indomethacin, fenamates, naproxen, and others. Despite the diversity of their chemical structures, these drugs shared therapeutic properties with aspirin. Fur-thermore, adverse events including gastric upset, GI ulcer-ation and bleeding, hypertension, edema, and renal damage were shared by all these drugs. In 1971 it was discovered that these drugs all acted by inhibiting PG biosynthesis, thereby providing a unifying explanation of their therapeu-tic actions and a rationale for grouping them together as NSAIDs.1

COX was isolated in 1976 from the endoplasmic reticu-lum of PG-forming cells.2,3 However, several groups specu-lated that there must be a second COX enzyme on the basis of observed biology. In 1990 investigators demonstrated that bacterial lipopolysaccharide (LPS) increased PG syn-thesis in human monocytes in vitro and in mouse peritoneal macrophages in vivo, but only the LPS-induced increase was inhibited by dexamethasone and required the de novo synthesis of new COX protein.4 This observation was the foundation of the concept for constitutive and inducible forms of COX. Soon thereafter a number of investigators working in different systems reported the discovery of an inducible second form of COX.3 Investigators went on to clone the gene and deduced its structure, and they found the gene product was homologous to COX, but to no other known protein. The observation that glucocorticoids inhib-ited the expression of COX-2 following a proinflammatory stimulus represented a link between the anti-inflammatory actions of NSAIDs and corticosteroids.

Because of the prediction that inhibiting COX-2 would block PG biosynthesis participating in the inflammatory

response but was not required for homeostasis, there was a tremendous push to develop drugs that specifically inhibit COX-2 without effect on COX-1 in the belief that these medications would provide clinical efficacy without adverse effects.2,5 Identification of new drugs that differentially inhibited COX-2 over COX-1 was accomplished quickly as existing NSAIDs were tested on the two COX isoforms and crystal structures revealed differences in the protein structures on which new drug development could be based.5,6

One hundred years after aspirin was introduced and 10 years after the discovery of COX-2, selective COX-2 inhibi-tors, celecoxib (Celebrex) and rofecoxib (Vioxx), were developed. In clinical trials, the safety and efficacy profiles of these and related drugs showed promise and the U.S. Food and Drug Administration (FDA) subsequently approved these COX-2-selective NSAIDs for treatment of arthritis and pain. After the introduction into clinical prac-tice, however, it became clear that the most highly COX-2-selective NSAIDs, particularly rofecoxib, were more likely than traditional NSAIDs to be associated with adverse cardiovascular events.7 This finding led to the voluntary withdrawal of rofecoxib and several other COX-2-selective NSAIDs from the market. Debate surrounding the relative risks of different NSAIDs to specific organ systems contin-ues to the present.

CYCLOOXYGENASE BIOLOGY AND BIOACTIVE LIPIDS

Therapeutic and adverse effects of NSAIDs are best under-stood in the context of COX biology. The COX enzymes are the first committed step in the synthesis of PG from arachidonic acid (AA) (Figure 59-1). AA is an omega-6 polyunsaturated fatty acid (PUFA) commonly found at the sn-2 position of cell membrane glycerophospholipids and cleaved from cell membranes by one of several different phospholipase A2 enzymes.8 Once generated, AA may enter several different pathways to generate bioactive lipids. It can be metabolized by COX to PGG2 and then PGH2 via its COX and peroxidase enzyme activities; by lipoxygenases (LOX) to hydroxyeicosatetraenoic acids (HETEs), hydro-peroxyeicosatetraenoic acids (HPETEs), or leukotrienes; or by the cytochrome p450 family of enzymes to HETEs, HPETEs, or epoxyeicosatrienoic acids (EETs).9,10 These bio-active lipids have diverse biologic activities in normal phys-iology and in pathologic conditions including inflammation, pain, cardiovascular disease, and cancer.

In the COX pathway, the unstable intermediate PGH2 spontaneously rearranges or is enzymatically converted by specific synthases to biologically active PG, of which there are many isoforms.3 Although phospholipase A2 activity is required to initiate PG synthesis, the overall regulation of the type and amount of PG produced in a given cell or tissue is determined by the expression levels of COX-1, COX-2, and terminal synthase enzymes.

Bioactive lipids synthesized via the COX and LOX path-ways are important mediators of inflammation, but alternate substrates and pathways can generate anti-inflammatory lipids and lipids important in the resolution of inflammation (Figure 59-2).11,12 Eicosapentaenoic acid (EPA) and docosa-hexaenoic acid (DHA), which are omega-3 PUFAs, can

873CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

mediating the action of prostanoids. Most cell types produce only one or a few prostanoid species; however, transcellular metabolism of unstable intermediates of COX may mediate qualitative and quantitative changes in the profile of eico-sanoids produced in a given tissue.14 PGs are exported from cells via the multidrug resistanceassociated protein family of efflux transporters, and the PG transporter mediates influx of PG and metabolism to stable products via hydroxy-prostaglandin dehydrogenase (see Figure 59-1).10

PGE2, the most abundant PG at sites of inflammation, can be produced by many different cell types via at least three different PGE synthases.15 Expression of the inducible form of PGE synthase, microsomal PGE synthase-1 (mPGES-1), is similar to expression of COX-2. mPGES-1 acts in concert with COX-2 to produce high levels of PGE2 during inflammation. mPGES-1 expression is blocked by NSAIDs, and experiments demonstrate that PGE2 itself, along with proinflammatory stimuli, is required for mPGES-1 upregulation.16 This positive regulatory loop contributes to the high levels of PGE2 at sites of inflammation and likely to the efficacy of NSAIDs. Other terminal PG synthases are not known to be highly regulated, and levels of other

also serve as substrates for COX and LOX and their metabolic pathways lead to generation of resolvins, protec-tins, and maresins, bioactive lipids associated with resolu-tion of inflammation.12,13 It is hypothesized that generation of anti-inflammatory lipids and lipids that accelerate resolu-tion of inflammation may explain why diets high in omega-3 fatty acids are associated with reduced inflammatory and cardiovascular diseases.11 In addition, ASA possesses the capacity to alter COX-2 activity such that an ASA-triggered alternative catalytic cascade also results in syn-thesis of anti-inflammatory products including lipoxins, protectins, and resolvins (see Figure 59-2).11

Prostaglandin Production and Action

PGs act as autocrine and paracrine mediators with effects limited to the immediate vicinity of their synthesis. That PG actions are limited to their site of synthesis reflects their ephemeral nature and predisposition to metabolic inactiva-tion to regulate promptly their potent biologic activities. Thus the enzymes required for production of a specific pros-tanoid must be co-localized and adjacent to the receptor

Figure 59-1 An overview of eicosanoid synthesis pathways. Arachidonic acid (AA) is a polyunsaturated fatty acid that constitutes the phospholipid domain of most cell membranes and is liberated from the cellular membranes by cytoplasmic phospholipase A2 (PLA2). Free AA can be metabolized to eicosanoids through three major pathways: the cyclooxygenase (COX), lipoxygenase (LOX), and the cytochrome P450 mono-oxygenase pathways. In the COX pathway, the key step is the enzymatic conversion of arachidonic acid to the intermediate prostaglandin G2 (PGG2), which is then reduced to an intermediate PGH2 by the peroxidase activity of COX. PGH2 is subsequently metabolized to prostanoids including PGs and thromboxanes (TX) by specific synthases. LOX converts AA to biologically active metabolites such as leukotrienes (LT) and hydroxyeicosatetraenoic acids (HETE). P450 metabolizes AA to epoxyeicosatrienoic acids (EET). In the 5-LOX pathway, AA is converted to an intermediary 5-HPETE, which is converted to LTA4. LTA4 is subsequently converted to 5-HETE, LTB4, LTC4, LTD4, and LTE4. Each of the PG and LT exerts its biologic effects by binding to its cognate G proteincoupled receptor. PGI2 can transactivate the nuclear peroxisome proliferator-activated receptor (PPAR), and a PGD2 dehydration product, 15dPGJ2, is a natural ligand for PPAR. The multidrug resistanceassociated protein (MRP) gene family is composed of efflux transporters for both PG and LT. PG transporter (PGT) is an influx transporter for PG. Hydroxyprostaglandin dehydrogenase 15-(NAD) (15-PGDH) mainly metabolizes intracellular PGE2 and PGF2 to stable 13,14-dihydro-15-keto-PGs. (Reprinted with permission from Wang D, Dubois RN: Eicosanoids and cancer, Nat Rev Cancer10:181193, 2010.)

15-HETE5-LOX

5-HETE

5-LOX and FLAP

PLA2

COX2

COX1

NSAIDs

COX-2-selective inhibitors

HETEs

HPETEs

Arachidonic acid

Intracellular

Lipoxygenase pathway P450 pathway

PG and TX synthase5-HPETE

15-PGDH

LTA4

PGH2 EETs

PGE2 PGD2 PGI2 TXA2PGE2

PGF2, PGD2, PGI2 or TXA2

13,14-dihydro15-keto-PGF2

15dPGJ2 PPAR PPARPGEM

LTA4hydrolase

LTB4

LTB4 LTC4 orLTD4

PGE2

BLT1 andBLT2

EP1-4 FP, DP, IP or TPCysLT1 andCysLT2

LTC4LTC4 synthase

LTD4

LTE4

12-LOX 15-LOX-1 15-LOX-2(mouseorthologue8-LOX)

12-HETE

8-HETE

MRP4 MRP4MRP1 MRPPGT

874 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

Biochemistry and Structural Biology

COX-1 and COX-2 are bifunctional enzymes that mediate a COX reaction whereby arachidonate plus two molecules of O2 are converted to the cyclic endoperoxide PGG2, fol-lowed by a hydroperoxidase reaction in which PGG2 under-goes a two-electron reduction to PGH2.8 COX enzymes are integral membrane proteins that sit within the inner leaflet of the lipid bilayer of intracellular phospholipid membranes of the nuclear envelope and the endoplasmic reticulum. The crystal structures of COX-1 and COX-2 have been solved, and they have essentially identical domain struc-tures.6 The COX enzymes are homodimers with each monomer consisting of three structural domains. The N-terminal, epidermal growth factorlike domain is involved in dimerization via hydrophobic interactions. The membrane-binding domain is composed of four amphi-pathic -helices lodged into half of the lipid bilayer to form a hydrophobic channel in the center of the large catalytic domain that contains the COX and peroxidase active sites and that constitutes about 80% of the protein. The catalytic domain is globular with two distinct intertwining lobes. The interface of these lobes creates a shallow cleft on the upper surface of the enzyme where the peroxidase active site is located and where heme is bound.

COX and hydroperoxidase reactions occur at distinct but structurally and functionally interconnected sites. The COX reaction is peroxide dependent and requires that the

important bioactive PGs are stably produced or their levels are increased when phospholipase activity and COX-2 levels are increased.

The actions of PG are mediated by cell surface G proteincoupled receptors (GPCRs). There are at least nine known PG receptors with additional splice variants (see Figure 59-1).17 The PG receptors belong to three clusters within a distinct subfamily of the GPCR superfamily with the lone exception of one of the PGD2 receptors (DP2), which belongs to the chemokine receptor subfamily. The relaxant receptors for prostacyclin (IP), PGD2 (DP1), and PGE2 (EP2 and EP4) signal through Gs-mediated increases in intracel-lular cyclic adenosine monophosphate (cAMP). The con-tractile receptors for thromboxane A2 (TP), PGF2 (FP), and PGE2 (EP1) signal through a Gq-mediated increase in intra-cellular calcium. An inhibitory receptor for PGE2 (EP3) couples to Gi and decreases cAMP formation. Note that PGE2 has at least four different receptors with a broad range of potential actions. EP4 in particular appears to mediate many of the proinflammatory activities of PGE2.18 Given the great diversity of PG receptors expressed by different cell types, PG signaling pathways constitute an enormously complex network controlling many biologic actions. Much work remains to understand fully all of the cellular signaling mechanisms by which PGs and their receptors elicit their respective biologic actions. This is particularly true as antagonists for many of these receptors show promise as novel targets for drug development.18

Figure 59-2 Biosynthesis of anti-inflammatory lipids. Function of essential polyunsaturated fatty acids in the production of families of bioactive lipid mediators. A, Arachidonic acid is the precursor of metabolites that function as proinflammatory mediators. Prostaglandins and leukotrienes play pivotal roles in the progression of inflammation. Through cell-cell interactions, exemplified by platelet leukocytes in the vasculature or polymorphonuclear cell-mucosa interactions, or both, lipoxins are generated. They serve as stop signals and promote resolution. They also serve as endogenous anti-inflammatory mediators self-limiting the course of inflammation. The essential omega-3 fatty acids eicosapentaenoic acid and docosahexaenoic acid (C20:5 and C22:6) are converted to new families of lipid mediators that are pivotal in promoting resolution (as in B). Resolvins of the E series such as RvE1 are generated from eicosapentaenoic acid, and resolvins of the D series such as RvD1 and the protectins such as neuroprotectin D1 (NPD1) are generated from docosahexaenoic acid, for which neural systems are enriched. B, Aspirin affects the formation of resolvin E1 by acetylating cyclooxy-genase (COX)-2 in vascular endothelial cells, which, in a stereoselective way, can generate 18 R-H(p)EPE (hydroperoxyeicosapentaenoic acid), which is picked up through transcellular metabolism by leukocytes and converted by a lipoxygenase (LOX)-like mechanism to resolvin E1. Aspirin also affects the formation of D-series resolvins and catalytically switches COX-2 to a 17 R-LOX-like mechanism that serves to generate 17 R-series resolvin D. Aspirin also affects the formation of protectins and neuroprotectins by a similar mechanism and generates compounds carrying the 17 R epimer at the alcohol at carbon-17 in neuroprotectin D1 and other protectins. DHA, docosahexaenoic acid; H(p)DHA, hydroperoxydocosahexaenoic acid. (Modified from Serhan CN, Savill J: Resolution of inflammation: the beginning programs the end, Nat Immunol 6:11911197, 2005.)

A B

Arachidonic acidC20:4

ProstaglandinsLeukotrienesEETsP450

Eicosanoids

Aspirin

LipoxinsEicosapentanoic acid

C20:5

ResolvinsE series

Docosahexanoic acidC22:6

Docosanoids

ResolvinsD seriesProtectinsNeuroprotectin D1

Proinflammatorymediators

Anti-inflammatoryproresolution

EPA 18 R-H(p)EPE LOX

LOX

RvE1

Protectinsneuroprotectins NPD1

Epoxidation

Aspirin:COX-2(acetylated)

DHALOX

17-H(p)DHA

17S-resolvin D series

RvD1

RvD2RvD3RvD4

Aspirin:COX-2

P450

875CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

critical for binding of most carboxylate-containing NSAIDs.19,20 Serine 530 is required for inhibition of COX by the phenylacetic acid NSAID diclofenac. Serine 530 is also the residue transacetylated by ASA and, along with valine 349, seems to govern the stereochemistry of the pocket such that after exposure to aspirin, AA is sterically blocked from a functional interaction with the catalytic domain in COX-1 and catalytic activity is completely inhibited.

A crucial structural difference between COX-1 and COX-2 is the substitution of the small amino acid valine in COX-2 for the isoleucine with a bulky side chain in COX-1 at position 523 that opens a side pocket in the hydrophobic channel in COX-2.6 Overall, COX-2 has a wider and somewhat more flexible interior channel, a struc-tural feature that has been exploited with respect to the development of COX-2-selective NSAIDs as shown in Figure 59-3.21

heme group at the peroxidase site undergo a two-electron oxidation. A tyrosine residue (tyrosine 385) located at the COX active site is involved as a reaction intermediate. The physiologic heme oxidant in vivo is not known, but it has been shown that the COX activity of COX-2 can be acti-vated at 10-fold lower concentrations of hydroperoxide than that of COX-1.8

NSAIDs function by blocking access of AA to the COX active site within a long, narrow, dead-end, hydrophobic channel whose entrance is framed by the four amphipathic helices of the membrane-binding domain. The channel extends into the globular catalytic domain and is about 8 wide. Significant narrowing of the channel occurs where arginine 120 protrudes into the channel. Arginine 120 is essential for binding both AA substrate and most carboxylate-containing NSAIDs in COX-1. By virtue of other differences in the hydrophobic channel, this residue is unessential for binding AA in COX-2, whereas it remains

Figure 59-3 Cyclooxygenase (COX)-1 and COX-2 substrate-binding channels. Schematic depiction of the structural differences between the substrate-binding channels of COX-1 and COX-2 that allowed the design of selective inhibitors. The amino acid residues, Val434, Arg513, and Val523, form a side pocket in COX-2 that is absent in COX-1. A, Nonselective inhibitors have access to the binding channels of both isoforms. B, The more voluminous residues in COX-1, Ile434, His513, and Ile532, obstruct access of the bulky side chains of the coxibs. (From Grosser T, Fries S, FitzGerald GA: Biological basis for the cardiovascular consequences of COX-2 inhibition: therapeutic challenges and opportunities. J Clin Invest 116:415, 2006.)

R

OH

O

Val532

OH

O

IIe532

Arg120Arg513

NN

N

N

NN N

N

Arg120

N

N

N

IIe434Val434

Val532

Arg120

Arg513

NN

N

N N

N

Val434

His513

IIe532

N

N Arg120NN

N

IIe434

His513

OO

OO

COX-1 COX-2

COX-1 COX-2

Active siteActive site

Active site

Active siteR

NonselectiveCOX inhibitor

COX-2selective inhibitor

OH

O

A

B

876 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

important role in normal reproductive, renal, cardiovascu-lar, and skeletal physiology.2,21,27

MECHANISM OF ACTION

Cyclooxygenase Inhibition

All of the NSAIDs are synthetic inhibitors of the COX active site, but subtle mechanistic differences in the manner in which individual NSAIDs interact and bind with the active site are responsible for some of the differences in their pharmacologic characteristics.28 ASA is the only covalent, irreversible modifier of COX-1 and COX-2, whereas all of the other NSAIDs are competitive inhibitors, competing with AA for binding in the active site. The competitive inhibitors are subdivided further on the basis of whether they bind to the COX active site in a time-dependent or time-independent manner.

Crystallographic studies have shown how ASA effec-tively acetylates serine 530 of COX-1. Similar to other NSAIDs, ASA diffuses into the COX-1 active site at the mouth of the channel and travels to the constriction created by arginine 120, where it is in the best orientation to transacetylate serine 530, leading to the complete and irreversible inhibition of COX-1.29 In COX-2, the channel of the active site is larger than COX-1, the orientation of ASA for serine 530 attack is not as good, and transacety-lation efficiency for COX-2 is 10-fold to 100-fold less than for COX-1. ASA can also trigger COX-2 to alter its catalytic activity to produce 15 R-HETE and lipoxins from AA and to generate anti-inflammatory lipids from omega-3-PUFA.11

The time it takes for an NSAID to inhibit the COX active site relative to how long it takes for it to leave the COX channel is a crucial factor in the inhibition of COX.30 Drugs such as ibuprofen exhibit such rapid rates that they essentially inhibit COX instantly but can be removed from the COX active site just as quickly when drug levels decrease. Both COX monomers must be inhibited by ibu-profen to block catalytic activity.22 Conversely, indometha-cin and diclofenac are time-dependent allosteric inhibitors that require seconds to minutes to bind to the COX active site and need only block one of the COX monomers to completely inhibit catalytic activity.22 These NSAIDs also need hours to exit the COX active site. Initially, most tra-ditional time-dependent NSAIDs form a loose complex with the COX active site before a stronger interaction is established. This complex is limited by the time it takes the drug to become properly oriented within the COX channel at arginine 120, the constriction site in the COX channel. This may involve a change in conformation to the open state to allow the drug to access the upper part of the COX catalytic site.

Drugs such as flurbiprofen and indomethacin form a salt bridge between the carboxylate moiety of the NSAID and the guanidinium moiety of arginine 120. Hydrophobic interactions between the aromatic rings and the hydropho-bic amino acids in the channel aid binding. Such interac-tions at the constriction point of the channel completely block the entry of substrate to the active site.31 Diclofenac interacts with serine 530, not arginine 120, but also blocks entry of substrate.32

Both enzymes are homodimers, but the monomers often behave asymmetrically as conformational heterodimers during catalysis and inhibition.22 That is, when a fatty acid binds to one monomer, the other monomer becomes cata-lytically active and only one monomer is catalytically active at any one time. The specific fatty acid bound to the non-catalytic monomer can regulate catalytic activity.23 Differ-ent NSAIDs interact differently with respect to allosteric inhibition of COX enzymes as one facet of their pharmacol-ogy.22 ASA acetylation occurs in only one of the two COX monomers, which completely inhibits the activity of COX-1. However, COX-2 retains the ability to form a reduced amount of PGH2 and alternate aspirin-triggered lipoxins from AA. The anti-inflammatory resolvins may be synthesized by ASA-acetylated COX-2 from omega-3 PUFA.22

Molecular Biology

In addition to the differences in their structures relevant for the pharmacology of COX-1 and COX-2 inhibition, there are physiologically relevant differences with respect to expression and regulation.2,3 Generally, COX-1 is constitu-tively expressed in most cells and its expression is minimally altered by inflammatory stimuli. The promoter region of COX-1 has the characteristics of a gene that is continuously transcribed and stably expressed. COX-1 activity is regu-lated by substrate (AA) availability. When there is an increase in substrate mobilization via phospholipase A2 acti-vation, there is a concordant increase in PG synthesis medi-ated by COX-1. COX-1 is the only isoform expressed in mature platelets and is the most highly expressed COX isoform in normal gastroduodenal mucosa.24 Because COX-1 is inhibited by nonselective NSAIDs, these physiologic properties may explain some of the common adverse effects of these drugs such as bleeding and GI ulceration.

In contrast, the COX-2 gene has the structure of a highly regulated product with binding sites for transcription factors such as nuclear factor B (NFB), cAMP-responsive element, and activating protein-1 (AP-1), which rapidly increase transcription in response to inflammatory signals.2 COX-2 expression is highly induced by proinflammatory cytokines such as tumor necrosis factor and interleukin-1 (IL-1), microbial products, and mitogens2,8 and is inhibited by glucocorticoids.2 COX-2 mRNA stability is a key regula-tor of COX-2 levels. The potential for instability of the COX-2 message is due to the presence of multiple AUUUA instability sequences in the 3 region that mediate a rapid degradation of mRNA, which ultimately suppresses COX-2 protein synthesis and PG production. Conversely, some stimuli including IL-1 may interfere with mRNA degrada-tion and increase COX-2 levels and PG production.25 COX-1 and COX-2 can be post-translationally modified. COX-1 is glycosylated at three asparagines involved in proper protein folding of the enzyme, whereas COX-2 can be glycosylated at four asparagines.25,26

The generalization that COX-1 expression is constitu-tive and COX-2 expression is inducible has its limitations, given that COX-2 is expressed constitutively in several organ systems and regulated by physiologic, as well as patho-logic, stimuli. COX-2 is basally expressed in the brain, kidney, pancreas, and blood vessels and therefore plays an

877CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

be modulated by NSAIDs. Some NSAIDs bind to and acti-vate members of the peroxisome proliferator-activated receptor (PPAR) family and other intracellular receptors. PPAR activation is thought to mediate anti-inflammatory activities. Selective COX-2 inhibitors may have unique structural features that promote COX-independent activities such as cell-cycle regulation, apoptosis, and antiangiogenesis.36

Mechanism of Acetaminophen and Other Analgesic Antipyretic Drugs

Acetaminophen (paracetamol) and dipyrone relieve pain and fever, but they are not anti-inflammatory. The precise mechanisms by which these drugs elicit their effects remain unclear. In the 1970s, it was proposed that acetaminophen worked by means of a central action by inhibiting COX activity primarily in the brain and not in peripheral tissues because they were not acidic and could cross the blood-brain barrier.37 Acetaminophen does inhibit COX-1 and COX-2, but variably so and dependent on cell and tissue type. Acetaminophen does not appear to inhibit by interac-tion with the COX active site; rather, it serves as a reducing co-substrate for the peroxidase site. The peroxide tone of cells and tissues in vivo may be responsible for inhibitor specificity, with platelets and activated macrophages being resistant to the action of acetaminophen and vascular endothelial cells being sensitive to its inhibitory effects on COX. Additionally, the inhibitory potency of acetamino-phen is determined by the concentration of the COX enzyme.37 This may be an additional factor for the lack of clinical anti-inflammatory effects because inflammation is associated with a markedly increased expression of COX- 2 enzyme. With the discovery of a COX-1 splice variant and studies showing that it is both highly expressed in brain and more sensitive to inhibition by acetaminophen, some authors proposed that the analgesic and antipyretic actions of acetaminophen could be explained by its ability to inhibit the COX-1 splice variants (called COX-3 by some despite the fact that this variant does not arise from a unique gene).38 However, more recent studies have rejected this mechanism as explanatory for acetaminophen effects.37,39

Salicylate has analgesic, antipyretic, and anti-inflammatory activity but, similar to acetaminophen and in contrast to ASA, is a poor COX inhibitor. Salicylate has also been shown to inhibit COX activity if substrate levels are low, and it is also dependent on the oxidative state of the enzyme, suggesting that this drug may inhibit COX by redox-related mechanisms.40

PHARMACOLOGY AND DOSING

Classification

NSAIDs are generally grouped according to their chemical structures, plasma half-life, and COX-1 versus COX-2- selectivity (Table 59-1 and Figure 59-4). Table 59-1 presents a representative compilation of common NSAIDs, formula-tions, dosages, half-lives, and precautions. Structurally, most NSAIDs are organic acids with low pK values that lend themselves to their accumulation at sites of inflammation,

COX-2 Selectivity

NSAIDs such as meloxicam, nimesulide, and etodolac show some selectivity for inhibiting COX-2 over COX-1. After the discovery of COX-2, efforts to further enhance COX-2 selectivity led to the development of celecoxib, rofecoxib, valdecoxib, etoricoxib, and lumiracoxib. The prototypical COX-2-selective NSAIDs, celecoxib and rofecoxib, are diaryl compounds containing a sulfonamide (celecoxib) and methylsulfone (rofecoxib) rather than a carboxyl group. Both drugs are weak time-independent inhibitors of COX-1 but strong time-dependent inhibitors of COX-2 that require their entry into and stabilized binding in the catalytic pocket. Because these drugs lack a carboxyl group, arginine 120 is not involved, but multiple sites of hydrogen and hydrophobic binding stabilize drugs at the catalytic site. The sulfur-containing phenyl ring of COX-2-selective NSAIDs plays a pivotal role in binding stability by occupy-ing the hydrophobic side pocket characteristic of the COX-2 catalytic site. If this side pocket is removed by mutagenesis, all isozyme selectivity is lost.6

COX isozyme selectivity is defined most commonly using the concentration of drug required to inhibit PG production by 50% in a particular assay system (inhibitory concentra-tion, or IC, 50). Ratios using values obtained for COX-1 IC50s compared with COX-2 IC50s can be calculated and used as a standard measure for comparing the degrees of selectivity of a particular NSAID for one or the other COX isoform.33 PG assay systems can vary widely, however, making it difficult to compare directly results from studies using different assay systems. To circumvent such problems, most clinicians have accepted the use of the in vitro whole-blood assay to compare NSAID selectivities. In this system, COX-1 inhibition is assessed as a function of the reduction of thromboxane made by platelets after clot formation. Inhibition of COX-2 is based on the inhibition of PGE2 production in a heparinized blood sample after LPS stimula-tion. A COX-2-selective NSAID lacks inhibitory effect on platelet COX-1 at concentrations at or above those that maximally inhibit COX-2.5,34

Cyclooxygenase-Independent Mechanisms of Action

At high, nonphysiologic concentrations, some NSAIDs seem to elicit effects on cellular pathways in vitro that do not involve the inhibition of COX. Because of the high doses of drug required and the use of in vitro systems, the relevance of these effects to in vivo activity is uncertain. Some NSAIDs inhibit phosphodiesterases associated with the metabolism of cAMP leading to increased intracellular cAMP levels and the subsequent general inhibition of peripheral blood lymphocyte responses to mitogen stimula-tion, monocyte and neutrophil migration, and neutrophil aggregation.35 NSAIDs scavenge free radicals, inhibit super-oxide production by polymorphonuclear neutrophils, reduce mononuclear cell phospholipase C activity, and inhibit inducible nitric oxide synthase activity. Sodium salicylate and ASA inhibit NFB activation, as do certain inactive enantiomers of flurbiprofen. Some reports indicate that other cell signaling molecules such as mitogen-activated protein kinases and the transcription factor AP-1 may also

878 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

Table 59-1 Common Nonsteroidal Anti-inflammatory Drugs (NSAIDs)

DrugBrand Name Available Formulations (mg)

Maximal Daily Dose (mg)

Tmax (hr)

Half-life (hr)

Dose Adjustment or Special Precautions

Salicylic Acids

Acetylsalicylic acid

Aspirin Tablets: 81,165, 325, 500, 650 3000 0.5 4-6 Decrease dose by 50% in renal failure patients and patients with hepatic insufficiency

Childrens: 81Suppository: 120, 200, 300, 600

Salsalate Disalcid Capsule: 500 3000 1.4 1Amigesic Tablet: 500, 750Salflex

Diflunisal Dolobid Tablets: 250, 500 1500 2-3 7-15 Decrease dose by 50% in renal failure

Acetic Acids

Diclofenac Voltaren Tablets: 25, 50, 75 225 1-2 2 Incidence of increased transaminase levels higher than with other NSAIDs

Voltaren XR Extended release: 100Cataflam

Diclofenac + misoprostol

Arthrotec Tablets: 50 or 75 plus misoprostol 200 g

200 1-2 2 Incidence of increased transaminase levels higher than with other NSAIDs

Indomethacin Indocin Caps: 25, 50 200 1-4 2-13 Approved for treatment of patent ductus arteriosus

Indocin SR Sustained release: 75Oral suspension: 25 mg/5 mLSuppositories: 50

Sulindac Clinoril Tablets: 150, 200 400 2-4 16 Prodrug metabolized to active compound

Decrease dose in renal disease, liver disease, and elderly patients

Ketorolac Toradol IM/IV: 15 or 30 mg/mL 120 IV/IM 0.3-1 4-6 Decrease dose by 50% in renal failure and elderly patients

Tablets: 10 40 mg PO Do not use > 5 daysTolmetin Tolectin Tablets: 200, 600 1800 0.5-1 1-1.5

Caps: 400Etodolac Lodine Caps: 200, 300 1200 1-2 6-7

Lodine XL Tablets: 400Extended release: 400, 500, 600

Propionic Acids

Ibuprofen Motrin Tablets: 200 (OTC), 300, 400, 600, 800

3200 1-2 2 Avoid in severe hepatic diseaseAdvilNuprenRufen

Naproxen Naprosyn Tablets: 125 (OTC), 250, 375, 500

1500 2-4 12-15 Decrease dose in renal disease, liver disease, and elderly patients

Aleve Sustained release: 375, 500Anaprox Suspension: 125 mg/5 mLEC-NaprosynNaprelan

Fenoprofen Nalfon Caps: 200, 300, 600 3200 1-2 2-3 Idiosyncratic nephropathy more frequent than with other NSAIDs

Ketoprofen Orudis Tablets: 12.5 (OTC) 300 0.5-2 2-4 Decrease dose in severe renal disease, hepatic disease, and elderly patients

Oruvail Caps: 25, 50, 75Sustained release: 100, 150,

200Flurbiprofen Ansaid Tablets: 50, 100 300 1.5-2 3-4Oxaprozin Daypro Tablets: 600 1800 or

26 mg/kg/day

3-6 49-60 Decrease dose in renal failure patients and patients < 50 kg

Fenamic Acids

Meclofenamate Meclomen Caps: 50, 100 400 0.5 2-3

Oxicams

Piroxicam Feldene Caps: 10, 20 20 2-5 3-86 Decrease dose in hepatic disease and elderly patients

Meloxicam Mobic Tab: 7.5, 15 15 5-6 20

Nonacidic Compounds

Nabumetone Relafen Tablets: 500, 750 2000 3-6 24 Food increases peak concentrationReduce dose in renal diseaseAvoid in severe liver diseaseLimit dose to 1 g/day in elderly

patients

879CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

DrugBrand Name Available Formulations (mg)

Maximal Daily Dose (mg)

Tmax (hr)

Half-life (hr)

Dose Adjustment or Special Precautions

COX-2 Selective Inhibitors

Celecoxib Celebrex Caps: 100, 200, 400 400 (800 mg in FAP)

3 11 Contraindicated with sulfonamide allergy

Etoricoxib* Arcoxia Tablets: 60, 90, 120 120 1-1.5 22 Contraindicated in severe renal or liver disease patients

Caution in mild-to-moderate disease

FAP, familial adenomatous polyposis; IM/IV, intramuscular/intravenous; OTC, over the counter; PO. Oral.

Table 59-1 Common Nonsteroidal Anti-inflammatory Drugs (NSAIDs)contd

Figure 59-4 Classification and representative structures of the traditional nonsteroidal anti-inflammatory drugs (NSAIDs) and cyclooxygenase-2 (COX-2)-selective NSAIDs. NSAIDs. *Selected NSAID structure from each subclass.

Carboxylic acids Enolic acids Nonacidiccompounds

OxicamsPyrazolonesFenamic acidsPropionic acidsAcetic acidsSalicylic acids

Phenylaceticacids

Carbo- andheterocyclic

acids

*AspirinDiflunisalTrisalicylateSalsalateSodium salicylate

*Diclofenac *EtodolacIndomethacinSulindacTolmetinKetorolac

FlurbiprofenKetoprofenOxaprozin*IbuprofenNaproxenFenoprofen

*Mefanamic *Phenyl-butazone

Piroxicam*Meloxicam

*Nabumetone

COX-2-selective inhibitors

*Celecoxib Etoricoxib

O

O

O CH3

OH

H

O

O

N

Cl

Cl

Na

-

+

O

O

OOH

OHNH

COOH

NH

COOH

O

O

OO

O O

N

NN

N

CH3O

CH3CH2CH2CH2

NH

S

S

N N

CH3

NH2SO2

CF3

areas that often exhibit lower pHs than uninvolved sites. Most often, there is a direct relationship between low pK and short half-life, but there are exceptions such as nabum-etone, which is nonacidic. Classifying NSAIDs on the basis of plasma half-life can be problematic given the fact that these drugs tend to accumulate in synovial fluid, where the concentration of drug may remain more stable than in the plasma. Short half-life NSAIDs potentially could be given less frequently than indicated by their plasma half-life. NSAIDs exhibiting longer half-lives require more time to reach steady-state plasma levels. Drugs with a half-life greater than 12 hours can be given once or twice a day. Plasma levels increase for a few days to several weeks

(depending on the specific half-life) but then tend to remain constant between doses. NSAIDs with longer half-lives also enable drug concentrations to equilibrate between the plasma and the synovial fluid, though total bound and unbound drug levels are usually lower in synovial fluid because there is less albumin in synovial fluid than in plasma. However, NSAIDs with longer half-lives or extended release formulation may be associated with increased propensity to cause adverse effects.41 COX-isozyme selectivity is likely to be a critically important factor in determining relative GI and cardiovascular risk, which should also be considered in addition to other pharmaco-logic properties for each NSAID.33

*Not approved by U.S. Food and Drug Administration.

880 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

ibuprofen, ketoprofen, and flurbiprofen, which exist as mix-tures of inactive (R) and active (S) enantiomers. Naproxen is composed of the active (S) enantiomer. Conversion of the propionic acid NSAIDs from the inactive (R) enantio-mer to the active (S) enantiomer occurs in vivo to various degrees, providing some basis for the variability in patient response. There is also genetic variability in the cytochrome P450 metabolic enzymes such that some individuals or ethnic groups metabolize drugs more slowly. For example, Asians are frequently slow metabolizers through the CYP2C9 pathway. Finally, the pharmacokinetics of some NSAIDs are affected by hepatic disease, renal disease, or old age.

Routes of Drug Delivery

NSAIDs are produced in a variety of dosage forms including intravenous, slow-release and sustained-release oral prepa-rations, topical preparations in various forms including gels and patches, and suppositories. Given the desire to reduce NSAID toxicity while preserving drug delivery to a specific site, efforts continue to alter drug formulation and delivery systems. Nanoparticles, liposomes, and microspheres are under investigation to allow dose reduction and specific targeting. Intra-articular delivery is under consideration, but because joints have efficient lymphatic clearance systems, the utility of this form of targeting remains to be proved.

Topical NSAID formulations were developed to reduce systemic exposure while preserving efficacy. Diclofenac, for example, is available as a solution, gel, or patch. The systemic effects are directly proportional to the surface area, and this method of delivery results in a relatively stable systemic diclofenac level compared with oral administration.44

Combination Drugs

NSAIDs have also been combined with agents having gas-troprotective effects into polypills that are currently avail-able on the market. This strategy may increase compliance with effective protective agents, thereby reducing adverse effects in clinical practice. Combining diclofenac with the synthetic PGE1 analogue misoprostol (Arthrotec) is shown to reduce risk of NSAID-related peptic ulcerations and mucosal injury, but utility of the combination is often limited by misoprostol-induced cramping and diarrhea.45 In population-based studies, Arthrotec was more effective than diclofenac and misoprostol co-prescription in preventing hospitalization for peptic ulcer disease or GI hemorrhage.46 The combination of enteric-coated naproxen and the proton pump inhibitor (PPI) esomeprazole (Vimovo) into a single pill has been approved by the U.S. Food and Drug Administration. This agent was shown to reduce endoscopi-cally detected gastric ulcers.47

A different strategy is nitric oxide releasing NSAIDs (NO-NSAIDs), which are synthesized by the ester linkage of an NO-releasing moiety to conventional NSAIDs includ-ing aspirin, flurbiprofen, diclofenac, sulindac, and others.48 The NO moiety is slowly released by enzymatic activity in vivo, likely by esterases, resulting in slow accumulation of the parent NSAID. The lower rate of GI ulceration

NSAID Metabolism

Almost all NSAIDs are more than 90% bound to plasma proteins. If total drug concentrations are increased beyond the point at which the binding sites on albumin are satu-rated, biologically active free drug concentrations increase disproportionately to the increasing total drug concentra-tion. The clearance of NSAIDs is usually by hepatic metab-olism with production of inactive metabolites that are excreted in the bile and urine. Most NSAIDs are metabo-lized through the microsomal cytochrome P450-containing mixed-function oxidase system. NSAIDs are most often metabolized by CYP3A, CYP2C9, or both. However, some are metabolized by other cytosolic hepatic enzymes.

Salicylate Metabolism and Aspirin Resistance

Salicylates are acetylated (e.g., aspirin) or nonacetylated (e.g., sodium salicylate, choline salicylate, choline magne-sium trisalicylate, salicylsalicylic acid).40 Although the nonacetylated salicylates are only weak inhibitors of COX in vitro, they are able to reduce inflammation in vivo. Aspirin is rapidly deacetylated to salicylate, both spontane-ously and enzymatically. Differences in formulation of these agents affect the absorption properties, but not bioavail-ability. Buffered aspirin tablets contain antacids that increase the pH of the microenvironment, whereas enteric coating slows absorption. The bioavailability of rectal aspirin sup-positories increases with retention time. Salicylates primar-ily bind to albumin and rapidly diffuse into most body fluids. Salicylate is metabolized principally by the liver and excreted primarily by the kidney. In the kidney, salicylate and its metabolites are freely filtered by the glomerulus, then reabsorbed and secreted by the tubules. Salicylate serum levels usually do not correlate well with dosage, however, and small increases in dosage may result in disproportionate increases in serum levels. The drug clearance rate is a func-tion of serum concentration. The primary factors regulating serum salicylate levels are urinary pH and metabolic enzyme activity.

The term aspirin resistance is broadly used to describe the failure of aspirin to prevent a thrombotic event whether due to pharmacologic resistance to the antiplatelet effects of aspirin or due to the inability of aspirin to overcome thrombophilia in a given clinical setting.42 Factors such as sex, genetic polymorphisms, and clinical factors including smoking, obesity, and diabetes may alter aspirin effects on platelet function. Lack of adherence and drug interactions also may play a role in aspirin resistance.

Pharmacologic Variability

Different patients can respond to the same NSAID in a variety of ways, and the basis for this individual variability remains unclear. Several pharmacologic factors related to NSAIDs may influence this variability such as dose response, plasma half-life, enantiomeric conversion, urinary excre-tion, and pharmacodynamic variation.43 Other important drug factors include protein binding, the metabolic profile of the drug, and the percentage of the drug that is available as the active (S) enantiomer. Some NSAIDs exist as two enantiomers; these include the propionic acid derivatives

881CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

production of PG, primarily PGE2, from vascular endothe-lial cells via COX-2 and mPGES-1.54 These PGs generate neuronal signals that activate the thermoregulatory center in the preoptic area of the anterior hypothalamus. PGE2 synthesis is stimulated by endogenous (e.g., interleukin-1) or exogenous (e.g., lipopolysaccharide) pyrogens. Mice with a targeted deletion of either the COX-2 or mPGES-1 genes fail to develop fever in response to inflammatory stimuli.55

Little evidence suggests that any NSAID has superior efficacy as an antipyretic. However, in fever associated with viral illnesses, aspirin should be avoided due to the associa-tion with hepatocellular failure (Reyes syndrome).56

Other Therapeutic Effects

Antiplatelet Effects

Aspirin and traditional NSAIDs inhibit platelet COX-1 to variable degrees. Except for aspirin, inhibition of platelet aggregation is reversible and depends on the concentration of drug in the platelet. Aspirin acetylates platelet COX-1, which cannot be resynthesized. The antiaggregation effect of as little as 80 mg of aspirin can last for up to 4 to 6 days, until the bone marrow can synthesize new platelets.57

On the basis of accumulated data showing its benefits, the FDA has approved ASA for use in the secondary pre-vention of cardiovascular disease. Major trials have shown that meaningful decreases in nonfatal myocardial infarction (MI), nonfatal stroke, and death can be realized by daily administration of ASA of 75 to 325 mg. Major vascular events can be reduced by 10 to 20 events for every 1000 patients treated, at a cost of one to two major GI bleeds.58

There was no reduction in rates of MI observed with the use of ASA, 100 mg every other day, in the Nurses Health Study of primary prevention of major vascular events, whereas rates of GI bleeding were increased. However, stroke rates were significantly reduced on this regimen.59 The U.S. Preventative Health Task Force has updated its recommendations to encourage use of low-dose ASA in men aged 45 to 79 and women aged 55 to 79.58 ASA used for primary prevention of cardiovascular events appears to lower the risk for MI in men and for stroke in women.60

Cancer Chemoprevention

A large body of epidemiologic and animal studies provides evidence that a high-fat diet can be associated with a risk for cancer. AA, one of the major ingredients of animal fats, and the eicosanoids derived from AA are shown to be important contributors to cancer development.10 Large-scale epidemiologic studies have long indicated that long-term NSAID use reduces the incidence of a variety of cancers including colon, intestinal, gastric, breast, and bladder, 40% to 50%.10 Given the ability of the NSAIDs to inhibit COX and PG production, the COX pathway imme-diately becomes implicated as playing an important role in the pathogenic process. It is well recognized that growth factors, tumor promoters, and oncogenes stimulate PG pro-duction via the induction of COX-2 and that human tumor-igenic tissues exhibit increased COX activity compared with their normal, nontumorigenic counterparts. COX-2 is overexpressed in 80% of colorectal cancer tissues. Among

associated with these drugs is likely related to NO-associated vasodilation and the relatively lower concentration of the parent NSAID.

THERAPEUTIC EFFECTS

Anti-inflammatory Effects

NSAIDs are frequently used as first-line agents for the symp-tomatic relief of many different inflammatory conditions. In double-blind, randomized clinical trials of inflammatory arthritis, NSAIDs have been compared with placebo, aspirin, and each other. Clinical trials of NSAID efficacy in rheumatoid arthritis (and osteoarthritis) most often employ a design whereby the current NSAID is discontinued and the patient must have an increase in symptoms or flare to enter the study. Although there is some variation in primary outcome measures, most include parameters that make up the American College of Rheumatology (ACR)-20. Effi-cacy superior to that of placebo is easily demonstrated for NSAIDs within 1 to 2 weeks in patients with active RA who are not receiving corticosteroids or other anti-inflammatory medications.49 Comparisons of adequate doses of traditional NSAIDs or COX-2-selective NSAIDs with one another almost always show comparable efficacy. Despite improvement in pain and stiffness with NSAIDs, these agents do not usually reduce acute phase reactants, nor do they modify radiographic progression. The anti-inflammatory effects of NSAIDs have also been demon-strated in rheumatic fever, juvenile rheumatoid arthritis, ankylosing spondylitis, gout, osteoarthritis, and systemic lupus erythematosus (SLE). Although not as rigorously proven, their efficacy is also accepted in treatment of reac-tive arthritis, psoriatic arthritis, acute and chronic bursitis, and tendinitis.

Analgesic Effects

Virtually all NSAIDs relieve pain when used in doses sub-stantially lower than those required to suppress inflamma-tion. The analgesic action of NSAIDs is due to inhibition of PG production in peripheral tissues and in the central nervous system (CNS). In the periphery, PGs do not induce pain per se but rather sensitize peripheral nociceptors to the effects of mediators such as bradykinin or histamine.50 PGs released during inflammation or other trauma lower the activation threshold of tetrodotoxin-resistant sodium chan-nels on sensory neurons. In the CNS, where NSAIDs and acetaminophen exert analgesic effects, PGs also play an important role in neuronal sensitization. COX-2 is consti-tutively expressed in the dorsal horn of the spinal cord, and its expression is increased during inflammation.51 Centrally generated PGE2 activates spinal neurons and also microglia that contribute to neuropathic pain.52 Both COX-1 and COX-2 play a role in nociception as demonstrated by reduc-tions in experimental pain in mice deficient in either COX-1 or COX-2.53

Antipyretic Effects

The NSAIDs and acetaminophen effectively suppress fever in humans and experimental animals. Fever results from the

882 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

mucosal lesions. Esterification of acidic NSAIDs such as the case for NO-NSAIDs suppresses mucosal injury.24

Acidic NSAIDs accumulate in gastric epithelial cells and lead to generation of reactive oxygen species. NSAIDs may also cause apoptosis, which appears independent of COX inhibition. They also associate with extracellular phospho-lipids resulting in attenuation of the hydrophobic surface barrier of the stomach.69 Phospholipids associated with cell membranes are disrupted, leading to increased permeability and acid or bile back diffusion.

The integrity of mucosal defense depends on generation of PGE2 and PGI2 from COX enzymes. COX-1 is abun-dantly expressed under basal conditions in gastric mucosa, whereas COX-2 is almost undetectable. However, both COX-1 and COX-2 are rapidly upregulated following injury or when there are pre-existing ulcers.70 This may explain the observation that concurrent Helicobacter pylori infection increases the risk of developing peptic ulcers and increases bleeding in NSAID users.71

It appears that concurrent COX-1 and COX-2 inhibition is associated with the highest propensity to develop gastric ulcers.41 This clinical observation is consistent with animal studies whereby mice deficient for a single COX enzyme or treated with drugs that specifically inhibit either COX-1 or COX-2 do not develop ulceration. Severe gastric lesions are seen when both enzymes are simultaneously blocked. However, when gastric mucosa is damaged, inhibition of either COX-1 or COX-2 is associated with development of ulcers.68 Traditional and COX-2-selective NSAIDs delay ulcer healing, with nonselective NSAIDs doing so to a

the PGs, PGE2 is most abundant in human neoplasms. The inducible mPGES-1 enzyme is highly expressed in tumors, and its absence suppresses intestinal tumorigenesis in animal models. Furthermore, the enzyme that metabolizes intracel-lular PGE2, 15-hydroxyprostaglandin dehydrogenase, is ubiquitously lacking in tumors, and mice with a genetic deletion of this enzyme have accelerated tumorigenesis.10 Many natural products including resveratrol (red wine), catechins (green tea), and curcumin (saffron) also inhibit COX, which may be an important mechanism underlying their putative cancer-preventing effects.61

A retrospective cohort study shows that aspirin and tra-ditional NSAIDs specifically reduce cancer risk in the sub-group of patients whose colon tumors express higher levels of COX-2.62 In a meta-analysis of ASA (75 mg daily and upward without dose dependence) effects on cancer, alloca-tion to aspirin reduced death due to cancer by more than 20%.63 On analysis of individual patient data, cancer death benefit was apparent only after 5 years follow-up, and benefit increased with scheduled duration of trial treatment. ASA effects appear greater on adenomatous cancer than other cancer types. Other studies demonstrated a reduction in both incidence of colorectal cancer and death from colorectal cancer, particularly for cancers of the proximal colon.64 Long-term low-dose ASA use also appears to reduce the risk of prostate cancer.65

Clinical trials also demonstrated that traditional and COX-2-selective NSAIDs could cause regression of polyps in patients with familial adenomatous polyposis (FAP).66 Celecoxib was subsequently approved by the FDA for reduc-tion of polyps in patients with FAP. Although NSAIDs are still among the most promising chemopreventative agents for cancer, cardiovascular and GI side effects have damp-ened enthusiasm.10 Alternative strategies for blocking effects of PGE2 and other eicosanoids need evaluation to determine their effectiveness as cancer prevention and adjunct therapy.

ADVERSE EFFECTS

The NSAIDs share a common spectrum of clinical toxici-ties, although the frequency of particular side effects varies with the compound (Table 59-2). Hazard of individual NSAIDs is related to the pharmacologic characteristics such as bioavailability and half-life, as well as potency for inhibition of COX-1 and COX-2.33,41,67

Gastrointestinal Tract Effects

NSAIDs cause GI injury through both topical and systemic effects. Following mucosal injury, NSAIDs inhibit the early events necessary to repair superficial injury, as well as later events of cell proliferation and angiogenesis leading to delayed ulcer healing.68 Topical injury initiates the initial mucosal erosions by disrupting the gastric epithelial cell barrier, but PG depletion is essential for the development of clinically significant gastric and duodenal ulcers. The propensity of specific NSAIDs to cause acute topical injury depends on the pKa, with most NSAIDs being weak organic acids. Aspirin is particularly prone to cause mucosal injury when administered orally.24 Nonacidic NSAIDs such as nabumetone, etodolac, and celecoxib do not cause acute

Table 59-2 Shared Toxicities of Nonsteroidal Anti-inflammatory Drugs

Organ System Toxicity

Gastrointestinal DyspepsiaEsophagitisGastroduodenal ulcersUlcer complications (bleeding, perforation

obstruction)Small bowel erosions and stricturesColitis

Renal Sodium retentionWeight gain and edemaHypertensionType IV renal tubular acidosis and

hyperkalemiaAcute renal failurePapillary necrosisAcute interstitial nephritisAccelerated chronic kidney disease

Cardiovascular Heart failureMyocardial infarctionStrokeCardiovascular death

Hepatic Elevated transaminasesReyes syndrome (aspirin only)

Asthma/allergic Aspirin-exacerbated respiratory disease* (susceptible patients)

RashHematologic CytopeniasNervous Dizziness, confusion, drowsiness

SeizuresAseptic meningitis

Bone Delayed healing

*Reduced risk in COX-2-selective nonsteroidal anti-inflammatory drugs.

883CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

and diclofenac ranged from 3.98 to 5.63. Drugs with a long half-life or slow-release formulation were associated with higher risk even accounting for dose.41 Profound and coin-cident inhibition of both COX-1 and COX-2 using whole blood assay, as seen for ketorolac, piroxicam, naproxen, ketoprofen, and indomethacin, was associated with a rela-tive risk of greater than 5 for GI bleeding and perforation.41 The use of low-dose aspirin, even in the absence of other risk factors, increases risk for bleeding and death. Many patients taking low-dose aspirin may do so without the knowledge of their physician; it is essential to query patients specifically on this point.

Other patient-specific factors influence the overall risk for GI ulcers and ulcer complications (Figure 59-5).45,77 A previous history of ulcer or ulcer complications is an impor-tant risk, especially if combined with other risks. Infection with H. pylori is likely to be associated with additive effect.71 It remains unclear if eradication of H. pylori would be useful in primary prevention of NSAID-induced ulcers, but this may be advantageous in those patients requiring long-term NSAIDs.45 Eradication alone is insufficient as a single strat-egy for secondary prevention of ulcer complications. This strategy appears to be most effective in reducing the bleed-ing risk of patients on low-dose aspirin but is less useful than use of PPIs in patients taking NSAIDs.45

Table 59-3 provides recommendations for patients who need NSAIDs and have GI risks.45 Misoprostol has consis-tently been shown to be effective in reduction of gastroduo-denal ulcers. Meta-analysis showed a reduction of 74% in gastric ulcers and 53% in duodenal ulcers when compared with placebo.78 The effectiveness of misoprostol is compa-rable with the PPI lansoprazole.79 However, the high preva-lence of abdominal cramping and diarrhea limit misoprostol use at full doses. For those who do not tolerate full doses (200 mcg four times daily), lower doses of 400 to 600 mcg/day may be useful and comparable with PPIs.

PPIs have been used extensively for prevention of NSAID-induced ulcers and are also used for ulcer healing. Their excellent tolerability and availability over-the-coun-ter have led to their dominance as pharmacologic agents for preventing NSAID-induced gastroduodenal ulcers. Studies have shown a reduction in endoscopic ulcer rate from 17% in patients taking traditional or COX-2-selective NSAIDs plus placebo to 5.2% and 4.6% in patients taking NSAIDs plus esomeprazole 20 mg or 40 mg, respectively.80 As noted previously, a combination pill containing naproxen and

greater degree.72 It should also be noted that GI bleeding may be related to the combination of injury and inhibition of platelet aggregation. This is an additional factor in the propensity of aspirin and other nonselective NSAIDs to cause clinically apparent ulcers.24,68

Dyspepsia

Nonulcer dyspepsia is the most common adverse event (10% to 20%) associated with use of NSAIDs and may account for poor tolerability.73 Dyspepsia is more often reported in younger as compared with older patients.74 Though expected to reduce dyspepsia, COX-2-selective NSAIDs are also associated with a substantial level of adverse GI symptoms.73 PPIs have been shown to reduce dyspepsia in controlled trials.74 Studies have shown that histamine-2-receptor antagonists (H2RAs) are also effective for reducing dyspepsia.75 Crucially from a clinical perspec-tive, subjective symptoms of dyspepsia, fecal blood loss, and endoscopic findings correlate poorly. Furthermore, only a minority of patients with serious GI events report anteced-ent dyspepsia.76

Gastritis and Gastroduodenal Ulcer

Up to 25% of chronic NSAID users will develop ulcer disease, and 2% to 4% will bleed or perforate. These GI events result in more than 100,000 hospital admissions annually in the United States and between 7000 and 10,000 deaths, especially in those patients with highest risk.45 The risk for ulcer complications appears highest within the first 3 months of use but remains present with longer-term therapy. A recent meta-analysis of observational studies on NSAIDs and upper GI bleeding or perforation of studies published between 2000 and 2008 demonstrated a relative risk of 4.50 (95% confidence interval [95% CI], 3.82 to 5.31) for traditional NSAIDs and 1.88 (195% CI, 0.96 to 3.71) for selective COX-2 inhibitors.41 For traditional NSAIDs, the risk of low or medium doses was associated with a lower risk than higher dose. Several NSAIDs had far higher than average risk including ketorolac 14.54 (95% CI, 5.87 to 36.04) and piroxicam 9.94 (95% CI, 5.99 to 16.50). The risks for celecoxib 1.42 (95% CI, 0.85 to 2.37) and ibuprofen 2.69 (95% CI, 2.17 to 3.33) were lower than NSAIDs overall. Relative risk for other commonly used NSAIDs including naproxen, indomethacin, meloxicam,

Figure 59-5 Established risk factors for upper gastrointestinal bleeding associated with nonste-roidal anti-inflammatory drug (NSAID) use. OR, odds ratio; RR, relative risk. (Adapted from Gutthann SP, Garca-Rodrguez LA, Raiford DS: Individual non-steroidal anti-inflammatory drugs and other risk factors for upper gastrointestinal bleeding and perfo-ration, Epidemiology 8:1824, 1997; Huang JQ, Sridhar S, Hunt RH: Role of Helicobacter pylori infec-tion and non-steroidal anti-inflammatory drugs in peptic ulcer disease: a meta-analysis, Lancet 359:1422, 2002; and Lanas A, Garca-Rodrguez LA, Arroyo MT, et al: Risk of upper gastrointestinal ulcer bleeding associated with selective cyclooxygenase-2 inhibitors, traditional non-aspirin non-steroidal anti-inflammatory drugs, aspirin and combinations, Gut 55:17311738, 2006.)

Previous complicated ulcer

Multiple NSAIDs (including aspirin)High-dose NSAID

Anticoagulant treatmentPrevious uncomplicated ulcer

Age 70-80 yearsHelicobacter pylori infection

Oral corticosteroids

13.5

8.9

7

6.4

6.1

5.6

3.5

2.2

0 2 4 6 8Adjusted RR (OR)

10 12 14 16

884 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

Colitis

NSAIDs cause erosions, ulcers, hemorrhage, perforations, strictures, and complications of diverticulosis in the large bowel.84 NSAID-induced injury is more common in the right colon (80%) but can occur in the transverse and left colon. NSAID-containing suppositories can cause erosions, ulcers, and stenoses in the rectum. NSAID colonopathy is in the differential diagnosis of inflammatory bowel disease. Patients with NSAID-induced colonopathy are typically older, and the erosions are more likely to be transverse or circular.85 There is also a concern that treatment with tra-ditional and COX-2-selective NSAIDs may exacerbate inflammatory bowel disease.86 NSAIDs are also implicated in the development of collagenous colitis.87

Renal Effects

PGs play a vital role in solute and renovascular homeosta-sis.88 It is becoming quite clear that PGs are produced by both COX-1 and COX-2, generally in different locations within the kidney, and that these PGs may play different physiologic roles in renal function.89,90 COX-1 is highly expressed in the renal vasculature, glomerular mesangial cells, and collecting duct. COX-2 expression is restricted to the vasculature, cortical thick ascending limb (specifi-cally in cells associated with the macula densa), and in medullary interstitial cells. COX-2 expression in the macula densa increases in high-renin states (e.g., salt restriction, angiotensin-converting enzyme inhibition, renovascular hypertension), and selective COX-2 inhibitors significantly decrease plasma renin levels and renal renin activity. COX-2 expression in the macula densa is reduced by angiotensin II and mineralocorticoids. Dehydration or hypertonicity appears to regulate COX-2 expression in the medullary interstitium. COX-2 is also necessary for normal renal development.

Sodium Excretion

PGs are known to regulate renal sodium resorption by their ability to inhibit active transport of sodium in both the thick ascending limb and the collecting duct and to increase renal water excretion by blunting the actions of vasopres-sin.91 The cellular source of COX-2-derived prostanoids that promote natriuresis remains uncertain, but it is possible that they may in large part be derived from the medullary interstitial cells. Sodium retention has been reported to occur in up to 25% of NSAID-treated patients and may be particularly apparent in patients who have an existing avidity for sodium, such as those with mild heart failure or liver disease.91 Decreased sodium excretion in NSAID-treated patients can lead to weight gain and peripheral edema. This effect may be sufficiently important to cause clinically important exacerbations of congestive heart failure.

Hypertension

NSAIDs may cause altered blood pressure, with average increases of mean arterial pressure of between 5 and 10 mm Hg. In addition, using NSAIDs may increase the risk of

esomeprazole has been approved for use. It may reduce non-compliance but will be associated with higher cost.

High-dose, twice-daily doses of H2RA reduce the risk of NSAID-induced endoscopic ulcers and are the least costly alternative. However, these agents are inferior to PPIs and, like PPIs, there are no randomized clinical outcome trials that evaluate the efficacy of H2RAs in chronic NSAID users.24

Esophageal Injury

Aspirin and NSAIDs are associated with esophagitis related to mechanisms similar to those in the gastric mucosa.81,82 Esophageal emptying may be slowed in the elderly, resulting in a prolonged exposure of the mucosa to the irritant action of aspirin and NSAIDs. Gastroesophageal reflux may be an aggravating factor and lead to stricture formation. Bleeding may also complicate esophagitis. NSAIDs should be pre-scribed with caution in the presence of gastroesophageal reflux disease.

Small Bowel Injury

The availability of video capsule endoscopy (VCE) and balloon enteroscopy has advanced the ability to detect small intestinal lesions in patients taking NSAIDs. NSAIDs can cause a concentric diaphragm-like stricture in the small bowel in addition to causing mucosal injury and bleeding. Two recent studies of patients on NSAIDs for at least 3 months using VCE demonstrated a prevalence of 70% to 80% for small bowel injuries.83 Furthermore, NSAID-induced small bowel injury is likely a common cause of obscure GI bleeding. NSAIDs that undergo enterohepatic circulation are likely to be associated with higher risk. Small bowel injury may be detected by anemia or symptoms of obstruction related to stricture.83 Strategies effective for gastroduodenal ulcers such as misoprostol or certain PPIs may also reduce the risk for small bowel mucosal injury. Strictures may require balloon endoscopy or surgical intervention.83

Table 59-3 Strategies for Gastrointestinal Risk Reduction45

Gastrointestinal Risk Potential Strategies

Low risk Intermittent NSAID useLow-dose NSAID

Moderate risk (1-2 risk factors)Age > 65High-dose NSAIDPrevious history of

uncomplicated ulcerConcurrent use of aspirin,

corticosteroids, or anticoagulants

Intermittent NSAID useNSAID + PPINSAID + misoprostolNSAID + high-dose H2RA*

High risk>2 Risk factorsHistory of previous complicated

ulcer, especially recent

Alternative treatmentCOX-2-selective NSAID + PPICOX-2-selective NSAID +

misoprostolHelicobacter pylori positive Consider eradication in

moderate- to high-risk patients

COX-2, cyclooxygenase-2; H2RA, histamine-2-receptor antagonist; NSAID, nonsteroidal anti-inflammatory drug; PPI, proton pump inhibitor.

*Less effective than PPI or misoprostol.

885CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

Cardiovascular Effects

The risk of adverse cardiovascular effects associated with NSAID use was not widely appreciated until COX-2- selective NSAIDs were introduced into clinical practice. Rofecoxib, a potent, highly specific COX-2 inhibitor with a long half-life, was shown to have a substantially increased risk of MI and stroke and removed from the market because of this adverse effect.7,67 The mechanisms for cardiovascular risks associated with all NSAIDs are likely related to an imbalance between complete inhibition of COX-1 and COX-2 across the dosing interval. The COX-1 isoform is responsible for generation of platelet TXA2, which facili-tates platelet aggregation and thrombus formation. In order to inhibit this activity, COX-1 must be inhibited by 95% or greater.97 Antithrombotic PGI2 synthesized by endothelial COX-2 is inhibited almost completely by both traditional and COX-2-selective NSAIDs. It is proposed that the rela-tionship between excess cardiovascular risk for all NSAIDs, not only COX-2-selective NSAIDs, is related to the degree of COX-2 inhibition absent complete inhibition of COX-1.98 Investigators have shown that drugs that inhibit COX-2 less than 90% at therapeutic concentrations in the whole blood assay present a relative risk for MI of 1.18 (95% CI, 1.02 to 1.38), whereas drugs that inhibit COX-2 to a greater degree present a relative risk of 1.60 (95% CI, 1.41 to 1.81).98

Relative inhibition of the COX isoforms is not the only mechanism that contributes to cardiovascular hazard. Other actions of NSAIDs including effects on blood pressure, endothelial function and nitric oxide production, and other renal effects may play a role in cardiovascular risks.67,99,100 Multiple analyses have demonstrated that the risk for car-diovascular hazard is significantly higher in those with pre-existing coronary artery disease. Some NSAIDs, notably ibuprofen and naproxen, may interfere with the irreversible inhibition of platelet COX-1 by aspirin, thereby increasing cardiovascular hazard in aspirin users.98

A number of large-scale randomized controlled trials comparing NSAIDs with placebo or with each other have been performed and analyzed to determine the risk of MI, stroke, cardiovascular death, death from any cause, and Antiplatelet Trialists Collaboration (APTC) composite outcomes.67 Because event rates in most of these studies were low, uncertainty regarding absolute and relative risk remains. For example, there were only 554 MIs in aggregate across all trials included in the most comprehensive analysis to date. Nevertheless, it appears from analyses of these aggregated clinical trials that all traditional and COX-2-selective NSAIDs except naproxen carry an excess risk of more than 30% compared with placebo.67 Pairwise com-parisons of the most commonly used traditional and COX-2-selective NSAIDs studied in clinical trials also suggest that naproxen may have lower cardiovascular risk.67 One meta-analysis explored the effects of dose and dosing regimen in a pooled analysis of six randomized placebo-controlled trials of celecoxib.101 Lower doses and once-daily regimens were associated with lower relative risks for the APTC outcomes. This finding confirms data from other studies that suggest avoiding continuous interference with PG biosynthesis is associated with lower cardiovas-cular risk.98

initiating antihypertensive therapy in older patients, with the magnitude of increased risk being proportional to the NSAID dose.92 Furthermore, in a large (n = 51,630) pro-spective cohort of women aged 44 to 69 without hyperten-sion in 1990, incident hypertension over the following 8 years was significantly more likely in frequent users of aspirin, acetaminophen, and NSAIDs.93 NSAIDs can atten-uate the effects of antihypertensive agents including diuret-ics, angiotensin-converting enzyme inhibitors, and -blockers, interfering with blood pressure control.

PGs stimulate renin release which, in turn, increases secretion of aldosterone and, subsequently, potassium secre-tion by the distal nephron. For this reason, NSAID-treated patients may develop hyporeninemic hypoaldosteronism that manifests as type IV renal tubular acidosis and hyper-kalemia.91 The degree of hyperkalemia is generally mild; however, patients with renal insufficiency or those that may otherwise be prone to hyperkalemia (e.g., patients with diabetes mellitus and those on angiotensin-converting enzyme inhibitors or potassium-sparing diuretics) may be at greater risk.

Acute Renal Failure and Papillary Necrosis

Acute renal failure is an uncommon consequence of NSAID treatment. This is due to the vasoconstrictive effects of NSAIDs and is reversible. In most cases, renal failure occurs in patients who have a depleted actual or effective intravas-cular volume (e.g., congestive heart failure, cirrhosis, renal insufficiency).91

Marked reduction in medullary blood flow may result in papillary necrosis that may arise from apoptosis of medullary interstitial cells. Inhibition of COX-2 may be a predisposing factor.90,94

Interstitial Nephritis

Another adverse renal effect resulting from NSAIDs involves an idiosyncratic reaction accompanied by massive proteinuria and acute interstitial nephritis. Hyper-sensitivity phenomena such as fever, rash, and eosinophilia may occur. This syndrome has been observed with most NSAIDs.

Chronic Kidney Disease

Use of analgesics, particularly acetaminophen and aspirin, has been associated with nephropathy leading to chronic renal failure. In one large case-control study, the regular use of aspirin or acetaminophen was associated with a risk of chronic renal failure 2.5 times as high as that for nonuse, and the risk increased significantly with an increas-ing cumulative lifetime dose.95 In subjects regularly using both acetaminophen and aspirin, the risk was also signifi-cantly increased compared with users of either agent alone. No association between the use of nonaspirin NSAIDs and chronic renal failure could be detected after adjusting for acetaminophen and aspirin use. Pre-existing renal or systemic disease was a necessary precursor to analgesic-associated renal failure and those without pre-existing renal disease had only a small risk of end-stage renal disease.95,96

886 PART 8 | PHARMACOLOGY OF ANTIRHEUMATIC DRUGS

normalization of the transaminase values, although rare, fatal outcomes have been reported with almost all NSAIDs. Those NSAIDs that appear most likely to be associated with hepatic adverse events are diclofenac and sulindac.

In clinical trial reports to the FDA, 5.4% of patients with rheumatoid arthritis who were treated with aspirin experi-enced persistent elevations of results in more than one liver function test. In children with viral illnesses, hepatocellular failure and fatty degeneration (Reyes syndrome) are associ-ated with aspirin ingestion.56

Asthma and Allergic Reactions

Asthma

Up to 10% to 20% of the general asthmatic population, especially those with the triad of vasomotor rhinitis, nasal polyposis, and asthma, are hypersensitive to aspirin. In these patients, ingestion of aspirin and nonspecific NSAIDs leads to severe exacerbations of asthma with naso-ocular reac-tions. Formerly termed aspirin-sensitive asthma, these patients are now characterized as having aspirin-exacerbated respiratory disease (AERD) because they have chronic upper and lower respiratory mucosal inflammation, sinusitis, nasal polyposis, and asthma independent of their hypersen-sitivity reactions. It is now thought that production of pro-tective PGs in the setting of AERD is derived from a COX-1. A number of studies have been reported demon-strating that the COX-2-specific NSAIDs, rofecoxib and celecoxib, fail to trigger asthma exacerbation or naso-ocular symptoms in patients with AERD.105,106 Nevertheless, these studies were performed as challenge tests rather than long-term placebo-controlled trials, and caution is advised. The fact that specific COX-2 inhibitors appear safe in AERD does not imply that other hypersensitivity reactions do not occur.

Allergic Reactions

A wide variety of cutaneous reactions have been associated with NSAIDs. Almost all the NSAIDs have been associated with cutaneous vasculitis, erythema multiforme, Stevens-Johnson syndrome, or toxic epidermal necrolysis. NSAIDs are also associated with urticaria/angioedema and anaphy-lactoid or anaphylactic reactions. Special note should be made that celecoxib and valdecoxib contain a sulfonamide group and should not be given to patients who report allergy to sulfa-containing drugs.

Hematologic Effects

Aplastic anemia, agranulocytosis, and thrombocytopenia are rarely associated with NSAIDs, but they are prominent among the causes of deaths attributed to these drugs. Because of the risk of hematologic effects, phenylbutazone is no longer recommended for use in any condition in the United States and has been taken off the market.107

Effects on the Immune System

Virtually all cell types composing the immune system produce and respond to PGs. PGE2 is a potent inhibitor of

Because clinical trials have been underpowered to spe-cifically address the relative cardiovascular risk of NSAIDs, investigators have turned to observational datasets. Using a large observational database with 8852 cases of nonfatal MI, a recent case-control study also identified a 35% increase in the risk of MI while using NSAIDs.98 This type of study also identifies naproxen as potentially having a lower risk. In this analysis, long half-life was an independent predictor of MI hazard. The effect of dose and slow-release formulation demonstrated that risk was a direct consequence of pro-longed drug exposure. It appears that the risk associated with these pharmacologic factors may be even more impor-tant than COX-2 specificity for most NSAIDs.67,98

A number of strategies have been suggested to mitigate cardiovascular risks associated with NSAID use (Table 59-4).102 These recommendations take into account a patients underlying risk, aspirin use, and the interaction between NSAIDs. In addition, the specific choice of NSAID should consider its pharmacologic properties.67,98

Heart Failure

NSAIDs are associated with reduced sodium excretion, volume expansion, increased preload, and hypertension. As a result of these properties, patients with pre-existing heart failure are at risk of decompensation with a relative risk of 3.8 (95% CI, 1.1 to 12.7). After adjusting for age, sex, and concomitant medication, the relative risk was 9.9 (95% CI, 1.7 to 57.0).103 Studies disagree as to whether NSAIDs are a risk for new heart failure, although the elderly may be at particular risk.103,104

Closure of the Ductus Arteriosus

The maintenance of an open ductus arteriosus and its closure during the postnatal period are regulated by PG. COX-1, COX-2, and EP4-deficient mice die from neonatal circulatory failure because the ductus arteriosus remains open. It is inadvisable for pregnant women to take NSAIDs during the last trimester of pregnancy because of the risk of a persistently patent ductus arteriosus.

Hepatic Effects

Small elevations of one or more liver tests may occur in up to 15% of patients taking NSAIDs, and notable elevations of alanine aminotransferase or aspartate aminotransferase (three or more times the upper limit of normal) have been reported in approximately 1% of patients in clinical trials of NSAIDs. Patients usually have no symptoms, and discon-tinuation or dose reduction generally results in

Table 59-4 Strategies for Reducing Cardiovascular Risk67,98,102

If using aspirin, take aspirin dose 2 hr before NSAID dose*Do not use NSAIDs within 3-6 mo of an acute cardiovascular event

or procedureCarefully monitor and control blood pressureUse low-dose, short half-life NSAIDs and avoid extended-release

formulations

NSAID, nonsteroidal anti-inflammatory drug.

*Especially ibuprofen. Celecoxib does not appear to interfere with aspirin actions.

887CHAPTER 59 | PROSTANOID BIOLOGY AND ITS THERAPEUTIC TARGETING

and highly expressed and regulated in osteoblasts. Parathy-roid hormone (PTH) is a strong inducer of COX-2. The production of PG by osteoblasts is an important mechanism for the regulation of bone turnover.27 The major effect of PGE2 is considered to occur indirectly via upregulation of receptor activator of NFB ligand (RANKL) expression and by inhibition of osteoprotegerin (OPG) expression in osteoblastic cells, which facilitated osteoclastogenesis. Genetic deletion of PTGS2 or COX-2-selective NSAIDs partially block the PTH- or 1,25-OH vitamin Dinduced formation of osteoclasts in organ cultures. Recently, a famil-ial disorder, primary idiopathic hypertrophic osteoarthropa-thy, was found to be associated with a mutation in the enzyme 15-hydroxyprostaglandin dehydrogenase, the enzyme that inactivates PGE2.118 These patients have chronically elevated PGE2 levels and digital clubbing with evidence of increased bone formation and resorption in the phalanges.

The role of endogenous PG and NSAIDs in skeletal pathology remains complex. LPS-induced bone loss can be ameliorated in mice lacking mPGES-1.119 Inflammatory bone loss likely reflects increased bone resorption and decreased bone formation.27 It has long been appreciated that NSAIDs can inhibit experimental fracture healing and reduce formation of heterotopic bone in patients.120 Fur-thermore, fracture healing is impaired in rats treated with specific COX-2 inhibitors and in mice with genetic deletion of the COX-2 gene.121 Given the effectiveness of NSAIDs as analgesics, it is important to understand the clinical concern regarding impaired fracture healing and NSAIDs. A recent meta-analysis found a pooled odds ratio for non-union with NSAID exposure of 3 (95% CI, 1.6 to 5.6).122 However, there was a significant association between lower-quality studies, and higher reported odds ratios for non-union was observed. When only higher-quality studies were considered, no statistically significant association between NSAID exposure and nonunion was identified.

The impact of NSAIDs on bone mineral density (BMD) also remains unclear.27 In older men, daily use of COX-2-selective NSAIDs was associated with lower hip and spine BMD compared with nonusers, but in postmenopausal women not taking hormone replace