Arthritis & Allied Conditions
15th Edition

Chapter 31
Nonsteroidal Anti-Inflammatory Drugs
John S. Sundy
The nonsteroidal antiinflammatory drugs (NSAIDs) are a diverse group of compounds that includes salicylates, nonselective cyclooxygenase (COX) inhibitors, and the newer class of selective COX-2 inhibitors. NSAIDs are among the most widely used class of medications in the United States. They are effective across a wide spectrum of disorders that includes fever, inflammation, mild to moderate pain, and the prevention of cardiovascular syndromes. NSAIDs have a critical role in the management of numerous rheumatologic disorders. Their use is likely to grow as a result of emerging applications such as the prevention of some forms of cancer and Alzheimer disease.
The medicinal use of NSAIDs dates back several millennia to the use of willow bark to treat musculoskeletal pain (1). The active ingredient of willow bark, salicin, was extracted in 1828, and later obtained in crystalline form (1,2). Subsequently, sodium salicylate and aspirin were synthesized in 1875 and 1899, respectively (1,3). The first nonaspirin NSAID, phenylbutazone, was approved by the U.S. Food and Drug Administration (FDA) in 1949, followed by indomethacin in 1963 (4). Numerous other NSAIDs followed for the next 30 years. The most recent significant advance has been the introduction of the selective COX-2 inhibitors—celecoxib, rofecoxib, and valdecoxib—beginning in 1999 (5).
The NSAIDs are a mainstay in the pharmacologic armamentarium of rheumatologists. For treatment of mild to moderate musculoskeletal pain and osteoarthritis (OA), the NSAIDs are generally considered the pharmacologic treatment of choice (6,7,8). In contrast, NSAIDs have assumed an adjunctive position in the management of immune-mediated inflammatory diseases such as rheumatoid arthritis (RA), owing to the early use of immunomodulating therapy for these conditions.
Despite the prevalent use of NSAIDs, drug-related adverse events are a primary concern for clinicians. The high prevalence of NSAID-induced peptic ulcer disease (PUD) provided the impetus for the recent development of the new class of highly selective COX-2 inhibitor. Ongoing concerns relate to exacerbation of hypertension and congestive heart failure (CHF) and impairment of renal function in some patients using NSAIDs. The therapeutic benefits and widespread use of NSAIDs requires that physicians of all specialties, especially rheumatologists, be well versed in
P.680

P.681

the appropriate prescribing and monitoring of patients taking NSAIDs. The efficacy of NSAIDs in specific rheumatologic disorders is discussed in detail throughout this text. This chapter will review the biochemical and pharmacologic rationale for the use of NSAIDs, and will review the strategies for anticipating and preventing adverse events of therapy.
CHEMISTRY
The NSAIDs may be divided into three major classes: aspirin and salicylates, nonselective NSAIDs, and selective COX-2 inhibitors (Fig. 31.1,Table 31.1). The prototype salicylate, aspirin, is the salicylic ester of acetic acid. Chemically, there are two classes of salicylates: esters of salicylic acid (e.g., salsalate), and salicylate esters of organic acids (e.g., acetylsalicylic acid or aspirin). Other compounds in clinical use include the salicylic acid salts (e.g., choline salicylate). A more clinically relevant classification scheme groups salicylates into acetylated (aspirin) and nonacetylated salicylates (all other salicylates in clinical use).
FIG. 31.1. Chemical structures of selected salicylates, nonselective nonsteroidal antiinflammatory drugs, and selective cyclooxygenase-2 inhibitors.
TABLE 31.1. Characteristics of currently approved nonsteroidal antiinflammatory drugs (NSAIDs)

Drug (trade names) Indications and uses Formulations Daily dose Metabolism Half-life Other considerations

Salicylates
Aspirin Arthralgia, dental pain, dysmenorrhea, fever, headache, JRA, migraine, mild pain, myalgia, OA, RA, prevention and treatment cardiovascular thrombosis Numerous Variable, depending on indication; maximum (adults) 2.4–5.4 g/day in 4 or more divided doses Hepatic and renal Acetylsalicylic acid 15–30 min; salicylate 2–30 h Dose in children body weight <25 kg is 60–90 mg/kg/day. Serum salicylate levels may need to be monitored at higher doses.
Choline magnesium trisalicylate (Trilisate) Fever, JRA, mild-moderate pain, OA, RA Solution Tablets 3 g/day in divided doses Hepatic and renal Low dose 2–3 h; high dose 15–30 h Dose in children <37 kg is 50 mg/kg/day.
            Serum salicylate levels may need to be monitored at higher doses.
Salsalate (Disalsid) Mild moderate pain OA, RA Capsules Tablets 2–4 g daily divided dose Hepatic and renal 1h Serum salicylate levels may need to be monitored at higher doses.
Nonselective NSAIDs
Naproxen (Naprosen, Anaprox) Ankylosing spondylitis, arthralgia, bursitis, dental pain, dysmenorrhea, fever, gout arthritis, headache, JRA, mild-moderate pain, myalgia, OA, RA, tendinitis Tablet Extended-release tablet Suspension 500–1,000 mg b.i.d. Hepatic and renal 10–20h Dosage in children age >2yr 10–15 mg/kg/day in 2 divided doses.
            Naproxen may falsely elevate urinary 17 ketosteroid concentrations and interfere with 5-hydroxyindoleacetic acid determination.
            Discontinue 72 h before testing.
Flurbiprofen (Ansaid) Arthralgia, mild-moderate pain, miosis inhibition, myalgia, OA, RA Tablets Ophthalmic solution 50–100 mg b.i.d.–t.i.d.; maximum 300 mg/day Hepatic 3–9h  
Diclofenac (voltaren, Arthrotec) Actinic keratoses, allergic conjunctivitis, ankylosing spondylitis, arthralgia, corneal ulcer, dysmenorrhea, headache, kerato-conjunctivitis, migraine, mild-moderate pain, myalgia, OA, post-operative occular inflammation, RA Tablets
Tablets (combination with misoprostol)
Ophthalmic solution
Topical solution
50–100 mg b.i.d.; maximum 225 mg/day Hepatic 1–2h Cholestyramine reduced bioavailability of diclofenac.
Diclofenac/misoprostol combination contraindicated in pregnancy because of abortifacient effect of misoprostol.
Sulindac (Clinoril) Ankylosing spondylitis, arthralgia, bursitis, gouty arthritis, moderate pain, OA, RA, tendinitis Tablets 150–200 mg b.i.d. Hepatic 8–16 h Nonindicated use in JRA 2–4 mg/kg/day suggested.
Oxaprozin (Daypro) Moderate pain, OA, RA Tablets 600–1200 mg qd Hepatic 36–92 h Non-indicated use in JRA of 10–20 mg/kg/day reported
Diffunisal (Dolobid) Mild-moderate pain, OA, RA Tablets 500–1,000 mg b.i.d. Hepatic 8–12 h 50% increase in acetaminophen plasma concentratrion following administration of diflunisal.
          68–138 h in severe renal disease Diflunisal is a salicylic acid derivative, association with Reye syndrome not known. Avoid in children.
            Diflunisal may falsely elevate serum salicylate levels.
Piroxicam (Feldene) Arthralgia, headache, moderate pain, myalgia, OA, RA Capsule 20 mg q.d. Hepatic with enterohepatic recirculation 50 h Particular caution in high-risk individuals.
Indomethacin (Indocin) Ankylosing spondylitis, arthralgia, bursitis, gouty arthritis, moderate pain, myalgia, OA, patent ductus arteriosus, RA, severe pain, tendinitis Capsules Extended-release capsules Suspension Suppositories Parenteral 25–50 mg t.i.d.–q.i.d. Hepatic Some enterohepatic recirculation Biphasic: 1 h initial; 2.6–11.2 h in second phase Increased serum aminoglycoside concentrations in neonates; monitor aminoglycoside levels closely in all patients.
          Prolonged half-life in neonates and premature neonates Indomethacin augments the hypothalamic-pituitary-adrenal axis response to dexamethasone. Possible false normal results in patients with depressed response.
Ibuprofen (Motrin) Arthralgia, dental pain, dysmenorrhea, fever, headache, JRA, migraine, mild-moderate pain, myalgia, OA, RA Numerous Adults: 400–800 mg t.i.d.–q.i.d.
Children: 5–10 mg/kg
Hepatic 2–4 h Safety demonstrated in children 6 mo of age and older.
Fenoprofen (Nalfon) Arthralgia, mild-moderate pain, myalgia, OA, RA Tablets Capsules 300–600 mg t.i.d.–q.i.d.; maximum 3,200 mg/day Hepatic
Enterohepatic recirculation
2.5–3.0 h Aspirin can decrease fenoprofen plasma concentrations by 50% and reduce half-life.
            Phenobarbital can decrease plasma concentrations of fenoprofen. Monitor barbiturate levels after initiation or withdrawal of fenoprofen.
            Elevated free and total triiodithyronine plasma concentrations by some methods
Etodolac (Lodine) Arthralgia, bone pain, dental pain, mild pain, moderate pain, myalgia, OA RA Tablets
Extended release tablets
600–1,200 mg daily Hepatic 6–7 h  
Ketoprofen (Orudis) Arthralgia, dental pain, dysmenorrhea, fever, headache, mild-moderate pain, myalgia, OA, RA Capsules
Extended release capsules
Tablets
75 mg t.i.d. or 50 mg q.i.d. Hepatic 1.1–4 h Increased plasma concentration of ketoprofen when administered with probenecid.
Ketorolac (Toradol) Allergic conjunctivitis, arthralgia, moderate pain, myalgia, occular pain, ocular pruritus, photophobia, post-operative ocular inflammation Tablets
Parenteral (i.m. or i.v.)
Ophthalmic solution
30 mg i.m./i.v. every 6h; maximum 120 mg/day; do not use longer than 5 days Hepatic Biphasic; terminal phase 4–6 h Elimination half-life of ketorolac is doubled during administration with probenecid. Concomitant use should be avoided.
      10 mg p.o. every 4–6 h; maximum of 40 mg daily for 5 days     Parenteral ketorolac can enhance the muscle relaxant effect of non-depolarizing skeletal muscle relaxants. Caution with concomitant use.
Meclofenamate, mefenamic acid (Ponstel) Arthralgia, dysmenorrhea, mild-moderate pain, OA, RA Capsule 50–100 mg t.i.d.–q.i.d.; maximum 400 mg/day Hepatic 2h Mefenamic acid may cause false-positive test result for urinary bile.
      Mefenamic acid: 250 mg every 6 h for 7 days; maximum 1,250 mg/day      
Meloxicam (Mobic) OA Tablets 7.5–15 mg q.d. Hepatic 15–30 h Cholestyramine may increase clearance meloxicam.
            No platelet inhibition at indicated doses.
Nabumetone (Relafen) Moderate pain, OA, RA Tablets 1,000 mg q.d.; maximum dose of 2,000 mg q.d. Hepatic 24 h  
Tolmetin (Tolectin) Arthralgia, JRA, moderate pain, myalgia, OA, RA Tablets Capsules 400 mg t.i.d.–q.i.d.; maximum dose 2,000 mg/day Hepatic Biphasic: initial 1–2 h; terminal 5 h Dosage in children age 2 yr and above 5–7 mg/kg/dose p.o. every 6–8 h.
            False-positive reaction for proteinuria on acid precipitation test; no effect on urine dipstick test for protein.
Selective COX-2 inhibitors
Celecoxib (Celebrex) Bone pain, dental pain, dysmenorrhea, FAP, headache, moderate-severe pain, OA, RA Capsules 100–200 mg b.i.d.; 400 mg b.i.d. in FAP Hepatic 11 h Reduce dose 50% in setting of moderate liver dysfunction.
            Fluconazole inhibits celecoxib metabolism in the liver. Use lowest celecoxib dose with concomitant fluconazole.
Rofecoxib (Vioxx) Bone pain, dental pain, dysmenorrhea, headache, mild-moderate pain, OA, RA Tablet Suspension 12.5–25 mg q.d.; 50 mg q.d. for 5 days for pain Hepatic 17 h  
Valdecoxib (Bextra) Dysmenorrhea, OA, RA Tablet 10 mg q.d.; 20 mg q.d. as needed for dysmenorrhea Hepatic 8–11h  

   b.i.d., twice daily; FAP, familial adenomatous polyposis; i.m., intramuscularly; i.v., intravenously; JRA, juvenile rheumatoid arthritis; OA, osteoarthritis; p.o., by mouth; q.d., daily; RA, rheumatoid arthritis; t.i.d., three times daily.
Among the nonselective NSAIDs are several subclasses grouped by chemical structure (Table 31.2). Despite the tradition of grouping nonselective NSAIDs by chemical structure, the designation carries no important significance for efficacy or safety. The third class of NSAIDs is the selective COX-2 inhibitors, which have negligible inhibition of COX-1 at therapeutic tissue concentrations (5).
TABLE 31.2. Chemical classification of nonselective nonsteroidal antiinflammatory drugs (NSAIDs)

Chemical class Associated NSAIDs

Acetic acids Diclofenac, indomethacin, sulindac, tolmetin
Fenamates Meclofenamate, mefenamic acid
Napthylalkanones Nabumetone
Oxicams Meloxicam, piroxicam
Propionic acids Enoprofen, flurbiprofen, ibuprofen, ketoprofen, naproxen, oxaprozin
Pyranocarboxylic acid Etodolac
Pyrrolizine carboxylic acid Ketorolac

Mechanism of Action
Despite the use of NSAIDs dating back to the early history of medicine, their mechanism of action was not identified until 1971 (9). The primary mechanism of action of NSAIDs is the inhibition of COX, a synthetic enzyme for prostaglandins (PGs). NSAIDs also have other potential mechanisms of action that may contribute to their clinical effectiveness.
Reduction of Prostaglandin Synthesis
The synthetic pathway for PG is reviewed in detail in Chapter 23. Briefly, arachidonic acid, the precursor to PGs, is produced in cells by cleavage of membrane phospholipids by phospholipase A. Cyclooxygenase and lipoxygenase convert arachidonic acid into PGs and leukotrienes, respectively (Fig. 31.2). PGG2 and PGH2 are the products of COX, and are subsequently converted to a variety of prostanoids by tissue-specific isomerases (5). For example, PGH2 is converted to thromboxane A2 in platelets.
FIG. 31.2. Prostaglandin synthesis pathway with sites of nonsteroidal antiinflammatory drug inhibition.
Through the independent work of several laboratories, two isoforms of COX, designated COX-1 and COX-2, have been identified (10,11,12,13,14,15). COX-1 and COX-2 are associated with membranes in the endoplasmic reticulum and cell nucleus (16). As a generalization, COX-1 is constitutively expressed in most cell types and produces PGs important in normal tissue homeostasis, such as the maintenance of gastrointestinal mucosa and regulation of platelet function (17,18). COX-2 is induced in response to a variety of proinflammatory mediators, growth factors, and tumor promoters (13,14). This dichotomy is somewhat oversimplified because COX-2 is constitutively expressed in the kidney and the central nervous system, and COX-1 expression has also been shown to be regulated in some sites (19,20,21,22).
The inhibition of COX enzyme activity is the primary mechanism of action of NSAIDs as a class. Both isoforms of COX are inhibited by aspirin and NSAIDs. Aspirin is unique in that it irreversibly inhibits both isoforms of COX by acetylating a serine moiety (serine 530 of COX-1 and serine 516 of COX-2) (23,24). Two mechanisms of COX inhibition by nonselective NSAIDs have been identified. First, NSAIDs can mediate time-independent inhibition of COX that is dependent on drug concentration. Second, some NSAIDs (e.g., indomethacin and flurbiprofen) have the additional feature of inducing time-dependent structural changes in the active site of COX that may lead to near-irreversible inhibition of enzyme activity (25). The clinical significance of these distinct inhibitory mechanisms has not been investigated.
Aspirin and the nonselective NSAIDs inhibit both isoforms of COX with varying ratios of effect on COX-1 and COX-2. The fundamental hypothesis that derived from the discovery of COX-1 and COX-2 was that the therapeutic benefits of NSAIDs resulted from COX-2 inhibition, whereas the adverse events associated with NSAIDs resulted from COX-1 inhibition. Highly selective COX-2 inhibitors were developed with the objective of ameliorating the harmful effects associated with traditional NSAIDs. This objective has been partially achieved; however, COX-2 inhibitors possess many of the adverse renal effects observed with nonselective NSAIDs (see later section on Adverse Events Associated with NSAIDs).
Other Potential Mechanisms of Action
Many NSAIDs possess pharmacologic properties distinct from COX inhibition. The salicylates, with the exception of aspirin, have minimal inhibitory action on COX. Yet, salicylates demonstrate significant antiinflammatory activity and inhibit the production of PG synthesis in cell culture and in whole animals (26,27). The mechanism of PG inhibition by salicylates probably relates to suppressing the expression of COX itself, rather than inhibiting its function (26).
Other pharmacologic properties of NSAIDs include the inhibition of transcription factors, cell growth factors, and molecules that regulate apoptosis. At supratherapeutic concentrations, sodium salicylate inhibits the gene transcription regulator nuclear factor κB, which may, in part, contribute to reduced expression of chemokines and nitric oxide, and reduced tumor necrosis factor-induced signaling activity (28,29,30,31). COX-2 selective and nonselective
P.682

P.683

P.684

P.685

P.686

NSAIDs have also been shown to inhibit angiogenesis through inhibition of mitogen-activated protein kinase (ERK2) in endothelial cells (32). Finally, when COX-2 is acetylated by aspirin, it acquires the ability to synthesize 15-hydroxyeicosatetraenoic acid, which is metabolized by 5-lipoxygenase to yield the antiinflammatory eicosanoid 15-epi-lipoxin A4 (33,34,35). This may contribute to the antiinflammatory actions of aspirin.
PHARMACODYNAMICS
Antiinflammatory Effect
PGs directly or indirectly mediate many of the manifestations of inflammation. The antiinflammatory properties of NSAIDs correlate with their ability to inhibit PG synthesis. The ability to inhibit COX is one method for assessing antiinflammatory activity. In vitro whole blood assays of COX-1 and COX-2 inhibition are used as a surrogate marker of antiinflammatory effectiveness and as a method for assessing the relative selectivity of an agent for COX-1 versus COX-2. An in vivo method for assessing the antiinflammatory effectiveness of NSAIDs is the administration of proinflammatory irritants, such as carrageenin, into the hind paw of rats. Other models include the induction of adjuvant arthritis, which results from administration of mycobacteria and oil into rat paws. Effective NSAIDs diminish the swelling and inflammatory infiltrates within the affected paw. Most compounds must demonstrate antiinflammatory effects in numerous in vitro and in vivo models to be considered for clinical development as an NSAID.
Analgesic Effect
One of the primary uses of NSAIDs is to relieve pain. The NSAIDs were originally thought to influence pain sensation strictly in the peripheral nervous system, but recent data indicate that both peripheral and central pain pathways are inhibited. Three types of pain are recognized: physiologic pain, inflammatory pain, and neuropathic pain (36). Physiologic pain is characterized by a high threshold for stimulation, transient symptoms, and well-localized sensation. Inflammatory pain is characterized by heightened sensitivity of normal pain stimuli in response to tissue inflammation or injury. Neuropathic pain results from injury to neurons in the periphery or in the spinal cord and is characterized by hyperalgesia to noxious stimuli, allodynia, and spontaneous onset of pain symptoms. PGs are hypothesized to play a role in the induction of all three types of pain, and may act in both the peripheral and central nervous system.
Inflammation or injury in peripheral tissues leads to local up-regulation of PGs and other inflammatory mediators. This coincides with markedly increased expression of COX-2 by local tissue cells and infiltrating inflammatory cells (37). Evidence indicating that PGs induce hyperalgesia at peripheral sites of inflammation includes the induction of hyperalgesia by local injection of PGE2 and reduction in the pain response by administration of neutralizing monoclonal antibodies to PGE2. Therefore, an important aspect of the analgesic properties of NSAIDs relates to inhibition of tissue PG production in sites of inflammation or injury.
PGs produced in the central nervous system (CNS) also regulate pain symptoms. Both COX-1 and COX-2 are expressed in the CNS (38,39). COX-2 is expressed constitutively in the CNS, and is also up-regulated by peripheral inflammatory stimuli. Both constitutive and induced
P.687

COX-2 expression is inhibited by systemic glucocorticoids (38,40). COX-2 is the predominant isoform expressed in the rat spinal cord (41,42), whereas COX-1, but not COX-2, is expressed in nociceptive neurons in the dorsal root ganglion (43). During systemic inflammatory stimulation, COX-1, but not COX-2, is stably expressed in dorsal root ganglia, whereas COX-2 is markedly up-regulated within the spinal cord.
Receptors for PGs are widely expressed in neuronal cells of the CNS. The EP3 receptor for PGE2 is expressed in ascending nociceptive pathways of the pontine parabrachial nucleus (44) and in thalamic nuclei (45) of rats. EP3 expression has also been detected in interleukin-1 (IL-1)-responsive neurons of nociceptive pathways in the preoptic area of the brain (46). Interestingly, COX-2 and EP3 expression colocalized to specific elements of the rat spinal cord that are associated with pain transmission (42). The expression of COX in association with receptors for PGs provides strong evidence for a role for PGE-induced hyperalgesia within the CNS (36).
Together these data demonstrate that PGs are important in regulating nociception in both the peripheral and CNS. The analgesic properties of NSAIDs most likely result from central and peripheral inhibition of PG production. Although comparative trials of the analgesic efficacy of NSAIDs have not been systematically studied, it is possible that variation in the analgesic effectiveness of NSAIDs may relate to differences in penetration of the drug into the CNS.
Antipyretic Effect
NSAIDs are effective at reducing fever in humans. Most evidence indicates that the febrile response is mediated through the production of PGE2 in the CNS (47). Evidence supporting this includes the detection of increased levels of PGE2 in the hypothalamus and the third ventricle during fever. Furthermore, deletion of the PGE2 receptor EP3 in mice abrogates the febrile response to pyrogens such as IL-1β and lipopolysaccharide (48). The production of PGE2 in the hypothalamus leads to the release of cyclic adenosine monophosphate, which functions as a neurotransmitter for neurons in the thermoregulatory center.
COX-2 is the primary source of PGE2 that leads to a febrile response. Mice deficient in COX-1 exhibit normal febrile responses to intraperitoneal administration of lipopolysaccharide or IL-1β (49), whereas COX-2-deficient mice do not develop fever in response to these stimuli (49). The highly selective COX-2 inhibitor rofecoxib reduced naturally-occurring fever in humans (50). These data indicate that NSAIDs reduce fever by inhibiting COX-2 activity in the CNS. The mechanism by which acetaminophen reduces fever despite negligible COX-2 inhibitory activity in vitro remains unresolved. Some evidence suggests that acetaminophen is modified by oxidation in the CNS to gain COX inhibitory activity. Evidence for a third isoform of COX expressed in the CNS has also been presented (51).
Antiplatelet Effect
Among the benefits of NSAIDs, the role of low-dose aspirin in inhibiting platelet function has perhaps had the most significant impact on public health (52). Aspirin, but not the nonacetylated salicylates, irreversibly inhibit COX-1 in platelets, resulting in the inability to produce thromboxane A2 in response to platelet activation (24). After exposure to aspirin, platelets are incapable of thromboxane production for the remainder of their circulating life span. As a result, the bleeding time of individuals treated with aspirin is prolonged for about 3 days after discontinuation of aspirin.
The nonselective NSAIDs inhibit COX reversibly. The effect of nonselective NSAIDs on platelet function depends on the drug’s avidity for COX-1 and its half-life. Some NSAIDs, such as naproxen, have significant platelet-inhibiting function that lasts for several hours after dosing. Selective COX-2 inhibitors have no effect on platelet function because COX-2 is not expressed in platelets.
Other Effects
NSAIDs also have other unique pharmacodynamic properties. A reduction in uterine PG levels induces closure of the ductus arteriosis shortly after birth. Indomethacin is used to induce closure of a patent ductus arteriosis in infants. The uterine cramping of primary dysmenorrhea is also mediated by PG production. The pain associated with uterine cramping represents an inflammatory pain process associated with spasm of the uterine musculature. NSAIDs are uniquely effective in treating the pain of primary dysmenorrhea (53,54,55).
A role for NSAIDs in the chemoprevention of malignancy and Alzheimer disease has been the focus of much recent research (56,57). The rationale for this work is the observation that COX-2 expression is up-regulated in some tumors and the amyloid plaques of Alzheimer disease.
Interest in chemoprevention of cancer using NSAIDs originated with the observation that sulindac-induced polyp regression in patients with established familial adenomatous polyposis (FAP) (58). A subsequent randomized trial demonstrated that both sulindac and celecoxib caused regression of polyps in established FAP (59). Numerous epidemiologic studies have demonstrated that aspirin and nonselective NSAID use is associated with reduced risk for nonfamilial colon polyps and colon cancer (60,61,62,63,64,65,66,67,68,69,70,71). Two recent randomized controlled trials have shown that low-dose aspirin reduces the risk for recurrent colon polyps in patients with a prior history of benign adenomas or carcinoma of the colon (72,73). Although these findings are intriguing, endoscopic surveillance remains the method of choice for the prevention of colon cancer (74).
P.688

Alzheimer disease is associated with inflammatory responses in regions of the brain affected by plaque formation. In addition, COX-induced oxidation influences glutamate signaling pathways in the brain that may damage neurons (75,76). Several observational studies have suggested a reduced risk for Alzheimer dementia in patients taking NSAIDs (57). Regular use of NSAIDs by participants in the Rotterdam Study, a prospective cohort study, had a significantly lower risk [relative risk (RR) 0.20; 95% confidence interval (CI), 0.05–0.83] for developing dementia during follow-up (77). However, a randomized controlled trial of rofecoxib, naproxen, or placebo in patients with mild-to-moderate Alzheimer disease showed no benefit in slowing the progression of dementia (78). Additional studies are needed before definitive conclusions can be drawn on the effect of NSAIDs on Alzheimer disease.
ADVERSE EVENTS ASSOCIATED WITH NSAIDS
Comparative studies have shown that the efficacy of NSAIDs for most rheumatologic disorders is similar regardless of the particular drug used. Therefore, preventing and anticipating adverse events is a primary consideration when NSAIDs are prescribed. About one third of patients using NSAIDs will develop a persistent drug-related adverse event, leading 10% to discontinue treatment (79). NSAID use also increases the risk for hospitalization and death. Over-the-counter availability of NSAIDs and inadequate assessment of risk factors for NSAID toxicity by physicians are factors that may contribute to the number of drug-related adverse events (80). The spectrum of drug toxicities associated with NSAIDs is shown in Table 31.3 and reviewed in this section.
TABLE 31.3. Drug toxicities associated with nonsteroidal antiinflammatory drugs (NSAIDs)

Gastrointestinal
   Gastroduodenal ulcer
   Dyspepsia
   Lower gastrointestinal tract (ulcer, stricture, hemorrhage)
   Hepatic toxicity (elevated liver enzymes, hepatic failure)
Renal
   Peripheral edema
   Acute reduction in renal function
   Hyperkalemia
   Interstitial nephritis with nephrotic syndrome
   Renal papillary necrosis
   Analgesic nephropathy
Cardiovascular
   Exacerbation of hypertension
   Exacerbation of congestive heart failure
   Acute coronary syndromes (?)
Allergic, Pseudoallergic and Immunologic
   NSAID-induced rhinitis and asthma
   NSAID-induced urticaria and angioedema
   Anaphylaxis
   Cutaneous reactions
   Hypersensitivity pneumonitis
   Aseptic meningitis
Other
   Salicylate toxicity
   Reye syndrome

Gastrointestinal Toxicity
Gastrointestinal toxicity is perhaps the most important adverse event related to NSAIDs. Much of the history of NSAIDs has been driven by efforts to reduce gastrointestinal toxicity. Aspirin was developed as a solution to the high rate of gastrointestinal intolerance to sodium salicylate (1,3). By the 1930s there was evidence that aspirin caused gastric ulceration (81). The advent of nonaspirin NSAIDs was driven, in part, by efforts to identify a better tolerated alternative to aspirin (82). However, the NSAIDs were also demonstrated to cause substantial gastrointestinal injury (83,84,85). The discovery of the mechanism of action of NSAIDs (9) and the identification of two isoforms of COX (10,11,12,13,14,15) led to renewed efforts to improve the gastrointestinal safety of NSAIDs through the development of COX-2 inhibitors.
Epidemiology
Several gastrointestinal adverse events are associated with NSAIDs. The most common is dyspepsia. However, the most important is gastroduodenal ulceration. The epidemiology of NSAID-induced gastroduodenal ulcer has been well established (reviewed in reference 79). Serious gastrointestinal complications occurred in 0.7% of OA patients who took an NSAID for 1 year (86). The incidence was 1.3% to 1.5% annually among patients with RA (86,87). Low-dose aspirin used for prophylaxis of cardiovascular events has also been associated with a twofold increased risk for bleeding peptic ulcer (88). Mortality associated with NSAID-induced gastrointestinal toxicity is estimated to be 0.22% per year, or over 16,500 NSAID-related deaths annually (79,86).
Mechanisms
The mechanisms by which NSAIDs cause gastrointestinal injury include both local and systemic effects (89). Local injury occurs as a result of the direct toxic effects of NSAIDs on mucosal cells. Aspirin and most NSAIDs are weak acids that remain in their nonionized form in the acidic environment of the stomach. As such, NSAIDs readily diffuse through the gastric mucous layer into the relatively neutral environment adjacent to epithelial cells. There, aspirin and NSAIDs become ionized, releasing hydrogen ions that injure the epithelium and increase mucosal permeability. The primary systemic effect of NSAIDs is inhibition of COX-1, which is necessary for the production of PGs important in maintaining the integrity of the gastric mucosa. The systemic effects of NSAIDs on mucosal PG
P.689

production in the gastrointestinal tract are thought to be the most important factor leading to ulcer formation (79).
Risk Factors
Risk factors for NSAID-induced gastrointestinal ulcers relate predominantly to comorbid medical conditions. In fact, prior symptoms of dyspepsia were reported by only 19% to 41% of patients using NSAIDs who presented with serious gastroduodenal ulcer (90,91). The risk for severe gastrointestinal complications is associated with advancing age, prior history of peptic ulcer, concomitant corticosteroid use, high doses of NSAIDs, concomitant anticoagulation, and severe systemic illness (79,92) (Table 31.4,Fig. 31.3). Other possible risk factors include Helicobacter pylori infection, smoking, and alcohol use. These data demonstrate that dyspepsia is not a sensitive indicator of gastrointestinal complications from NSAIDs. Instead, risk factors for gastrointestinal toxicity should be assessed before NSAIDs are prescribed, even for short-term use.
TABLE 31.4. Risk factors for severe nonsteroidal antiinflammatory drug (NSAID)-induced gastrointestinal ulcer

Established risk factors
   Advanced age
   History of ulcer
   Concomitant use of corticosteroids
   Higher doses of NSAIDs, including use of more than one NSAID
   Concomitant administration of anticoagulants
   Serious systemic disorder
Possible risk factors
   Concomitant infection with Helicobacter pylori
   Cigarette smoking
   Consumption of alcohol

   From Wolfe MM, Lichtenstein DR, Singh G. Gastrointestinal toxicity of nonsteroidal antiinflammatory drugs. N Engl J Med 1999;340:1888–1899, with permission.
FIG. 31.3. Risk of nonsteroidal antiinflammatory drug-induced upper gastrointestinal bleeding increases with age. The relative risk for bleeding increases from 1.8 for individuals age 50 to 80 years to 9.2 for individuals older than 80 years. (Asterisk signifies individuals 25–49 years of age as a reference group) (92).
Two types of risk factors for gastrointestinal toxicity have been defined in patients using NSAIDs: baseline risk and NSAID-attributable risk (93). Baseline risk is attributable to preexisting factors such as advanced age, prior PUD, or H. pylori infection. NSAID-attributable risk is that which exists in those patients without other ulcer risk factors. Although the incidence of gastrointestinal toxicity may be highest in those with high baseline risk (e.g., multiple ulcer risk factors), strategies used to reduce the risk for ulcers may be most effective in those patients who have only NSAID-attributable risk.
Prevention Strategies
Several strategies have been shown to reduce the incidence of gastrointestinal injury in patients using NSAIDs. When evaluating clinical trials of interventions to protect against NSAID-induced ulcers, it is important to draw a distinction between reduction in symptomatic ulcers and reduction in surrogate markers of ulcers such as superficial gastric erosions or ulcers detected by endoscopy. The most clinically-relevant outcome is reduction in painful ulcers, and ulcer complications such as perforation, bleeding, and obstruction. However, given the low frequency of these events in patients using NSAIDs, few studies have been performed that have sufficient power to detect a reduction in clinically-important ulcers. A summary of clinical trials of strategies to reduce NSAID-induced ulcers is summarized in the next section (Table 31.5).
TABLE 31.5. Effectiveness and tolerability of strategies to reduce NSAID-induced gastrointestinal ulcer risk
The following interventions have been shown to reduce endoscopic or symptomatic ulcers in patients taking NSAIDs:
  • Misoprostol. Misoprostol is an oral PGE1 analogue that is approved for the prevention of NSAID-induced gastric ulcers. Its pharmacologic actions may be mediated by a cytoprotective effect on gastric mucosa and by inhibition of acid secretion from parietal cells. Coadministration of misoprostol with NSAIDs reduced endoscopic ulcers by 71% and symptomatic ulcers by 51% (87,94). Disadvantages of misoprostol include a high incidence (27%) of diarrhea and other gastrointestinal symptoms that often leads to discontinuation of treatment; especially at the most effective dosage of 800 μg/day (94). The requirement for dosing twice to four times daily may also lead to poor compliance.
  • Proton pump inhibitors. Proton pump inhibitors (PPIs) bind to the H+/K+ adenosine triphosphatase pump on the surface of parietal epithelial cells and inhibit secretion of hydrogen ions into the gastric lumen. PPIs are indicated for the treatment of gastric and duodenal ulcers. When used in combination with NSAIDs, PPIs reduce the risk for endoscopic gastroduodenal ulcers by 77%. PPIs are well tolerated and significantly reduce dyspepsia symptoms
    P.690

    P.691

    in patients taking NSAIDs (94). No trials have assessed the effectiveness of PPIs in preventing clinically significant NSAID-induced ulcers. However, among patients with a history of gastrointestinal bleeding and H. pylori infection, a PPI reduced the risk for rebleeding more effectively than eradication of H. pylori (4.4% vs. 19%) (95). These data provide circumstantial evidence that PPIs reduce the risk for clinically important ulcers.
  • High-dose histamine (H2) receptor antagonists. H2 antagonists reduce gastric acid secretion and accelerate the healing of peptic ulcers. When administered at standard doses, H2 antagonists prevented endoscopically-detected duodenal, but not gastric ulcers in subjects taking NSAIDs. Double-dose H2 antagonists reduced both gastric and duodenal ulcers by 56% and 74%, respectively (94). H2 antagonists are well tolerated and may reduce the severity of dyspepsia symptoms associated with NSAID use.
  • Highly selective COX-2 inhibitors. The use of COX-2 inhibitors to minimize ulcer risk is an alternate strategy to combinations of drugs used in conjunction with nonselective NSAIDs. The rationale for using COX-2 inhibitors to prevent gastroduodenal ulcers was previously described. Pooled analysis of several endoscopic studies in subjects using celecoxib or rofecoxib demonstrated a 76% reduction in gastroduodenal ulcer risk compared with nonselective NSAIDs (94). Furthermore, two large randomized controlled trials have shown that both celecoxib (CLASS trial) and rofecoxib (VIGOR trial) reduce the risk for symptomatic gastroduodenal ulcers when compared with nonselective NSAIDs (96,97).
There exists ongoing controversy regarding the interpretation of the results of both trials. In the CLASS study, the predefined primary end point of a reduction in complicated ulcers was not achieved, whereas the secondary end point of reduction in complicated and symptomatic ulcers was achieved. Factors that may have contributed to this outcome include the inclusion in the CLASS trial of patients using low-dose aspirin. The VIGOR trial achieved its primary end point of reduction in complicated and symptomatic ulcers. However, the overall rate of serious adverse events was similar in both treatment arms, owing to a greater incidence of cardiovascular events in the rofecoxib group. Hypotheses that may explain these findings are discussed below in the section on cardiovascular adverse events.
Each of the strategies described above shows substantial evidence of effectiveness in preventing NSAID-induced ulcer. The only agents shown to reduce the risk for clinically-important ulcers are misoprostol, celecoxib, and rofecoxib. Given the high rate of gastrointestinal symptoms and the frequent dosing interval required with misoprostol, the COX-2 inhibitors have emerged as the preferred method for avoiding gastroduodenal ulcers in patients using NSAIDs. Regardless, all patients initiating therapy with NSAIDs who have one or more risk factors for gastroduodenal ulcer are candidates for treatment with one of the strategies above for reducing ulcer risk. Ongoing studies will answer important remaining questions regarding the cardiovascular safety of COX-2 inhibitors, the impact of daily low-dose aspirin on the gastrointestinal safety of COX-2 inhibitors, and the relative gastrointestinal safety of COX-2 inhibitors versus nonselective NSAIDs administered with a PPI.
Role of Helicobacter pylori Infection
Infection with H. pylori represents the primary overall risk factor for peptic ulcer disease in the general population. The possibility that H. pylori and NSAIDs interact to influence the risk for NSAID-induced ulcer has been the focus of several clinical trials (98). Most studies have shown that H. pylori does not increase the risk for endoscopic or clinical ulcer formation in NSAID users (99,100,101,102,103,104), whereas three studies indicated that eradication of H. pylori reduces the risk for recurrent ulcer (105,106,107). Studies investigating the effect of H. pylori infection on ulcer healing have yielded mixed results (108,109,110,111). Two studies showed no effect (110,111), and two studies showed the counterintuitive observation that H. pylori infection enhanced ulcer healing (108,109). Several case control studies indicate an increased risk for ulcer bleeding in patients with H. pylori infection who take NSAIDs (98). However, two prospective randomized trials demonstrate no benefit of H. pylori eradication in preventing recurrent bleeding in patients taking NSAIDs (95,112). The conflicting results regarding the importance of H. pylori in NSAID-induced ulcer require further studies. At present, screening for H. pylori infection in patients initiating NSAID therapy is not indicated (98). However, patients diagnosed with ulcer, regardless of their use of NSAIDs, should be screened for H. pylori and treated if infection is detected.
Dyspepsia
Dyspepsia comprises a number of gastrointestinal symptoms that are common in users and nonusers of NSAIDs (113,114). Its pathophysiologic mechanism is not well understood. Dyspepsia symptoms occurred at a rate of 69 to 85 events per 100 patient-years of therapy in a pooled analysis of comparative clinical trials of rofecoxib and nonselective NSAIDs (115). Dyspepsia is also a primary reason for discontinuation of NSAID therapy. The percentage of patients in the VIGOR and CLASS trials who discontinued treatment due to dyspepsia and related gastrointestinal symptoms (excluding ulcers) was as follows: naproxen 4.9%, rofecoxib 3.5%, celecoxib 8.7%, and diclofenac/ ibuprofen 10.7% (96,97). It is likely that the discontinuation rate is higher outside of a clinical trial setting. The clinician is challenged by the finding that dyspepsia symptoms correlate with gastroduodenal ulcers in just 50% of patients, and up to 40% of patients with gastroduodenal ulcer have
P.692

no prior symptoms of dyspepsia (116,117). Both H2 receptor antagonists (118,119,120,121) and PPIs (108,109) reduce the incidence of dyspepsia in patients taking NSAIDs. However, Singh and colleagues have shown that asymptomatic patients taking standard doses of H2 receptor antagonists had a higher rate of ulcer complications than those not using these medications (91). It is possible that H2 receptor antagonists masked symptoms of dyspepsia but did not protect against the development of clinically-important ulcers (79). Patients who develop dyspepsia while on NSAIDs should be treated with a PPI rather than an H2 receptor antagonist in order to diminish dyspepsia symptoms, and potentially reduce the risk for ulcer formation. Gastroduodenal ulcer should be ruled out in patients who do not have prompt resolution of dyspepsia symptoms, because ulcer may be present in about 50% (116).
Lower Gastrointestinal Tract Adverse Events
Lower gastrointestinal tract complications of NSAIDs include stricture, ulceration, and hemorrhage of the small bowel or colon (122,123,124,125,126,127,128,129,130,131). Estimates of the annual rate of lower tract complications range from 0.9% to 4% per year (130,131,132). Recent prospective data indicate that COX-2 inhibitors may be less likely to cause complicated lower gastrointestinal tract adverse events (132).
Hepatic Toxicity
The spectrum of NSAID-induced hepatic toxicity ranges from elevated liver transaminases to fulminant liver failure (133). Severe liver injury due to NSAIDs is rare, resulting in 2.2 hospitalizations per 100,000 population annually (134,135). Hepatocellular injury, characterized by elevated transaminases, is the most common form of NSAID-induced liver toxicity (133). Laboratory abnormalities usually resolve within several days to weeks if the NSAID is discontinued. Small elevations in transaminases of less than three times the upper limit of normal do not require discontinuation of NSAIDs unless they are associated with clinical signs of liver injury, reduced serum albumin levels, or prolonged prothrombin time which indicates impaired liver synthetic function. Cholestatic injury or mixed hepatocellular/cholestatic injury has also been described. Sulindac is one of the most common causes of NSAID-induced liver toxicity, and typically induces cholestatic or mixed liver injury (133,136). Aspirin causes predictable dose-related hepatocellular toxicity. Manifestations usually occur when blood levels are 25 mg/dL or higher (133). The nonacetylated salicylates may induce non-dose-related idiosyncratic toxicity. Several NSAIDs have been associated with a greater than acceptable risk for fulminant hepatic failure and have been removed from the market; they include, benoxaprofen, ibufenac, and cinchophen (133).
Renal Toxicity
Up to 5% of patients taking NSAIDs will develop a clinically-apparent renal adverse event (137). The most common manifestation is peripheral edema (138). Other toxicities include acute reduction in renal function, hyperkalemia, interstitial nephritis, and papillary necrosis (139). Renal adverse events are almost always reversible when detected early (140). Both nonselective and selective COX-2 inhibitors seem to have similar effects on the kidney (141).
COX-1 and COX-2 are expressed in discrete locations in the human kidney (19,142) and produce PGs important in regulating sodium and water reabsorption (20,143,144,145). COX-1 is expressed in the collecting duct, interstitial cells, the endothelium, and vascular smooth muscle cells. COX-2 is expressed in endothelial cells and vascular smooth muscle cells, as well as in podocytes of the glomerulus. Furthermore, COX-2 expression could be detected in macula densa cells of human kidney tissue obtained from patients with hyperreninemia (146). The constitutive and regulated expression of COX in the kidney illustrates the important role of local PG production in regulating normal and “stressed” kidney function. Decreased production of PGs is the primary mechanism underlying most NSAID-related renal adverse events (140).
Edema
NSAID-induced edema occurs in 1% to 5% of patients (137,147,148), and results from increased sodium reabsorption due to decreased PGE2 production (149,150). Edema is usually reversible and mild, associated with 1 to 2 kg of weight gain (137). Onset is most pronounced during the first several days of NSAID use and may improve after renal handling of sodium returns toward baseline (151,152). Edema does not necessarily correlate with increased mean blood pressure. Some patients may develop severe edema that can progress to CHF (153). Risk factors for peripheral edema include CHF, diuretic use, cirrhosis, diabetes mellitus, renal insufficiency, and older age (138). NSAIDs may be continued in patients with mild edema in the absence of increased blood pressure, reduced renal function, or exacerbation of underlying disease such as CHF. However, patients require frequent and careful monitoring to detect onset of hypertension or the development of electrolyte or renal function abnormalities. In this setting, clinicians must determine whether the clinical benefits of NSAID therapy offset the risks.
Acute Reduction in Renal Function
Renal function is unaffected by NSAIDs in normal, healthy individuals. However, NSAIDs may cause rapid loss of renal function when given to patients with effective intravascular volume depletion or impaired organ perfusion. The mechanism of toxicity is inhibition of PG-
P.693

mediated vasodilation, which results in renal ischemia. Patients at risk for acute loss in renal function include those with CHF, poor underlying renal function, cirrhosis, diabetes mellitus, advanced age, and dehydration due to diuretics or underlying medical conditions (140,154,155,156,157,158,159,160). Acute renal failure is more commonly associated with higher doses of NSAIDs, as well as NSAIDs with longer half-lives (161). Patients with risk factors for acute reduction in renal function should be reevaluated soon after initiating NSAIDs for signs of edema, weight gain, impaired renal function, or hyperkalemia. Acute reduction in renal function is reversible and usually resolves within several days if detected early (140). Renal failure may require dialysis if not detected early (158).
Hyperkalemia
NSAIDs may induce hyperkalemia by inhibiting PG-stimulated renin release in the kidney (140). Reduced renin leads to reduced aldosterone production and subsequent reduction in potassium excretion (140). Risk factors for NSAID-induced hyperkalemia are use of angiotensin-converting enzyme inhibitors, potassium-sparing diuretics, and potassium supplements (140,162,163,164,165). Underlying medical conditions that may contribute to hyperkalemia include heart failure, renal insufficiency, multiple myeloma, and diabetes mellitus (166,167,168,169). Discontinuation of NSAIDs corrects the hyperkalemia. However, NSAID-induced hyperkalemia has been reported to present with renal failure, quadriparesis, and fatal arrhythmia (162,163,170,171,172,173).
Interstitial Nephritis with Nephrotic Syndrome
A rare complication of NSAID use is interstitial nephritis, usually in association with nephrotic syndrome. Onset may range from 2 weeks to 18 months from initiation of NSAIDs (174). The syndrome is unique in that eosinophilia and urine eosinophils are usually not present. Patients usually present with clinical manifestations of nephrotic syndrome. The urine sediment demonstrates microscopic hematuria and tubular epithelial cell casts (140). There are no known risk factors, and the pathophysiologic mechanism is unknown. Renal biopsy demonstrates a unique pattern of interstitial nephritis with minimal change glomerulonephritis (175,176). Proteinuria often resolves with discontinuation of the NSAID. The benefit of corticosteroid treatment for interstitial nephritis has not been definitively shown (157,175). A trial of corticosteroids is recommended if proteinuria has not begun to resolve within 2 weeks of stopping NSAIDs (140).
Renal Papillary Necrosis
Acute renal papillary necrosis is a rare and irreversible form of NSAID-induced toxicity (177,178,179). The usual clinical scenario involves high doses of NSAIDs in a setting of severe dehydration. The clinical presentation may be minimally symptomatic or may mimic passage of a renal stone (140). The mechanism of injury is ischemic necrosis of the distal nephron caused by loss of PG-dependent vasodilation secondary to high local concentrations of NSAIDs. Affected individuals have difficulty forming a maximally concentrated urine (140).
Chronic renal papillary necrosis is the pathologic description of the clinical entity analgesic nephropathy (180,181). Long-term, daily use of drug combinations containing two or more analgesics plus caffeine or codeine is associated with analgesic nephropathy (182). The analgesic combination most commonly associated with nephropathy is aspirin and phenacetin. The latter, a prodrug that is metabolized to acetaminophen, is no longer available in most countries. Other analgesic combinations associated with analgesic nephropathy include aspirin and acetaminophen, and aspirin or acetaminophen in conjunction with the pyrazolone class of analgesic drugs (not available in the United States) (183). There is no indication that acetaminophen or NSAIDs used as single agents cause analgesic nephropathy.
The clinical presentation of analgesic nephropathy is usually limited to polyuria and, occasionally, episodes of microscopic or gross hematuria during sloughing of necrosed renal papillae. As analgesic nephropathy progresses, the clinical manifestations are the nonspecific findings of renal failure (182). Women are affected more often than men. Patients may present with end-stage renal disease as the initial manifestation of analgesic nephropathy. Computed tomography of the abdomen usually demonstrates small kidneys with a “bumpy” contour, and calcifications of the renal papillae (184,185). Discontinuation of analgesic medications may arrest the progression to renal failure if detected early. Urologic malignancy, usually transitional cell carcinoma, may be a late complication in up to 8% of individuals with analgesic nephropathy (186,187,188).
Cardiovascular Adverse Reactions
Hypertension
Meta-analyses of numerous clinical trials performed prior to the advent of COX-2 inhibitors have shown that NSAIDs induce small elevations in blood pressure. Pope and colleagues compiled data from 54 clinical trials and determined that mean blood pressure was increased in NSAID users (189). Normotensive subjects had a mean increase in blood pressure of 1.1 mm Hg, whereas hypertensive subjects had an increase of 3.3 mm Hg. Similarly, Johnson and colleagues pooled data from 50 clinical trials and determined that mean blood pressure increased by 5 mm Hg in patients taking NSAIDs (190). The hypertensive effect was must pronounced in subjects with preexisting hypertension. Certain NSAIDs were more commonly associated with elevated blood pressure. Both meta-analyses found that
P.694

naproxen and indomethacin had the greatest effect, whereas sulindac and aspirin were not significantly different from placebo. Although mean changes in blood pressure were low in the total cohort, some patients experienced large increases in blood pressure while taking NSAIDs.
COX-2 inhibitors have also been shown to have effects on blood pressure that are similar to nonselective NSAIDs. Trials of celecoxib and rofecoxib demonstrated increases in blood pressure (<3 mmHg systolic; <1 mmHg diastolic) that were not statistically significant (191). In a pooled analysis of clinical trials of rofecoxib in patients with OA or RA, the incidence of investigator-defined hypertension was related to daily dose: 12.5 mg, 2.8%; 25 mg, 4.0%; and 50 mg, 8.2% (192). In long-term gastrointestinal safety studies, the incidences of investigator-defined hypertension was 9.7% (rofecoxib 50 mg daily, twice the recommended daily dose) and 5.5% (naproxen 500 mg twice daily) (96). The incidences of investigator-defined hypertension in celecoxib long-term gastrointestinal safety studies were 2.0% (celecoxib 400 mg twice daily, twice the recommended daily dose), 2.0% (diclofenac 75 mg twice daily), and 3.1% (ibuprofen 800 mg three times daily) (97). The conclusion from these results is that the incidence of hypertension in patients taking COX-2 inhibitors is similar to that in patients taking nonselective NSAIDs.
In general, NSAIDs cause small mean increases in blood pressure. However, a small subset of patients may experience clinically significant increases in blood pressure that require intervention—either discontinuation of NSAIDs or intensified antihypertensive therapy. Risk factors for NSAID-induced hypertension include older age and preexisting hypertension (189). In fact, NSAID use is a predictor of hypertension in elderly patients (193). Furthermore, a prospective analysis of elderly patients demonstrated a 70% increased risk for initiating antihypertensive medications in NSAID users versus nonusers (194). All patients initiating long-term NSAID therapy should be monitored for increased blood pressure. Particular caution should be taken in those with preexisting hypertension and older patients.
Congestive Heart Failure
NSAIDs have been associated with CHF in several studies (195,196,197,198,199,200,201,202). Among elderly patients with known heart disease, NSAID use was associated with a 10-fold increased risk for hospital admission for a first episode of CHF, suggesting that NSAIDs may induce CHF (200,201). However, a large 7-year prospective cohort study by Feenstra and colleagues demonstrated that NSAID use did not cause new-onset CHF, but was highly associated with relapse of CHF (202) (Fig. 31.4). Among patients who filled one or more prescriptions for NSAIDs, the risk for incident heart failure was not significantly elevated (RR 1.2; 95% CI 0.8–1.8). However, after a first episode of CHF, patients had a 10-fold increased risk for recurrence (RR 9.9; 95% CI 1.7–57.0) when prescribed NSAIDs. This observation is consistent with the observed risk for fluid retention and elevated blood pressure in patients with reduced left ventricular function who use NSAIDs. Risk factors for exacerbation of CHF by NSAIDs include higher daily dose and the use of NSAIDs with longer half-lives (201). It is estimated that up to 19% of hospital admissions for CHF may be attributable to NSAIDs (201). NSAIDs should be avoided in patients with heart failure. Close monitoring of fluid status, blood pressure, and renal function is mandatory in any patient with a history of heart disease who is taking an NSAID.
FIG. 31.4. Relative risk for exacerbation of congestive heart failure (CHF) in older patients using nonsteroidal antiinflammatory drugs (NSAIDs) (202). Patients without prior heart failure experienced no increased risk for CHF associated with NSAID use. Patients with a prior history of heart failure had a 10-fold increased risk for CHF exacerbation associated with NSAID use.
Myocardial Infarction and Stroke
The importance of low-dose aspirin in the prevention and management of acute coronary syndromes and stroke is well documented (203). A somewhat paradoxic concern relates to a potential association between selective COX-2 inhibitors and increased risk for myocardial infarction (204). Subjects receiving rofecoxib in the VIGOR trial experienced a higher risk for vascular events than those taking naproxen (20 versus 4 myocardial infarctions per 2,699 person-years of follow-up) (96). No increase in vascular events was observed in a metanalysis of randomized trials of rofecoxib (205). Similarly, celecoxib was not associated with an increased risk for vascular events in comparison with diclofenac or ibuprofen in the CLASS trial (97). Plausible hypotheses proposed to explain this observation include the known antiplatelet effects of naproxen relative to rofecoxib (206,207). Alternatively, selective COX-2 inhibition reduces prostacyclin production by endothelial cells, which normally prevents platelet aggregation and causes vasodilation. Prostacyclin inhibition in the absence of inhibition of
P.695

thromboxane through COX-1 inhibition (which induces platelet aggregation) could potentially establish a prothrombotic state that might increase the risk for vascular events (208,209). Additional studies are needed to determine whether COX-2 inhibitors are associated with increased risk for vascular events.
Allergic and Pseudoallergic Reactions to NSAIDs
NSAIDs cause a variety of allergic, pseudoallergic, and immunologic reactions (210,211) (Table 31.3). NSAID-induced rhinitis and asthma (NIRA) is classically characterized by asthma, rhinitis or nasal polyps, and “aspirin sensitivity” (Samter triad). Chronic sinusitis has also been determined to be part of the syndrome of NIRA. Invariably, patients with NIRA have preexisting airway disease that often presents as persistent sinusitis or nasal polyps following a respiratory infection (210,211). Onset usually occurs in adulthood, and women are affected slightly more commonly than men. Symptoms often progress from isolated rhinitis or sinusitis to also include asthma. Exposure to aspirin or an NSAID may result in increased nasal congestion, rhinorrhea, or wheezing within 15 minutes to an hour (210). Symptoms may resolve within an hour or persist for a day or more. NIRA has been reported in 5% to 40% of patients with asthma, depending on the method used to assess sensitivity to NSAIDs (210,212,213).
The mechanism underlying AERS is not well established. The leading hypothesis is that inhibition of COX-1 reduces levels of PGE2, which normally inhibits the production of leukotrienes by 5-lipoxygenase. Loss of 5-lipoxygenase inhibition leads to excessive production of leukotrienes, resulting in increased airway inflammation and bronchial responsiveness (214). Evidence supporting this hypothesis includes the finding of increased levels of urinary leukotrienes in aspirin-sensitive asthmatics compared with asthmatics who are not sensitive (215). Further evidence includes the ability of leukotriene-inhibiting drugs, such as montelukast, to reduce symptoms of asthma in patients with aspirin-sensitive disease (216,217). Because the presumed mechanism of action is not through the development of a specific immune response to the drug, symptoms may occur after the first dose of an NSAID.
There are important implications of NIRA in rheumatology practice. Physicians should be cautious about prescribing NSAIDs for patients with a history of nasal polyps, chronic sinusitis, or asthma. However, less than 10% of patients with asthma in the absence of nasal polyps will have increased symptoms when taking NSAIDs (210). Therefore, patients with asthma should not automatically be excluded from using NSAIDs. Definitive diagnosis of NIRA requires a provocative aspirin challenge, usually performed by an allergist.
Strategies that potentially enable use of aspirin or NSAIDs in patients with NIRA include (a) aspirin desensitization (218), (b) use of a selective COX-2 inhibitor (219,220,221,222,223), and (c) coadministration of a leukotriene inhibitor (216,217). Aspirin desensitization is usually performed by an allergist experienced in the management of patients undergoing provocative oral drug challenges. Progressive doses of aspirin are administered orally until a full dose (325 mg) is tolerated. About two thirds of patients are able to tolerate NSAIDs after 1 year of treatment (218). The patient may continue to use aspirin or another NSAID without concern for NIRA as long as a daily dose is administered. Interruption in aspirin or NSAID dosing for more than 24 hours requires that the desensitization procedure be repeated.
NSAID-induced urticaria/angioedema occurs in individuals with preexisting chronic urticaria or angioedema. Patients with idiopathic urticaria or angioedema may experience an exacerbation of symptoms after administration of an NSAID (224,225). Inhibition of COX-1 is thought to be the primary mechanism underlying increased urticaria or angioedema. Symptoms may occur after the first dose of the NSAID. The diagnosis may be confirmed by oral NSAID challenge. There are no in vitro tests or skin tests to confirm the presence of NSAID-induced urticaria or angioedema. About one third of patients with inactive urticaria symptoms have cross-reactivity to multiple NSAIDs, whereas two thirds of patients with active hives will be cross-reactive to multiple NSAIDs. Management strategies include (a) NSAID avoidance, (b) coadministration of an antihistamine or leukotriene antagonist (226,227,228), (c) a trial of an alternate NSAID, or (d) use of a selective COX-2 inhibitor (229,230,231,232). Desensitization is not effective in alleviating NSAID-induced urticaria or angioedema.
Angioedema and urticaria induced by NSAIDs may occur in patients with no preexisting history of chronic urticaria. Usually patients are sensitive to a single NSAID. The presumed mechanism is the induction of immunoglobulin E (IgE) antibody to the drug or a drug metabolite (210). As a result, symptoms do not occur after the first administration of the NSAID. Readministration of the NSAID after an acute episode of urticaria or angioedema could result in anaphylaxis. The appropriate management strategy is avoidance of the suspect drug and selection of an alternate NSAID when needed. Rarely, patients without a history of chronic urticaria or angioedema may develop urticaria or angioedema in response to multiple NSAIDs. Many of these patients ultimately develop chronic urticaria and would be classified as having NSAID-induced urticaria or angioedema as described above (233). Anaphylaxis may also occur after repeat use of a particular NSAID. The mechanism is presumed to be IgE mediated. Among patients presenting to emergency rooms with anaphylaxis in response to medications, half were sensitive to NSAIDs (234).
A variety of other allergic, pseudoallergic, and immunologic adverse events are associated with NSAIDs (211).
P.696

Cutaneous reactions associated with NSAIDs include maculopapular eruptions, fixed drug eruptions, erythema multiforme, Stevens-Johnson syndrome, leukocytoclastic vasculitis, pseudoporphyria, and photosensitivity responses (235,236,237,238). Aseptic meningitis has been associated with several NSAIDs (239,240). Hypersensitivity pneumonitis has rarely been associated with NSAIDs (241).
Salicylate Toxicity
Overdose of salicylates results in a characteristic syndrome of salicylate toxicity. Acute toxicity results from ingestion of more than 150 mg/kg of sodium salicylate. Lethal toxicity may occur after ingestion of more than 500 mg/kg. Clinical manifestations include mixed respiratory alkalosis and metabolic acidosis, altered mental status, hypernatremia, hypokalemia, dehydration, fever, and hyper- or hypoglycemia. Serum salicylate levels may be used to estimate the severity of intoxication (242). Treatment of acute salicylate overdose includes induction of emesis in conscious patients, gastric lavage, administration of activated charcoal, and supportive therapy.
Chronic salicylate toxicity, or salicylism, results from prolonged high doses of salicylates (242). Generally, the clinical manifestations are associated with administration of 100 mg/kg/day for 2 or more days. Factors that may lead to chronic intoxication include dehydration or concomitant use of bismuth compounds that contain salicylates. Symptoms include hearing loss, tinnitus, dizziness, confusion, tachycardia, nausea, and vomiting. Hyperventilation, mixed respiratory acidosis and metabolic acidosis, and hypoglycemia may also be present. Measurement of the serum salicylate level is not helpful in assessing the severity of chronic intoxication. Mild intoxication is treated by discontinuing salicylates. More severe toxicity is treated with supportive care and measures described above for acute intoxication.
Reye Syndrome
Reye syndrome is characterized by acute onset of fatty liver and encephalopathy in children less than 15 years of age (243). Typically, the syndrome is preceded by a viral infection such as chicken pox or influenza. Use of aspirin has been linked to Reye syndrome in several case-control studies (244,245). Clinical manifestations include persistent vomiting and stupor that progresses to seizures and coma. Laboratory findings may include elevated serum transaminases, prolonged prothrombin time, elevated serum ammonia, metabolic acidosis, and hypoglycemia. Jaundice is typically absent. Treatment is with infusions of fresh frozen plasma and glucose. The mortality rate exceeds 50%. Reye syndrome is rare, perhaps owing to the nearly universal avoidance of aspirin in children (246).
DRUG INTERACTIONS
Drug-drug interactions may occur among the entire class of NSAIDs, or may be unique to a particular NSAID (247,248). Known interactions include the displacement of drugs from serum proteins, altered clearance or metabolism of drugs, and altered pharmacologic activity of drugs (e.g., antihypertensives) (248). Most drug-drug interactions involving NSAIDs are clinically insignificant in most patients. However, the marked variation of individual responses to drugs can lead to significant toxicity due to interactions.
One of the most common drug-drug interactions involving NSAIDs is blunting of the effect of antihypertensives. As a class, NSAIDs elevate mean blood pressure by 5 mm Hg in patients with controlled hypertension (189). This effect was most pronounced in patients treated with angiotensin-converting enzyme inhibitors, β blockers, or calcium channel blockers. NSAIDs also blunt the effects of diuretics (249). The risk for hyperkalemia is also increased in patients taking NSAIDs and angiotensin-converting enzyme inhibitors (162). The mechanism underlying these interactions is related to the inhibition of renal PG production by NSAIDs.
NSAIDs may increase serum levels of methotrexate by interfering with the function of a renal tubular anion transporter (250). The interaction of NSAIDs with methotrexate is most important in patients receiving high-dose methotrexate as chemotherapy, in which severe hematologic and gastrointestinal toxicity has been reported. Combinations of low-dose methotrexate and NSAIDs are routinely used in rheumatology practice. The systemic clearance of methotrexate was shown to be significantly reduced by NSAIDs in some studies (251,252), but not in others (253). All studies demonstrated significant variability between individuals in the effect of NSAIDs on methotrexate clearance. Patients taking methotrexate and NSAIDs should be carefully monitored for signs of methotrexate toxicity, especially in the setting of renal insufficiency.
NSAIDs interfere with excretion of lithium, which may lead to increased serum lithium concentrations (254). The effects occur within 5 to 7 days of initiating or discontinuing NSAIDs. Usually, the effect is not clinically important, but lithium toxicity in the setting of NSAID use has been reported (255). Patients should be monitored for lithium toxicity or decreased effectiveness of lithium when NSAIDs are initiated or withdrawn, respectively.
Patients taking NSAIDs may also frequently be taking low-dose aspirin for its antiplatelet effects. Because NSAIDs and aspirin bind to similar regions of COX-1, Catella-Lawson and colleagues hypothesized that certain NSAIDs may potentially inhibit the antiplatelet effect of aspirin (152). Their studies confirmed that ibuprofen administered prior to aspirin inhibited the antiplatelet effect of aspirin. Diclofenac, acetaminophen, and the COX-2 inhibitor rofecoxib did not inhibit the antiplatelet effect of aspirin. It is not yet known if ibuprofen negates the clinical
P.697

benefits of daily low-dose aspirin, but use of this drug in patients taking low-dose aspirin for cardioprotection is not advised.
Other important drug interactions involving NSAIDs relate to the additive effects of shared toxicities. For instance, both NSAIDs and oral bisphosphonates are associated with increased risk for gastric ulcers. One study demonstrated that women taking alendronate and naproxen were more likely to develop endoscopically-detected gastric ulcers than women taking either agent alone (256). Similarly, combined use of NSAIDs with agents that impair renal function may lead to additive toxicity. An example is the increased risk for elevated serum creatinine and potassium in patients taking cyclosporine A and NSAIDs (257,258). Caution should be observed when combining NSAIDs with other nephrotoxic agents such as aminoglycosides, ganciclovir, cisplatin, and amphotericin B. Finally, NSAIDs can displace warfarin from plasma proteins, causing prolongation of the prothrombin time. The combination of impaired platelet function and gastrointestinal toxicity associated with nonselective NSAIDs and anticoagulation with warfarin is thought to lead to a significant increase in the incidence of bleeding gastroduodenal ulcers (79). With the availability of selective COX-2 inhibitors that have reduced ulcer risk and no effect on platelet function, aspirin and nonselective NSAIDs should be avoided in patients taking warfarin.
Inhibition of enzymes important in NSAID metabolism may lead to impaired clearance. Many NSAIDs (ibuprofen, celecoxib, diclofenac, flurbiprofen, meloxicam, and indomethacin) are metabolized by the 2C9 isozyme of cytochrome P450 (259,260,261,262). Drugs that inhibit 2C9 function may cause reduced rates of NSAID metabolism. Examples include fluconazole, metabolites of leflunomide, and voriconazole. The clinical significance of these interactions has not been studied. Many of the cytochrome P450 isozyme genes are polymorphic, and the importance of these polymorphisms with regard to drug toxicity is an area of intense research in the field of pharmacogenetics.
Salicylates may interact with other drugs as a result of high-protein binding (247). Warfarin is displaced from plasma proteins by salicylates. Other drugs that may be displaced from plasma proteins by salicylates are sulfonylureas, valproic acid, phenytoin, sulfonamides, and penicillins. The clinical significance of these interactions is usually related to high doses of salicylates.
PRACTICAL CONSIDERATIONS IN THE DOSING AND ADMINISTRATION OF NSAIDS
Drug Selection and Toxicity Monitoring
Among large groups of patients, most NSAIDs show similar levels of efficacy in disorders such as OA or RA. Therefore, considerations in the selection of NSAIDs may be focused on issues such as avoidance of adverse events. Therapy should be initiated with low doses that are titrated upward as needed, especially in the elderly and those with risk factors for adverse events such as hypertension and heart disease. Individual variability in clinical responsiveness to NSAIDs may be marked. Therefore, if the desired clinical response has not been achieved after 2 to 4 weeks of therapy, it is reasonable to switch to another NSAID. Some patients may try three to four NSAIDs before identifying a treatment that is effective and well tolerated.
A plan for monitoring potential drug toxicity should be implemented for each patient starting an NSAID. Patients should be advised to seek medical attention if persistent symptoms of dyspepsia or edema develop. Hypertensive patients should monitor their blood pressure frequently during the first several weeks of therapy and advised to report even modest increases. The elderly and patients with a history of heart disease or liver disease should be reevaluated after 2 to 6 weeks of therapy. Appropriate monitoring may include measurement of blood pressure, assessment of edema, and laboratory evaluation for signs of impaired renal function or hyperkalemia. Guidelines from the American College of Rheumatology recommend monitoring complete blood count, and possibly serum creatinine, aspartate aminotransferase (AST), and alanine aminotransferase (ALT) at least yearly (263). More frequent laboratory monitoring may be indicated in the elderly. Patients taking concomitant angiotensin-converting enzyme inhibitors or diuretics are recommended to have serum creatinine monitored weekly for 3 weeks upon initiating NSAID therapy.
Use in Children
NSAIDs are generally well tolerated in children; however, a limited number of systematic studies of efficacy and pharmacokinetics have been performed (264). Prolonged clearance of NSAIDs may be observed in neonates as a result of relatively lower glomerular filtration rates and immature hepatic cytochrome P450 system (265). The weight-adjusted elimination half-lives of NSAIDs in children are similar to those on adults (266). The gastrointestinal and renal safety profile of NSAIDs may be superior to that of adults, based on a small number of studies (264,265). However, severe and irreversible renal failure has been reported in neonates who received indomethacin to induce closure of a patent ductus arteriosus (267). Preliminary results of clinical trials using COX-2 inhibitors in children indicate that these drugs are well tolerated, although celecoxib was more rapidly cleared in children compared with adults (268,269).
Fertility, Pregnancy, and Lactation
PGs are involved in the implantation of embryos into the endometrium, and animal studies have shown that some
P.698

NSAIDs may reduce the number of successfully implanted embryos. However, there are no data to indicate that NSAIDs affect fertility in humans. NSAIDs cross the placenta, and tissue levels in the fetus are comparable with those in the mother. Limited studies with aspirin, naproxen, and ibuprofen do not indicate that these NSAIDs have teratogenic effects (270,271,272). Use of NSAIDs during the third trimester of pregnancy has been associated with premature closure of the ductus arteriosus or neonatal pulmonary hypertension (273). NSAIDs may also be associated with oligohydramnios, renal insufficiency in the fetus, maternal or fetal hemorrhage, and delayed onset and duration of labor (139,273). In general, NSAIDs are not recommended during pregnancy. An exception to this is the use of low-dose aspirin in the management of pregnant women with antiphospholipid antibody syndrome, a history of placental insufficiency, or risk factors for preeclampsia (273). NSAIDs are excreted into breast milk in many cases. Accordingly, use of NSAIDs in nursing mothers is not recommended.
FUTURE DIRECTIONS
The history of NSAIDs continues to be written with advances that improve the efficacy and safety of this important drug class. During the next 2 to 3 years several new COX-2 inhibitors will be introduced. Ongoing safety studies and cost-benefit analyses will better establish the most appropriate use of COX-2 inhibitors and nonselective NSAIDs (274). An emerging class of nitric oxide-releasing NSAIDs may offer a new strategy for added drug safety (275,276). Finally, it is hoped that the emerging field of pharmacogenetics will provide the tools to understand the mechanism behind the marked individual variability in both the clinical responses and adverse effects associated with NSAIDs.
REFERENCES
1. Jones R. Nonsteroidal anti-inflammatory drug prescribing: past, present, and future. Am J Med 2001;110(suppl 1A):4–7.
2. Leroux H. J Chim Med 1830;6:341.
3. Sneader W. The discovery of aspirin: a reappraisal. BMJ 2000;321 (7276):1591–1594.
4. Vieson K. Overview: Nonsteroidal antiinflammatory drugs. In: Clinical pharmacology. Vol. 2003. Gold Standard Multimedia, 2000.
5. FitzGerald GA, Patrono C. The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med 2001;345:433–442.
6. Simon L, Lipman A, Jacox A, et al. Guideline for the management of arthritis pain in osteoarthritis, rheumatoid arthritis, and juvenile chronic arthritis. Glenview, IL: American Pain Society, 2002 (APS Clinical Practice Guidelines Series, No. 2).
7. American College of Rheumatology Subcommittee on Osteoarthritis Guidelines. Recommendations for the medical management of osteoarthritis of the hip and knee: 2000 update. Arthritis Rheum 2000; 43:1905–1915.
8. The management of persistent pain in older persons. J Am Geriatr Soc 2002;50(suppl):205–224.
9. Vane JR. Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nat New Biol 1971;231:232–235.
10. Pash JM, Bailey JM. Inhibition by corticosteroids of epidermal growth factor-induced recovery of cyclooxygenase after aspirin inactivation. FASEB J 1988;2:2613–2618.
11. Sebaldt RJ, Sheller JR, Oates JA, et al. Inhibition of eicosanoid biosynthesis by glucocorticoids in humans. Proc Natl Acad Sci U S A 1990;87:6974–6978.
12. Masferrer JL, Zweifel BS, Seibert K, et al. Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J Clin Invest 1990;86:1375–1379.
13. Xie WL, Chipman JG, Robertson DL, et al. Expression of a mitogen-responsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci U S A 1991;88:2692–2696.
14. Kujubu DA, Herschman HR. Dexamethasone inhibits mitogen induction of the TIS10 prostaglandin synthase/cyclooxygenase gene. J Biol Chem 1992;267:7991–7994.
15. O’Banion MK, Winn VD, Young DA. cDNA cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc Natl Acad Sci U S A 1992;89:4888–4892.
16. Toh H, Yokoyama C, Tanabe T, et al. Molecular evolution of cyclooxygenase and lipoxygenase. Prostaglandins 1992;44:291–315.
17. Morita I, Schindler M, Regier MK, et al. Different intracellular locations for prostaglandin endoperoxide H synthase-1 and -2. J Biol Chem 1995;270:10902–10908.
18. Smith WL. Prostanoid biosynthesis and mechanisms of action. Am J Physiol 1992;263(2 part 2):F181–F191.
19. Komhoff M, Grone HJ, Klein T, et al. Localization of cyclooxygenase-1 and -2 in adult and fetal human kidney: implication for renal function. Am J Physiol 1997;272(4 part 2):F460–F468.
20. Harris RC, McKanna JA, Akai Y, et al. Cyclooxygenase-2 is associated with the macula densa of rat kidney and increases with salt restriction. J Clin Invest 1994;94:2504–2510.
21. Rocca B, Spain LM, Pure E, et al. Distinct roles of prostaglandin H synthases 1 and 2 in T-cell development. J Clin Invest 1999;103:1469–1477.
22. Smith CJ, Morrow JD, Roberts LJ 2nd, et al. Induction of prostaglandin endoperoxide synthase-1 (COX-1) in a human promonocytic cell line by treatment with the differentiating agent TPA. Adv Exp Med Biol 1997;400A:99–106.
23. Loll PJ, Picot D, Garavito RM. The structural basis of aspirin activity inferred from the crystal structure of inactivated prostaglandin H2 synthase. Nat Struct Biol 1995;2:637–643.
24. Roth GJ, Majerus PW. The mechanism of the effect of aspirin on human platelets. I. Acetylation of a particulate fraction protein. J Clin Invest 1975;56:624–632.
25. Wu KK. Biochemical pharmacology of nonsteroidal anti-inflammatory drugs. Biochem Pharmacol 1998;55:543–547.
26. Wu KK, Sanduja R, Tsai AL, et al. Aspirin inhibits interleukin 1-induced prostaglandin H synthase expression in cultured endothelial cells. Proc Natl Acad Sci U S A 1991;88:2384–2387.
27. Weissmann G. Prostaglandins as modulators rather than mediators of inflammation. J Lipid Mediat 1993;6:275–286.
28. Kopp E, Ghosh S. Inhibition of NF-kappa B by sodium salicylate and aspirin. Science 1994;265:956–959.
29. Gautam SC, Pindolia KR, Noth CJ, et al. Chemokine gene expression in bone marrow stromal cells: downregulation with sodium salicylate. Blood 1995;86:2541–2550.
30. Farivar RS, Brecher P. Salicylate is a transcriptional inhibitor of the inducible nitric oxide synthase in cultured cardiac fibroblasts. J Biol Chem 1996;271:31585–31592.
31. Schwenger P, Skolnik EY, Vilcek J. Inhibition of tumor necrosis factor-induced p42/p44 mitogen-activated protein kinase activation by sodium salicylate. J Biol Chem 1996;271:8089–8094.
32. Jones MK, Wang H, Peskar BM, et al. Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: insight into mechanisms and implications for cancer growth and ulcer healing. Nat Med 1999;5:1418–1423.
33. Lecomte M, Laneuville O, Ji C, et al. Acetylation of human prostaglandin endoperoxide synthase-2 (cyclooxygenase-2) by aspirin. J Biol Chem 1994;269:13207–13215.
34. Claria J, Serhan CN. Aspirin triggers previously undescribed bioactive eicosanoids by human endothelial cell-leukocyte interactions. Proc Natl Acad Sci U S A 1995;92:9475–9479.
35. Serhan CN, Takano T, Maddox JF. Aspirin-triggered 15-epi-lipoxin A4 and stable analogs on lipoxin A4 are potent inhibitors of acute inflammation. Receptors and pathways. Adv Exp Med Biol 1999;447: 133–149.
P.699

36. Ito S, Okuda-Ashitaka E, Minami T. Central and peripheral roles of prostaglandins in pain and their interactions with novel neuropeptides nociceptin and nocistatin. Neurosci Res 2001;41:299–332.
37. Seibert K, Zhang Y, Leahy K, et al. Pharmacological and biochemical demonstration of the role of cyclooxygenase 2 in inflammation and pain. Proc Natl Acad Sci U S A 1994;91:12013–12017.
38. Yamagata K, Andreasson KI, Kaufmann WE, et al. Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron 1993;11:371–386.
39. Breder CD, Smith WL, Raz A, et al. Distribution and characterization of cyclooxygenase immunoreactivity in the ovine brain. J Comp Neurol 1992;322:409–438.
40. Breder CD, Saper CB. Expression of inducible cyclooxygenase mRNA in the mouse brain after systemic administration of bacterial lipopolysaccharide. Brain Res 1996;713:64–69.
41. Beiche F, Scheuerer S, Brune K, et al. Up-regulation of cyclooxygenase-2 mRNA in the rat spinal cord following peripheral inflammation. FEBS Lett 1996;390:165–169.
42. Beiche F, Klein T, Nusing R, et al. Localization of cyclooxygenase-2 and prostaglandin E2 receptor EP3 in the rat lumbar spinal cord. J Neuroimmunol 1998;89:26–34.
43. Chopra B, Giblett S, Little JG, et al. Cyclooxygenase-1 is a marker for a subpopulation of putative nociceptive neurons in rat dorsal root ganglia. Eur J Neurosci 2000;12:911–920.
44. Engblom D, Ek M, Hallbeck M, et al. Distribution of prostaglandin EP(3) and EP(4) receptor mRNA in the rat parabrachial nucleus. Neurosci Lett 2000;281:163–166.
45. Nakamura K, Kaneko T, Yamashita Y, et al. Immunohistochemical localization of prostaglandin EP3 receptor in the rat nervous system. J Comp Neurol 2000;421:543–569.
46. Ek M, Arias C, Sawchenko P, et al. Distribution of the EP3 prostaglandin E(2) receptor subtype in the rat brain: relationship to sites of interleukin-1-induced cellular responsiveness. J Comp Neurol 2000; 428:5–20.
47. Coceani F, Akarsu ES. Prostaglandin E2 in the pathogenesis of fever. An update. Ann NY Acad Sci 1998;856:76–82.
48. Ushikubi F, Segi E, Sugimoto Y, et al. Impaired febrile response in mice lacking the prostaglandin E receptor subtype EP3. Nature 1998;395:281–284.
49. Li S, Wang Y, Matsumura K, et al. The febrile response to lipopolysaccharide is blocked in cyclooxygenase-2(2/2), but not in cyclooxygenase-1(2/2) mice. Brain Res 1999;825:86–94.
50. Schwartz JI, Chan CC, Mukhopadhyay S, et al. Cyclooxygenase-2 inhibition by rofecoxib reverses naturally occurring fever in humans. Clin Pharmacol Ther 1999;65:653–660.
51. Chandrasekharan NV, Dai H, Roos KL, et al. COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesic/ antipyretic drugs: cloning, structure, and expression. Proc Natl Acad Sci U S A 2002;99:13926–13931.
52. Awtry EH, Loscalzo J. Aspirin. Circulation 2000;101:1206–1218.
53. Owen PR. Prostaglandin synthetase inhibitors in the treatment of primary dysmenorrhea. Outcome trials reviewed. Am J Obstet Gynecol 1984;148:96–103.
54. Morrison BW, Daniels SE, Kotey P, et al. Rofecoxib, a specific cyclooxygenase-2 inhibitor, in primary dysmenorrhea: a randomized controlled trial. Obstet Gynecol 1999;94:504–508.
55. Daniels SE, Talwalker S, Torri S, et al. Valdecoxib, a cyclooxygenase-2-specific inhibitor, is effective in treating primary dysmenorrhea. Obstet Gynecol 2002;100:350–358.
56. Subbaramaiah K, Dannenberg AJ. Cyclooxygenase 2: a molecular target for cancer prevention and treatment. Trends Pharmacol Sci 2003;24:96–102.
57. Aisen PS. Evaluation of selective COX-2 inhibitors for the treatment of Alzheimer’s disease. J Pain Symptom Manage 2002;23(suppl):35–40.
58. Waddell WR, Loughry RW. Sulindac for polyposis of the colon. J Surg Oncol 1983;24:83–87.
59. Steinbach G, Lynch PM, Phillips RK, et al. The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–1952.
60. Giovannucci E, Egan KM, Hunter DJ, et al. Aspirin and the risk of colorectal cancer in women. N Engl J Med 1995;333:609–614.
61. Giovannucci E, Rimm EB, Stampfer MJ, et al. Aspirin use and the risk for colorectal cancer and adenoma in male health professionals. Ann Intern Med 1994;121:241–246.
62. Greenberg ER, Baron JA, Freeman DH Jr, et al. Reduced risk of large-bowel adenomas among aspirin users. The Polyp Prevention Study Group. J Natl Cancer Inst 1993;85:912–916.
63. Peleg, II, Maibach HT, Brown SH, et al. Aspirin and nonsteroidal anti-inflammatory drug use and the risk of subsequent colorectal cancer. Arch Intern Med 1994;154:394–399.
64. Thun MJ, Namboodiri MM, Calle EE, et al. Aspirin use and risk of fatal cancer. Cancer Res 1993;53:1322–1327.
65. Thun MJ, Namboodiri MM, Heath CW Jr. Aspirin use and reduced risk of fatal colon cancer. N Engl J Med 1991;325:1593–1596.
66. Collet JP, Sharpe C, Belzile E, et al. Colorectal cancer prevention by non-steroidal anti-inflammatory drugs: effects of dosage and timing. Br J Cancer 1999;81:62–68.
67. Logan RF, Little J, Hawtin PG, et al. Effect of aspirin and non-steroidal anti-inflammatory drugs on colorectal adenomas: case-control study of subjects participating in the Nottingham faecal occult blood screening programme. BMJ 1993;307:285–289.
68. Muscat JE, Stellman SD, Wynder EL. Nonsteroidal antiinflammatory drugs and colorectal cancer. Cancer 1994;74:1847–1854.
69. Peleg, II, Lubin MF, Cotsonis GA, et al. Long-term use of nonsteroidal antiinflammatory drugs and other chemopreventors and risk of subsequent colorectal neoplasia. Dig Dis Sci 1996;41:1319–1326.
70. Reeves MJ, Newcomb PA, Trentham-Dietz A, et al. Nonsteroidal anti-inflammatory drug use and protection against colorectal cancer in women. Cancer Epidemiol Biomarkers Prev 1996;5:955–960.
71. Smalley W, Ray WA, Daugherty J, et al. Use of nonsteroidal anti-inflammatory drugs and incidence of colorectal cancer: a population-based study. Arch Intern Med 1999;159:161–166.
72. Sandler RS, Halabi S, Baron JA, et al. A randomized trial of aspirin to prevent colorectal adenomas in patients with previous colorectal cancer. N Engl J Med 2003;348:883–890.
73. Baron JA, Cole BF, Sandler RS, et al. A randomized trial of aspirin to prevent colorectal adenomas. N Engl J Med 2003;348:891–899.
74. Imperiale TF. Aspirin and the prevention of colorectal cancer. N Engl J Med 2003;348:879–880.
75. McGeer PL, McGeer EG. The inflammatory response system of brain: implications for therapy of Alzheimer and other neurodegenerative diseases. Brain Res Brain Res Rev 1995;21:195–218.
76. Breitner JC. Inflammatory processes and antiinflammatory drugs in Alzheimer’s disease: a current appraisal. Neurobiol Aging 1996;17: 789–794.
77. in t’ Veld BA, Ruitenberg A, Hofman A, et al. Nonsteroidal antiinflammatory drugs and the risk of Alzheimer’s disease. N Engl J Med 2001; 345:1515–1521.
78. Aisen PS, Schafer KA, Grundman M, et al. Effects of rofecoxib or naproxen vs placebo on Alzheimer disease progression: a randomized controlled trial. JAMA 2003;289:2819–2826.
79. Wolfe MM, Lichtenstein DR, Singh G. Gastrointestinal toxicity of nonsteroidal antiinflammatory drugs. N Engl J Med 1999;340:1888–1899.
80. Tamblyn R, Berkson L, Dauphinee WD, et al. Unnecessary prescribing of NSAIDs and the management of NSAID-related gastropathy in medical practice. Ann Intern Med 1997;127:429–438.
81. Douthwaite AH, Lintott GAM. Gastroscopic observation of effect of aspirin and certain other substances on stomach. Lancet 1938;2:1222–1225.
82. Bjorkman D. Commentary: Gastrointestinal safety of coxibs and outcomes studies: what’s the verdict? J Pain Symptom Manage 2002;23 (suppl):11–14.
83. Sun DC, Roth SH, Mitchell CS, et al. Upper gastrointestinal disease in rheumatoid arthritis. Am J Dig Dis 1974;19:405–410.
84. Levy M. Aspirin use in patients with major upper gastrointestinal bleeding and peptic-ulcer disease. A report from the Boston Collaborative Drug Surveillance Program, Boston University Medical Center. N Engl J Med 1974;290:1158–1162.
85. Silvoso GR, Ivey KJ, Butt JH, et al. Incidence of gastric lesions in patients with rheumatic disease on chronic aspirin therapy. Ann Intern Med 1979;91:517–520.
86. Singh G, Triadafilopoulos G. Epidemiology of NSAID induced gastrointestinal complications. J Rheumatol 1999;26(suppl 56):18–24.
P.700

87. Silverstein FE, Graham DY, Senior JR, et al. Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti-inflammatory drugs. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 1995; 123:241–249.
88. Weil J, Colin-Jones D, Langman M, et al. Prophylactic aspirin and risk of peptic ulcer bleeding. BMJ 1995;310:827–830.
89. Schoen RT, Vender RJ. Mechanisms of nonsteroidal anti-inflammatory drug-induced gastric damage. Am J Med 1989;86:449–458.
90. Armstrong CP, Blower AL. Non-steroidal anti-inflammatory drugs and life threatening complications of peptic ulceration. Gut 1987;28: 527–532.
91. Singh G, Ramey DR, Morfeld D, et al. Gastrointestinal tract complications of nonsteroidal anti-inflammatory drug treatment in rheumatoid arthritis. A prospective observational cohort study. Arch Intern Med 1996;156:1530–1536.
92. Hernandez-Diaz S, Rodriguez LA. Association between nonsteroidal anti-inflammatory drugs and upper gastrointestinal tract bleeding/ perforation: an overview of epidemiologic studies published in the 1990s. Arch Intern Med 2000;160:2093–2099.
93. MacDonald TM. Epidemiology and pharmacoeconomic implications of non-steroidal anti-inflammatory drug-associated gastrointestinal toxicity. Rheumatology (Oxford) 2000;39(suppl 2):13–20; discussion 57–59.
94. Rostom A, Dube C, Wells G, et al. Prevention of NSAID-induced gastroduodenal ulcers. Cochrane Database Syst Rev 2002: CD002296.
95. Chan FK, Chung SC, Suen BY, et al. Preventing recurrent upper gastrointestinal bleeding in patients with Helicobacter pylori infection who are taking low-dose aspirin or naproxen. N Engl J Med 2001;344:967–973.
96. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med 2000; 343:1520–1528.
97. Silverstein FE, Faich G, Goldstein JL, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: A randomized controlled trial. Celecoxib Long-term Arthritis Safety Study. JAMA 2000;284:1247–1255.
98. Laine L. Review article: The effect of Helicobacter pylori infection on nonsteroidal anti-inflammatory drug-induced upper gastrointestinal tract injury. Aliment Pharmacol Ther 2002;16(suppl 1):34–39.
99. Laine L, Cominelli F, Sloane R, et al. Interaction of NSAIDs and Helicobacter pylori on gastrointestinal injury and prostaglandin production: a controlled double-blind trial. Aliment Pharmacol Ther 1995;9:127–135.
100. Thillainayagam AV, Tabaqchali S, Warrington SJ, et al. Interrelationships between Helicobacter pylori infection, nonsteroidal antiinflammatory drugs and gastroduodenal disease. A prospective study in healthy volunteers. Dig Dis Sci 1994;39:1085–1089.
101. Kim JG, Graham DY. Helicobacter pylori infection and development of gastric or duodenal ulcer in arthritic patients receiving chronic NSAID therapy. The Misoprostol Study Group. Am J Gastroenterol 1994;89:203–207.
102. Goggin PM, Collins DA, Jazrawi RP, et al. Prevalence of Helicobacter pylori infection and its effect on symptoms and non-steroidal anti-inflammatory drug induced gastrointestinal damage in patients with rheumatoid arthritis. Gut 1993;34:1677–1680.
103. Lanza FL, Evans DG, Graham DY. Effect of Helicobacter pylori infection on the severity of gastroduodenal mucosal injury after the acute administration of naproxen or aspirin to normal volunteers. Am J Gastroenterol 1991;86:735–737.
104. Bannwarth B, Dorval E, Caekaert A, et al. Influence of Helicobacter pylori eradication therapy on the occurrence of gastrointestinal events in patients treated with conventional nonsteroidal antiinflammatory drugs combined with omeprazole. J Rheumatol 2002;29: 1975–1980.
105. Taha AS, Dahill S, Morran C, et al. Neutrophils, Helicobacter pylori, and nonsteroidal anti-inflammatory drug ulcers. Gastroenterology 1999;116:254–258.
106. Chan FK, Sung JJ, Chung SC, et al. Randomised trial of eradication of Helicobacter pylori before non-steroidal anti-inflammatory drug therapy to prevent peptic ulcers. Lancet 1997;350:975–979.
107. Labenz J, Blum AL, Bolten WW, et al. Primary prevention of diclofenac associated ulcers and dyspepsia by omeprazole or triple therapy in Helicobacter pylori positive patients: a randomised, double blind, placebo controlled, clinical trial. Gut 2002;51:329–335.
108. Yeomans ND, Tulassay Z, Juhasz L, et al. A comparison of omeprazole with ranitidine for ulcers associated with nonsteroidal antiinflammatory drugs. Acid Suppression Trial: Ranitidine versus Omeprazole for NSAID-associated Ulcer Treatment (ASTRONAUT) Study Group. N Engl J Med 1998;338:719–726.
109. Hawkey CJ, Karrasch JA, Szczepanski L, et al. Omeprazole compared with misoprostol for ulcers associated with nonsteroidal antiinflammatory drugs. Omeprazole versus Misoprostol for NSAID-induced Ulcer Management (OMNIUM) Study Group. N Engl J Med 1998;338: 727–734.
110. Hawkey CJ, Tulassay Z, Szczepanski L, et al. Randomised controlled trial of Helicobacter pylori eradication in patients on non-steroidal anti-inflammatory drugs: HELP NSAIDs study. Helicobacter Eradication for Lesion Prevention. Lancet 1998;352:1016–1021.
111. Agrawal NM, Campbell DR, Safdi MA, et al. Superiority of lansoprazole vs ranitidine in healing nonsteroidal anti-inflammatory drug-associated gastric ulcers: results of a double-blind, randomized, multicenter study. NSAID-Associated Gastric Ulcer Study Group. Arch Intern Med 2000;160:1455–1461.
112. Chan FK, Sung JJ, Suen R, et al. Does eradication of Helicobacter pylori impair healing of nonsteroidal anti-inflammatory drug associated bleeding peptic ulcers? A prospective randomized study. Aliment Pharmacol Ther 1998;12:1201–1205.
113. Talley NJ, Zinsmeister AR, Schleck CD, et al. Smoking, alcohol, and analgesics in dyspepsia and among dyspepsia subgroups: lack of an association in a community. Gut 1994;35:619–624.
114. Talley NJ, Evans JM, Fleming KC, et al. Nonsteroidal antiinflammatory drugs and dyspepsia in the elderly. Dig Dis Sci 1995;40: 1345–1350.
115. Watson DJ, Harper SE, Zhao PL, et al. Gastrointestinal tolerability of the selective cyclooxygenase-2 (COX-2) inhibitor rofecoxib compared with nonselective COX-1 and COX-2 inhibitors in osteoarthritis. Arch Intern Med 2000;160:2998–3003.
116. Larkai EN, Smith JL, Lidsky MD, et al. Gastroduodenal mucosa and dyspeptic symptoms in arthritic patients during chronic nonsteroidal anti-inflammatory drug use. Am J Gastroenterol 1987;82: 1153–1158.
117. Pounder R. Silent peptic ulceration: deadly silence or golden silence? Gastroenterology 1989;96(part 2; suppl):626–631.
118. Bijlsma JW. Treatment of NSAID-induced gastrointestinal lesions with cimetidine: an international multicentre collaborative study. Aliment Pharmacol Ther 1988;2(suppl 1):85–95.
119. Lanza FL, Aspinall RL, Swabb EA, et al. Double-blind, placebo-controlled endoscopic comparison of the mucosal protective effects of misoprostol versus cimetidine on tolmetin-induced mucosal injury to the stomach and duodenum. Gastroenterology 1988;95:289–294.
120. Saunders JH, Oliver RJ, Higson DL. Dyspepsia: incidence of a non-ulcer disease in a controlled trial of ranitidine in general practice. BMJ (Clin Res) 1986;292:665–668.
121. Van Groenendael JH, Markusse HM, Dijkmans BA, et al. The effect of ranitidine on NSAID related dyspeptic symptoms with and without peptic ulcer disease of patients with rheumatoid arthritis and osteoarthritis. Clin Rheumatol 1996;15:450–456.
122. Matsuhashi N, Yamada A, Hiraishi M, et al. Multiple strictures of the small intestine after long-term nonsteroidal anti-inflammatory drug therapy. Am J Gastroenterol 1992;87:1183–1186.
123. Hershfield NB. Endoscopic demonstration of non-steroidal anti-inflammatory drug-induced small intestinal strictures. Gastrointest Endosc 1992;38:388–390.
124. Saw KC, Quick CR, Higgins AF. Ileocaecal perforation and bleeding—are non-steroidal anti-inflammatory drugs (NSAIDs) responsible? [comment]. J R Soc Med 1990;83:114–115.
125. Lang J, Price AB, Levi AJ, et al. Diaphragm disease: pathology of disease of the small intestine induced by non-steroidal anti-inflammatory drugs. J Clin Pathol 1988;41:516–526.
126. Bjarnason I, Price AB, Zanelli G, et al. Clinicopathological features of nonsteroidal antiinflammatory drug-induced small intestinal strictures. Gastroenterology 1988;94:1070–1074.
127. Sukumar L. Recurrent small bowel obstruction associated with piroxicam. Br J Surg 1987;74:186.
P.701

128. Fang WF, Broughton A, Jacobson ED. Indomethacin-induced intestinal inflammation. Am J Dig Dis 1977;22:749–760.
129. Sturges HF, Krone CL. Ulceration and stricture of the jejunum in a patient on long-term indomethacin therapy. Am J Gastroenterol 1973;59:162–169.
130. Bjorkman D. Nonsteroidal anti-inflammatory drug-associated toxicity of the liver, lower gastrointestinal tract, and esophagus. Am J Med 1998;105(suppl):17–21.
131. Kessler WF, Shires GT 3rd, Fahey TJ 3rd. Surgical complications of nonsteroidal antiinflammatory drug-induced small bowel ulceration. J Am Coll Surg 1997;185:250–254.
132. Laine L, Connors LG, Reicin A, et al. Serious lower gastrointestinal clinical events with nonselective NSAID or coxib use. Gastroenterology 2003;124:288–292.
133. Tolman KG. Hepatotoxicity of non-narcotic analgesics. Am J Med 1998;105(suppl):13–19.
134. Fry SW, Seeff LB. Hepatotoxicity of analgesics and anti-inflammatory agents. Gastroenterol Clin North Am 1995;24:875–905.
135. Carson JL, Strom BL, Duff A, et al. Safety of nonsteroidal anti-inflammatory drugs with respect to acute liver disease. Arch Intern Med 1993;153:1331–1336.
136. Tarazi EM, Harter JG, Zimmerman HJ, et al. Sulindac-associated hepatic injury: analysis of 91 cases reported to the Food and Drug Administration. Gastroenterology 1993;104:569–574.
137. Whelton A, Hamilton CW. Nonsteroidal anti-inflammatory drugs: effects on kidney function. J Clin Pharmacol 1991;31:588–598.
138. Harris RC Jr. Cyclooxygenase-2 inhibition and renal physiology. Am J Cardiol 2002;89:10D–17D.
139. Henrich WL, Agodoa LE, Barrett B, et al. Analgesics and the kidney: summary and recommendations to the Scientific Advisory Board of the National Kidney Foundation from an Ad Hoc Committee of the National Kidney Foundation. Am J Kidney Dis 1996;27: 162–165.
140. Whelton A. Nephrotoxicity of nonsteroidal anti-inflammatory drugs: physiologic foundations and clinical implications. Am J Med 1999; 106(suppl):13–24.
141. Swan SK, Rudy DW, Lasseter KC, et al. Effect of cyclooxygenase-2 inhibition on renal function in elderly persons receiving a low-salt diet. A randomized, controlled trial. Ann Intern Med 2000;133: 1–9.
142. Smith WL, Bell TG. Immunohistochemical localization of the prostaglandin-forming cyclooxygenase in renal cortex. Am J Physiol 1978;235:F451–F457.
143. Yang T, Singh I, Pham H, et al. Regulation of cyclooxygenase expression in the kidney by dietary salt intake. Am J Physiol 1998; 274:F481–F489.
144. Ichihara A, Imig JD, Inscho EW, et al. Cyclooxygenase-2 participates in tubular flow-dependent afferent arteriolar tone: interaction with neuronal NOS. Am J Physiol 1998;275:F605–F612.
145. Harding P, Sigmon DH, Alfie ME, et al. Cyclooxygenase-2 mediates increased renal renin content induced by low-sodium diet. Hypertension 1997;29:297–302.
146. Komhoff M, Jeck ND, Seyberth HW, et al. Cyclooxygenase-2 expression is associated with the renal macula densa of patients with Bartter-like syndrome. Kidney Int 2000;58:2420–2424.
147. Whelton A, Maurath CJ, Verburg KM, et al. Renal safety and tolerability of celecoxib, a novel cyclooxygenase-2 inhibitor. Am J Ther 2000;7:159–175.
148. Gertz BJ, Krupa D, Bolognese JA, et al. A comparison of adverse renovascular experiences among osteoarthritis patients treated with rofecoxib and comparator non-selective non-steroidal anti-inflammatory agents. Curr Med Res Opin 2002;18:82–91.
149. Brater DC. Effects of nonsteroidal anti-inflammatory drugs on renal function: focus on cyclooxygenase-2-selective inhibition. Am J Med 1999;107(suppl):65–70; discussion 70–71.
150. Whelton A. Renal and related cardiovascular effects of conventional and COX-2-specific NSAIDs and non-NSAID analgesics. Am J Ther 2000;7:63–74.
151. Schwartz JI, Vandormael K, Malice MP, et al. Comparison of rofecoxib, celecoxib, and naproxen on renal function in elderly subjects receiving a normal-salt diet. Clin Pharmacol Ther 2002;72: 50–61.
152. Catella-Lawson F, McAdam B, Morrison BW, et al. Effects of specific inhibition of cyclooxygenase-2 on sodium balance, hemodynamics, and vasoactive eicosanoids. J Pharmacol Exp Ther 1999; 289:735–741.
153. Epstein M. Renal prostaglandins and the control of renal function in liver disease. Am J Med 1986;80:46–55.
154. Oates JA, FitzGerald GA, Branch RA, et al. Clinical implications of prostaglandin and thromboxane A2 formation (1). N Engl J Med 1988;319:689–698.
155. Walshe JJ, Venuto RC. Acute oliguric renal failure induced by indomethacin: possible mechanism. Ann Intern Med 1979;91: 47–49.
156. Arisz L, Donker AJ, Brentjens JR, et al. The effect of indomethacin on proteinuria and kidney function in the nephrotic syndrome. Acta Med Scand 1976;199:121–125.
157. Kleinknecht C, Broyer M, Gubler MC, et al. Irreversible renal failure after indomethacin in steroid-resistant nephrosis. N Engl J Med 1980;302:691.
158. Blackshear JL, Napier JS, Davidman M, et al. Renal complications of nonsteroidal antiinflammatory drugs: identification and monitoring of those at risk. Semin Arthritis Rheum 1985;14: 163–175.
159. Favre L, Glasson P, Vallotton MB. Reversible acute renal failure from combined triamterene and indomethacin: a study in healthy subjects. Ann Intern Med 1982;96:317–320.
160. McCarthy JT, Torres VE, Romero JC, et al. Acute intrinsic renal failure induced by indomethacin: role of prostaglandin synthetase inhibition. Mayo Clin Proc 1982;57:289–296.
161. Henry D, Page J, Whyte I, et al. Consumption of non-steroidal anti-inflammatory drugs and the development of functional renal impairment in elderly subjects. Results of a case-control study. Br J Clin Pharmacol 1997;44:85–90.
162. Hay E, Derazon H, Bukish N, et al. Fatal hyperkalemia related to combined therapy with a COX-2 inhibitor, ACE inhibitor and potassium rich diet. J Emerg Med 2002;22:349–352.
163. Jolobe OM. Nephrotoxicity in the elderly due to co-prescription of ACE inhibitors and NSAIDs. J R Soc Med 2001;94:657–658.
164. Akbarpour F, Afrasiabi A, Vaziri ND. Severe hyperkalemia caused by indomethacin and potassium supplementation. South Med J 1985; 78:756–767.
165. Mor R, Pitlik S, Rosenfeld JB. Indomethacin- and moduretic-induced hyperkalemia. Isr J Med Sci 1983;19:535–537.
166. Galler M, Folkert VW, Schlondorff D. Reversible acute renal insufficiency and hyperkalemia following indomethacin therapy. JAMA 1981;246:154–155.
167. Nicholls MG, Espiner EA. Indomethacin-induced azotaemia and hyperkalaemia: a case study. N Z Med J 1981;94:377–379.
168. Findling JW, Beckstrom D, Rawsthorne L, et al. Indomethacin-induced hyperkalemia in three patients with gouty arthritis. JAMA 1980;244:1127–1128.
169. Paladini G, Tonazzi C. Indomethacin-induced hyperkalemia and renal failure in multiple myeloma. Acta Haematol 1982;68:256–260.
170. Kelley M, Bastani B. Ketorolac-induced acute renal failure and hyperkalemia. Clin Nephrol 1995;44:276–277.
171. Patel P, Mandal B, Greenway MW. Hyperkalaemic quadriparesis secondary to chronic diclofenac treatment. Postgrad Med J 2001;77: 50–51.
172. Apostolou T, Sotsiou F, Yfanti G, et al. Acute renal failure induced by nimesulide in a patient suffering from temporal arteritis. Nephrol Dial Transplant 1997;12:1493–1496.
173. Pal B, Hutchinson A, Bhattacharya A, et al. Cardiac arrest due to severe hyperkalaemia in patient taking nabumetone and low salt diet. BMJ 1995;311:1486–1487.
174. Abraham PA, Keane WF. Glomerular and interstitial disease induced by nonsteroidal anti-inflammatory drugs. Am J Nephrol 1984; 4:1–6.
175. Bender WL, Whelton A, Beschorner WE, et al. Interstitial nephritis, proteinuria, and renal failure caused by nonsteroidal anti-inflammatory drugs. Immunologic characterization of the inflammatory infiltrate. Am J Med 1984;76:1006–1012.
176. Levin ML. Patterns of tubulo-interstitial damage associated with nonsteroidal antiinflammatory drugs. Semin Nephrol 1988;8:55–61.
P.702

177. Husserl FE, Lange RK, Kantrow CM Jr. Renal papillary necrosis and pyelonephritis accompanying fenoprofen therapy. JAMA 1979;242: 1896–1898.
178. Shah GM, Muhalwas KK, Winer RL. Renal papillary necrosis due to ibuprofen. Arthritis Rheum 1981;24:1208–1210.
179. Morales A, Steyn J. Papillary necrosis following phenylbutazone ingestion. Arch Surg 1971;103:420–421.
180. Murray TG, Goldberg M. Analgesic-associated nephropathy in the U.S.A.: epidemiologic, clinical and pathogenetic features. Kidney Int 1978;13:64–71.
181. Nanra RS, Stuart-Taylor J, de Leon AH, et al. Analgesic nephropathy: etiology, clinical syndrome, and clinicopathologic correlations in Australia. Kidney Int 1978;13:79–92.
182. De Broe ME, Elseviers MM. Analgesic nephropathy. N Engl J Med 1998;338:446–452.
183. Elseviers MM, De Broe ME. A long-term prospective controlled study of analgesic abuse in Belgium. Kidney Int 1995;48:1912–1919.
184. Elseviers MM, De Schepper A, Corthouts R, et al. High diagnostic performance of CT scan for analgesic nephropathy in patients with incipient to severe renal failure. Kidney Int 1995;48:1316–1323.
185. Elseviers MM, Waller I, Nenoy D, et al. Evaluation of diagnostic criteria for analgesic nephropathy in patients with end-stage renal failure: results of the ANNE study. Analgesic Nephropathy Network of Europe. Nephrol Dial Transplant 1995;10:808–814.
186. Piper JM, Tonascia J, Matanoski GM. Heavy phenacetin use and bladder cancer in women aged 20 to 49 years. N Engl J Med 1985; 313:292–295.
187. Blohme I, Johansson S. Renal pelvic neoplasms and atypical urothelium in patients with end-stage analgesic nephropathy. Kidney Int 1981;20:671–675.
188. McCredie M, Stewart JH, Carter JJ, et al. Phenacetin and papillary necrosis: independent risk factors for renal pelvic cancer. Kidney Int 1986;30:81–84.
189. Johnson AG, Nguyen TV, Day RO. Do nonsteroidal anti-inflammatory drugs affect blood pressure? A meta-analysis. Ann Intern Med 1994; 121:289–300.
190. Pope JE, Anderson JJ, Felson DT. A meta-analysis of the effects of nonsteroidal anti-inflammatory drugs on blood pressure. Arch Intern Med 1993;153:477–484.
191. Komers R, Anderson S, Epstein M. Renal and cardiovascular effects of selective cyclooxygenase-2 inhibitors. Am J Kidney Dis 2001;38: 1145–1157.
192. Vioxx. Prescribing information, Merck, Inc.
193. Johnson AG, Simons LA, Simons J, et al. Non-steroidal anti-inflammatory drugs and hypertension in the elderly: a community-based cross-sectional study. Br J Clin Pharmacol 1993;35:455–459.
194. Gurwitz JH, Avorn J, Bohn RL, et al. Initiation of antihypertensive treatment during nonsteroidal anti-inflammatory drug therapy. JAMA 1994;272:781–786.
195. Schooley RT, Wagley PF, Lietman PS. Edema associated with ibuprofen therapy. JAMA 1977;237:1716–1717.
196. Tashima CK, Rose M. Letter: Pulmonary edema and salicylates. Ann Intern Med 1974;81:274–275.
197. Nevins M, Berque S, Corwin N, et al. Phenylbutazone and pulmonary oedema. Lancet 1969;2:1358.
198. Van den Ouweland FA, Gribnau FW, Meyboom RH. Congestive heart failure due to nonsteroidal anti-inflammatory drugs in the elderly. Age Ageing 1988;17:8–16.
199. Feenstra J, Grobbee DE, Mosterd A, et al. Adverse cardiovascular effects of NSAIDs in patients with congestive heart failure. Drug Saf 1997;17:166–180.
200. Heerdink ER, Leufkens HG, Herings RM, et al. NSAIDs associated with increased risk of congestive heart failure in elderly patients taking diuretics. Arch Intern Med 1998;158:1108–1112.
201. Page J, Henry D. Consumption of NSAIDs and the development of congestive heart failure in elderly patients: an underrecognized public health problem. Arch Intern Med 2000;160:777–784.
202. Feenstra J, Heerdink ER, Grobbee DE, et al. Association of nonsteroidal anti-inflammatory drugs with first occurrence of heart failure and with relapsing heart failure: the Rotterdam Study. Arch Intern Med 2002;162:265–270.
203. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ 2002;324:71–86.
204. Baigent C, Patrono C. Selective cyclooxygenase 2 inhibitors, aspirin, and cardiovascular disease: a reappraisal. Arthritis Rheum 2003;48: 12–20.
205. Konstam MA, Weir MR, Reicin A, et al. Cardiovascular thrombotic events in controlled, clinical trials of rofecoxib. Circulation 2001; 104:2280–2288.
206. Van Hecken A, Schwartz JI, Depre M, et al. Comparative inhibitory activity of rofecoxib, meloxicam, diclofenac, ibuprofen, and naproxen on COX-2 versus COX-1 in healthy volunteers. J Clin Pharmacol 2000;40:1109–1120.
207. Watson DJ, Rhodes T, Cai B, et al. Lower risk of thromboembolic cardiovascular events with naproxen among patients with rheumatoid arthritis. Arch Intern Med 2002;162:1105–1110.
208. FitzGerald GA, Cheng Y, Austin S. COX-2 inhibitors and the cardiovascular system. Clin Exp Rheumatol 2001;19(suppl 25): 31–36.
209. Cheng Y, Austin SC, Rocca B, et al. Role of prostacyclin in the cardiovascular response to thromboxane A2. Science 2002;296:539–541.
210. Namazy JA, Simon RA. Sensitivity to nonsteroidal anti-inflammatory drugs. Ann Allergy Asthma Immunol 2002;89:542–550; quiz 550, 605.
211. Stevenson DD, Sanchez-Borges M, Szczeklik A. Classification of allergic and pseudoallergic reactions to drugs that inhibit cyclooxygenase enzymes. Ann Allergy Asthma Immunol 2001;87:177–180.
212. McDonald JR, Mathison DA, Stevenson DD. Aspirin intolerance in asthma. Detection by oral challenge. J Allergy Clin Immunol 1972;50:198–207.
213. Giraldo B, Blumenthal MN, Spink WW. Aspirin intolerance and asthma. A clinical and immunological study. Ann Intern Med 1969; 71:479–496.
214. Arm JP, O’Hickey SP, Spur BW, et al. Airway responsiveness to histamine and leukotriene E4 in subjects with aspirin-induced asthma. Am Rev Respir Dis 1989;140:148–153.
215. Christie PE, Tagari P, Ford-Hutchinson AW, et al. Urinary leukotriene E4 concentrations increase after aspirin challenge in aspirin-sensitive asthmatic subjects. Am Rev Respir Dis 1991;143:1025–1029.
216. Mastalerz L, Gawlewicz-Mroczka A, Nizankowska E, et al. Protection against exercise-induced bronchoconstriction by montelukast in aspirin-sensitive and aspirin-tolerant patients with asthma. Clin Exp Allergy 2002;32:1360–1365.
217. Dahlen SE, Malmstrom K, Nizankowska E, et al. Improvement of aspirin-intolerant asthma by montelukast, a leukotriene antagonist: a randomized, double-blind, placebo-controlled trial. Am J Respir Crit Care Med 2002;165:9–14.
218. Stevenson DD. Aspirin desensitization in patients with AERD. Clin Rev Allergy Immunol 2003;24:159–168.
219. Gyllfors P, Bochenek G, Overholt J, et al. Biochemical and clinical evidence that aspirin-intolerant asthmatic subjects tolerate the cyclooxygenase 2-selective analgetic drug celecoxib. J Allergy Clin Immunol 2003;111:1116–1121.
220. Woessner KM, Simon RA, Stevenson DD. The safety of celecoxib in patients with aspirin-sensitive asthma. Arthritis Rheum 2002;46: 2201–2206.
221. Martin-Garcia C, Hinojosa M, Berges P, et al. Safety of a cyclooxygenase-2 inhibitor in patients with aspirin-sensitive asthma. Chest 2002;121:1812–1817.
222. Szczeklik A, Nizankowska E, Bochenek G, et al. Safety of a specific COX-2 inhibitor in aspirin-induced asthma. Clin Exp Allergy 2001; 31:219–225.
223. Marks F, Harrell K, Fischer R. Successful use of cyclooxygenase-2 inhibitor in a patient with aspirin-induced asthma. South Med J 2001; 94:256–257.
224. Moore-Robinson M, Warin RP. Effect of salicylates in urticaria. BMJ 1967;4:262–264.
225. Sanchez-Borges M, Capriles-Hulett A, Caballero-Fonseca F. NSAID-induced urticaria and angioedema: a reappraisal of its clinical management. Am J Clin Dermatol 2002;3:599–607.
P.703

226. Perez C, Sanchez-Borges M, Capriles E. Pretreatment with montelukast blocks NSAID-induced urticaria and angioedema. J Allergy Clin Immunol 2001;108:1060–1061.
227. Pacor ML, Di Lorenzo G, Corrocher R. Efficacy of leukotriene receptor antagonist in chronic urticaria. A double-blind, placebo-controlled comparison of treatment with montelukast and cetirizine in patients with chronic urticaria with intolerance to food additive and/or acetylsalicylic acid. Clin Exp Allergy 2001;31:1607–1614.
228. Asero R. Leukotriene receptor antagonists may prevent NSAID-induced exacerbations in patients with chronic urticaria. Ann Allergy Asthma Immunol 2000;85:156–157.
229. Quiralte J, Saenz de San Pedro B, Florido JJ. Safety of selective cyclooxygenase-2 inhibitor rofecoxib in patients with NSAID-induced cutaneous reactions. Ann Allergy Asthma Immunol 2002;89: 63–66.
230. Pacor ML, Di Lorenzo G, Biasi D, et al. Safety of rofecoxib in subjects with a history of adverse cutaneous reactions to aspirin and/or non-steroidal anti-inflammatory drugs. Clin Exp Allergy 2002;32: 397–400.
231. Nettis E, Di PR, Ferrannini A, et al. Tolerability of rofecoxib in patients with cutaneous adverse reactions to nonsteroidal anti-inflammatory drugs. Ann Allergy Asthma Immunol 2002;88:331–334.
232. Sanchez Borges M, Capriles-Hulett A, Caballero-Fonseca F, et al. Tolerability to new COX-2 inhibitors in NSAID-sensitive patients with cutaneous reactions. Ann Allergy Asthma Immunol 2001;87: 201–204.
233. Asero R. Intolerance to nonsteroidal anti-inflammatory drugs might precede by years the onset of chronic urticaria. J Allergy Clin Immunol 2003;111:1095–1098.
234. Kemp SF, Lockey RF, Wolf BL, et al. Anaphylaxis. A review of 266 cases. Arch Intern Med 1995;155:1749–1754.
235. Davidson KA, Ringpfeil F, Lee JB. Ibuprofen-induced bullous leukocytoclastic vasculitis. Cutis 2001;67:303–307.
236. De Silva B, Banney L, Uttley W, et al. Pseudoporphyria and nonsteroidal antiinflammatory agents in children with juvenile idiopathic arthritis. Pediatr Dermatol 2000;17:480–483.
237. Cutaneous reactions to analgesic-antipyretics and nonsteroidal anti-inflammatory drugs. Analysis of reports to the spontaneous reporting system of the Gruppo Italiano Studi Epidemiologici in Dermatologia. Dermatology 1993;186:164–169.
238. Moore DE. Drug-induced cutaneous photosensitivity: incidence, mechanism, prevention and management. Drug Saf 2002;25:345–372.
239. Greenberg GN. Recurrent sulindac-induced aseptic meningitis in a patient tolerant to other nonsteroidal anti-inflammatory drugs. South Med J 1988;81:1463–1464.
240. Marinac JS. Drug- and chemical-induced aseptic meningitis: a review of the literature. Ann Pharmacother 1992;26:813–822.
241. Goodwin SD, Glenny RW. Nonsteroidal anti-inflammatory drug-associated pulmonary infiltrates with eosinophilia. Review of the literature and Food and Drug Administration Adverse Drug Reaction reports. Arch Intern Med 1992;152:1521–1524.
242. Yip L, Dart RC, Gabow PA. Concepts and controversies in salicylate toxicity. Emerg Med Clin North Am 1994;12:351–364.
243. Glasgow JF, Middleton B. Reye syndrome—insights on causation and prognosis. Arch Dis Child 2001;85:351–353.
244. Forsyth BW, Horwitz RI, Acampora D, et al. New epidemiologic evidence confirming that bias does not explain the aspirin/Reye’s syndrome association. JAMA 1989;261:2517–2524.
245. Hall SM. Reye’s syndrome and aspirin: a review. Br J Clin Pract Suppl 1990;70:4–11.
246. Belay ED, Bresee JS, Holman RC, et al. Reye’s syndrome in the United States from 1981 through 1997. N Engl J Med 1999;340: 1377–1382.
247. Roberts LJ 2nd, Morrow JD. Analgesic-antipyretic and antiinflammatory agents and drugs employed in the treatment of gout. In: Hardman JG, Limbird LE, eds. Pharmacological basis of therapeutics, 9th ed. New York: McGraw-Hill, 2001:687–731.
248. Brater DC. Drug-drug and drug-disease interactions with nonsteroidal anti-inflammatory drugs. Am J Med 1986;80:62–77.
249. Watkins J, Abbott EC, Hensby CN, et al. Attenuation of hypotensive effect of propranolol and thiazide diuretics by indomethacin. BMJ 1980;281:702–705.
250. Uwai Y, Saito H, Inui K. Interaction between methotrexate and nonsteroidal anti-inflammatory drugs in organic anion transporter. Eur J Pharmacol 2000;409:31–36.
251. Kremer JM, Hamilton RA. The effects of nonsteroidal antiinflammatory drugs on methotrexate (MTX) pharmacokinetics: impairment of renal clearance of MTX at weekly maintenance doses but not at 7.5 mg. J Rheumatol 1995;22:2072–2077.
252. Tracy TS, Krohn K, Jones DR, et al. The effects of a salicylate, ibuprofen, and naproxen on the disposition of methotrexate in patients with rheumatoid arthritis. Eur J Clin Pharmacol 1992;42: 121–125.
253. Iqbal MP, Baig JA, Ali AA, et al. The effects of non-steroidal anti-inflammatory drugs on the disposition of methotrexate in patients with rheumatoid arthritis. Biopharm Drug Dispos 1998;19:163–167.
254. Verbeeck RK. Pharmacokinetic drug interactions with nonsteroidal anti-inflammatory drugs. Clin Pharmacokinet 1990;19:44–66.
255. Ragheb M. The clinical significance of lithium-nonsteroidal anti-inflammatory drug interactions. J Clin Psychopharmacol 1990;10: 350–354.
256. Graham DY, Malaty HM. Alendronate and naproxen are synergistic for development of gastric ulcers. Arch Intern Med 2001;161: 107–110.
257. Kovarik JM, Mueller EA, Gerbeau C, et al. Cyclosporine and nonsteroidal antiinflammatory drugs: exploring potential drug interactions and their implications for the treatment of rheumatoid arthritis. J Clin Pharmacol 1997;37:336–343.
258. Tugwell P, Ludwin D, Gent M, et al. Interaction between cyclosporin A and nonsteroidal antiinflammatory drugs. J Rheumatol 1997;24: 1122–1125.
259. Davies NM, McLachlan AJ, Day RO, et al. Clinical pharmacokinetics and pharmacodynamics of celecoxib: a selective cyclo-oxygenase-2 inhibitor. Clin Pharmacokinet 2000;38:225–242.
260. Klose TS, Ibeanu GC, Ghanayem BI, et al. Identification of residues 286 and 289 as critical for conferring substrate specificity of human CYP2C9 for diclofenac and ibuprofen. Arch Biochem Biophys 1998; 357:240–248.
261. Yamazaki H, Inoue K, Chiba K, et al. Comparative studies on the catalytic roles of cytochrome P450 2C9 and its Cys- and Leu-variants in the oxidation of warfarin, flurbiprofen, and diclofenac by human liver microsomes. Biochem Pharmacol 1998;56:243–251.
262. Chesne C, Guyomard C, Guillouzo A, et al. Metabolism of meloxicam in human liver involves cytochromes P4502C9 and 3A4. Xenobiotica 1998;28:1–13.
263. Guidelines for monitoring drug therapy in rheumatoid arthritis. American College of Rheumatology Ad Hoc Committee on Clinical Guidelines. Arthritis Rheum 1996;39:723–731.
264. Hollingworth P. The use of non-steroidal anti-inflammatory drugs in paediatric rheumatic diseases. Br J Rheumatol 1993;32:73–77.
265. Berde CB, Sethna NF. Analgesics for the treatment of pain in children. N Engl J Med 2002;347:1094–1103.
266. Olkkola KT, Maunuksela EL. The pharmacokinetics of postoperative intravenous ketorolac tromethamine in children. Br J Clin Pharmacol 1991;31:182–184.
267. Cuzzolin L, Dal Cere M, Fanos V. NSAID-induced nephrotoxicity from the fetus to the child. Drug Saf 2001;24:9–18.
268. Pickering AE, Bridge HS, Nolan J, et al. Double-blind, placebo-controlled analgesic study of ibuprofen or rofecoxib in combination with paracetamol for tonsillectomy in children. Br J Anaesth 2002; 88:72–77.
269. Stempak D, Gammon J, Klein J, et al. Single-dose and steady-state pharmacokinetics of celecoxib in children. Clin Pharmacol Ther 2002;72:490–497.
270. Briggs GG, Freeman RK, Yaffe SJ. Drugs in pregnancy and lactation: a reference guide to fetal and neonatal risk, 6th ed. Philadelphia: Lippincott Williams & Wilkins, 2002.
271. Werler MM, Mitchell AA, Shapiro S. The relation of aspirin use during the first trimester of pregnancy to congenital cardiac defects. N Engl J Med 1989;321:1639–1642.
P.704

272. Collins E, Turner G. Letter: Salicylates and pregnancy. Lancet 1973; 2:1494.
273. Niebyl JR. Drugs in pregnancy and lactation. In: Gabbe SG, Niebyl JR, Simpson JL, eds. Obstetrics: normal and problem pregnancies, 4th ed. New York: Churchill Livingstone, 2002:221–50.
274. Laine L. Gastrointestinal safety of coxibs and outcomes studies: what’s the verdict? J Pain Symptom Manage 2002;23(suppl):5–10; discussion 11–14.
275. Holm L, Phillipson M, Perry MA. NO-flurbiprofen maintains duodenal blood flow, enhances mucus secretion contributing to lower mucosal injury. Am J Physiol Gastrointest Liver Physiol 2002;283: G1090–G1097.
276. Wallace JL, Reuter B, Cicala C, et al. Novel nonsteroidal anti-inflammatory drug derivatives with markedly reduced ulcerogenic properties in the rat. Gastroenterology 1994;107:173–179.