Coxibs Beyond the GI Tract: Renal and Cardiovascular Issues

Vol. 25 No. 2S February 2003 Journal of Pain and Symptom Management S41 Perspectives in Pain Management: The Role of Coxibs Coxibs Beyond the GI Tract: Renal and Cardiovascular Issues Anthony N. DeMaria, MD and Matthew R. Weir, MD Division of Cardiology (A.N.D.), University of California at San Diego School of Medicine, San Diego, California; and Division of Nephrology (M.R.W.), University of Maryland School of Medicine, Baltimore, Maryland, USA Abstract Although the coxibs have demonstrated superior gastrointestinal safety compared to traditional non-selective NSAIDs, questions remain regarding their effects on the renal and cardiovascular systems. In terms of renal function, both Type 1 and Type 2 cyclooxygenase (COX-1 and COX-2) are expressed constitutively in the kidney. Prostaglandins do not play a major role in the maintenance of renal function in healthy individuals but they become profoundly important in certain clinical situations such as renal stress, or volume depletion. In such situations the effects of the coxibs and non-selective NSAIDs are likely to be similar. The incidence of renal side effects is low (1 5%) and the patients at risk of renal complications are well defined and can usually be identified prospectively and followed up as appropriate. From the cardiovascular point of view, questions have been raised as to whether the coxibs have a prothrombotic effect. Here we review the available evidence and consider various hypotheses for an apparent increase in cardiovascular events reported in one coxib study (the VIGOR trial). Because of a lack of anti-platelet activity, coxibs are not suited for the provision of cardiovascular prophylaxis, and in patients at risk of myocardial infarction the prophylactic use of aspirin should always be considered. Although evidence suggests that use of coxibs with low-dose aspirin is safer than the combination of traditional NSAIDs with aspirin, further studies are required to confirm that this is the case. J Pain Symptom Manage 2003;25S:S41 S49. 2003 U.S. Cancer Pain Relief Committee. Published by Elsevier. All rights reserved. Key Words Renal, cardiovascular, hypertension, sodium excretion, safety, prostacyclin, COX-1, COX-2, cyclooxygenase, myocardial infarction Introduction The development of non-selective NSAIDs and, more recently, selective inhibitors of the cyclooxygenase-2 (COX-2) isoform, has contributed greatly towards the effective management Address reprint requests to: Matthew R. Weir, MD, Division of Nephrology, N3W143, University of Maryland Hospital, 22 South Greene Street, Baltimore, MD 21201, USA. of patients with arthritis and pain disorders. Although selective COX-2 inhibitors (coxibs) have demonstrated an improved gastrointestinal (GI) tolerability profile compared with non-selective NSAIDs 1 4 the two drug types do, nonetheless, share a number of renal syndromes that are exhibited in a specific group of at-risk patients. In addition to these well-characterized events, questions have arisen about the relative risk of cardiovascular effects associated with 2003 U.S. Cancer Pain Relief Committee 0885-3924/03/$ see front matter Published by Elsevier. All rights reserved. PII S0885-3924(02)00630-9

S42 DeMaria and Weir Vol. 25 No. 2S February 2003 the use of coxibs, particularly with regard to myocardial infarction. 5 Prostaglandins are produced by the breakdown of arachidonic acid by the cyclooxygenase enzyme, and are widely distributed throughout the body. Autacoid prostaglandins have a number of important functions, including protection of the gastric mucosa, maintenance of hemostasis and clotting, and actions upon renal physiology, including modulation of sodium absorption and glomerular filtration rate (GFR). Although prostaglandins do not play a major role in the maintenance of renal function in healthy individuals, they are known to become profoundly important in certain clinical situations. 6 In patients experiencing renal stress, such as renal hypoperfusion or low sodium intake, COX-2 expression (and therefore prostaglandin production) is upregulated in the kidney. The compensatory vasodilation evoked by these prostaglandins plays a central role in the maintenance of renal perfusion in these patients. 7 In terms of cardiovascular function, the COX-1 isoform is constitutive within platelets and is associated with the production of thromboxane, a potent vasoconstrictor, and strongly promotes platelet aggregation. Thus, COX-1 is important in maintaining hemostasis, and does this in a manner that facilitates thrombosis. Conversely, both COX-1 and COX-2 isoforms can cause endothelial cells to produce prostacyclin, a powerful vasodilator, particularly in response to increased blood flow shear forces. It is therefore possible that, by blocking thromboxane production, COX-1 inhibition might reduce the potential for thrombosis and vasoconstriction, while COX-2 inhibition might potentially block prostacyclin production and thus diminish vasodilation and inhibition of platelet aggregation. Non-specific NSAIDs and aspirin block both COX isoforms, and therefore have a balanced effect of reducing the prothrombotic effects of thromboxane and the antithrombotic effects of prostacyclin. Selective COX-2 inhibition by coxibs has the potential to blunt the vasodilatory and platelet inhibitory effects of prostacyclin while leaving the production of thromboxane unopposed. Accordingly, there is a theoretical basis for an association between coxib use and the occurrence of thrombotic phenomena. 8 This review explores the potential effects of non-selective NSAIDs and coxibs upon the renal and cardiovascular systems, and weighs the evidence for an effect of the COX-2 specific inhibitors on the cardiovascular system. Role of COX-1 and COX-2 in the Kidney Both COX isoforms are present constitutively in the human kidney. COX-1 is found in the glomerulus and afferent arteriole, while COX- 2 is located in the podocytes, thick ascending limb of the loop of Henle, macula densa and afferent arteriole (See Fig. 1). 9,10 These distinct areas of distribution correlate with the different effects on renal function displayed by COX-1 and COX-2. Prostaglandins produced by COX- 1 primarily affect renal homeostasis, and promote dilation of the renal vascular bed, lowering of renal vascular resistance and increases in renal perfusion. 11 In contrast, prostaglandins produced by COX-2 have diuretic and natriuretic effects (Fig. 2). Under physiological conditions, prostaglandins do not play a major role in the maintenance of basal renal function. However, in situations Fig. 1. Distribution of COX-1 and COX-2 in the human kidney. Fig. 2. Role of prostaglandins in the kidney.

Vol. 25 No. 2S February 2003 Coxibs: Renal and Cardiovascular Issues S43 where renal hemodynamics are compromised, such as in subjects with volume depletion, prostaglandins have a profound effect on the maintenance of renal perfusion and GFR by virtue of their vasodilatory effects. In these prostaglandindependent states, NSAID-mediated prostaglandin inhibition can result in acute renal failure. 12 Renal Effects of NSAIDs and Coxibs The frequency of NSAID-associated renal side effects is low, and has been reported to occur in 1 5% of NSAID users. 13 The commonlyreported renal effects of NSAIDs are reductions in renal blood flow, GFR and sodium and potassium excretion, which can manifest as fluid retention, edema, hypertension and hyperkalemia. These effects only tend to result in clinically appreciable problems in patients with prostaglandin-dependent renal function. As such, though the large number of NSAID users magnifies the incidence of nonsteroidal renal syndromes, the patient groups at risk are well defined and can usually be identified prospectively. Edema and GFR Effects The activity of NSAIDs and coxibs on COX-2 results in inhibition of prostaglandin E2 production, predominantly in the thick ascending limb of the loop of Henle. This tends to lead to increased sodium and water reabsorption and may manifest as small weight increases and edema, and a decreased sensitivity to diuretics. 12 In patients with a propensity for sodium retention, such as those with liver disease or congestive heart failure, these effects are more pronounced. A study in healthy young males receiving a lowsodium diet, (in whom renal function may have been more dependent on prostaglandins), reported a transient decrease in GFR and renal blood flow with celecoxib. 14 Although these effects were not seen with the comparator drug naproxen, the two drugs were associated with similar decreases in sodium and potassium excretion. In healthy elderly subjects, in whom no control was used for salt intake, naproxen was found to produce a transient decrease in GFR with the first dose, while no effect was noted with celecoxib. Both drugs were associated with a slight, transient decrease sodium excretion. 15 In healthy, sodium-replete, elderly subjects the use of rofecoxib and indomethacin was associated with a significant decrease in sodium excretion during the first 72 hours of therapy, which normalized over 12 days in the rofecoxib group. 16 Subjects receiving indomethacin also experienced a significant decrease in GFR compared with rofecoxib and placebo. In salt-restricted, elderly subjects with mild-to-moderate renal dysfunction, in whom prostaglandins would be expected to play a role in renal function, significant reductions in GFR were recorded with rofecoxib and indomethacin compared with placebo. 17 Hypertension The use of non-selective NSAIDs may induce dose-related fluid retention and blood pressure increases. As such, sodium intake can influence blood pressure in subjects taking these drugs, and this effect may be heightened in higher-risk individuals, such as the elderly or those with hypertension, heart disease or kidney disease. Few studies have been undertaken to investigate the blood pressure effects of NSAIDs and coxibs. Early meta-analyses reported modest increases in blood pressure in hypertensive patients receiving short-term NSAID therapy, 18,19 although these studies have been questioned regarding methodological shortcomings. Population-based studies have reported significant increases in the initiation of antihypertensive agents in NSAID users, 20 and increases in blood pressure in subjects with previously controlled hypertension. 21 It has been postulated that NSAIDs may interact with the prostaglandin-mediated activity of ACE inhibitors, beta-blockers and diuretics while having little effect on calcium antagonists, whose activity is unrelated to prostaglandin production. 12,22 24 Studies have reported a blunting of the antihypertensive effects of enalapril and captopril, and lesser effects on nifedipine, amlodipine and losartan with indomethacin. 25 27 The blood pressure effects of the coxibs have been investigated as part of studies in arthritis, though the lack of documentation of baseline values makes analysis of the findings difficult. The Celecoxib Long-term Arthritis Safety Study (CLASS) reported a low incidence of hypertension, with a significantly higher incidence in patients receiving indomethacin compared with celecoxib or diclofenac. 2 The SUC- CESS-I trial (SUCcessive Celecoxib Efficacy and Safety Study), which studied the GI tolerability of celecoxib, diclofenac and naproxen in osteoarthritis, reported a low rate of hypertension in all treatment groups. 28 A similarly low incidence of hypertension was recorded in a

S44 DeMaria and Weir Vol. 25 No. 2S February 2003 long-term study in patients with osteoarthritis receiving rofecoxib or diclofenac. 29 Although the blood pressure effects of the coxibs have not been compared directly in many studies, one trial has reported a similar increase in blood pressure in normotensive osteoarthritis sufferers receiving rofecoxib or celecoxib, and a higher incidence in subjects receiving acetaminophen. 30 Furthermore, the destabilizing effects of celecoxib and rofecoxib on blood pressure control and other cardiorenal events have been investigated in elderly osteoarthritis patients receiving antihypertensive therapy. 31 Increases in mean systolic blood pressure and edema were significantly more common in rofecoxib-treated patients. However, this study was not randomized for antihypertensive therapy and there were significantly higher numbers of ACE-inhibitor and diuretic users in the celecoxib group and higher numbers of calcium channel blocker users in the rofecoxib group. Changes in antihypertensive therapy during the study were not accounted for. In addition, due to the different half-lives of celecoxib and rofecoxib (approximately 8 11 hours and 17 hours, respectively), the blood pressure effects of these drugs would have been affected by the timing of blood pressure measurements. The findings from these studies indicate that 1 3% of subjects receiving NSAIDs or coxibs will experience clinically significant increases in blood pressure. Those patients most at risk include elderly subjects and individuals with congestive heart failure, liver disease, kidney disease or hypertension, particularly those patients treated with an ACE inhibitor or diuretic. However, some of these increases in blood pressure may represent undiagnosed or borderline hypertension. When nonsteroidal renal syndromes occur, they may necessitate a change in antihypertensive medication. Similarities Between NSAIDs and Coxibs Clinical experience of the use of NSAIDs and coxibs in various patient populations indicates that, though coxibs demonstrate an improved GI tolerability profile, there are similarities between the two drug types in their effects on renal hemodynamics and homoeostasis. Consequently, the same precautions applied for the use of NSAIDs in patients at risk of nephrotoxic events should be used for the coxibs. Other Nephrotoxic Syndromes In addition to these events, several uncommon nephrotoxic syndromes have been reported with NSAID use that are unrelated to the inhibition of COX expression or prostaglandin synthesis. These conditions generally resolve upon treatment discontinuation, and include acute interstitial nephritis, which may be accompanied by minimal-change glomerulonephritis or heavy proteinuria, nephrotic syndrome, chronic interstitial nephritis, papillary necrosis and acute renal failure. These syndromes are seldom documented with coxibs, though the use of these agents is not as widespread. Patient Advice The majority of patients receiving NSAIDs or coxibs are unlikely to suffer adverse renal events during therapy. Those patients at risk of these events can be identified prospectively and monitored accordingly. As in all medical therapy, follow up is of great importance. In those patients at risk, the use of lower drug doses may be appropriate, with adjustments being made to the dose of antihypertensive medication where necessary. Patients should also be advised to lower their dietary salt intake, in particular to avoid highly processed foods. They should also be cautioned to avoid the use of over-the-counter analgesics. (Table 1). Cardiovascular Effects of NSAIDs and COXIBS A recent article by Mukherjee et al. suggested that the use of coxibs may be associated with an increased risk of cardiovascular events, particularly myocardial infarction. 5 This analysis compared the incidence of myocardial infarction in patients receiving rofecoxib in the VIOXX Gastrointestinal Outcomes Research (VIGOR) study 1 and those receiving celecoxib in the Celecoxib Long-term Arthritis Safety Study (CLASS) 2 Table 1 Action in Case of Increased Blood Pressure on NSAIDs/Coxibs Reduce dose of anti-inflammatory therapy Restrict dietary sodium intake Sodium particularly high in processed foods Inquire about home remedies/over-the-counter medications Adjust anti-hypertensive medication as appropriate

Vol. 25 No. 2S February 2003 Coxibs: Renal and Cardiovascular Issues S45 with the incidence in healthy patients receiving placebo in four older aspirin primary prevention trials. The incidence of myocardial infarction was reported to be comparable in patients receiving rofecoxib in the VIGOR study and celecoxib in the CLASS study (0.74 vs. 0.80, respectively), but significantly lower in the placebo group (0.52) employed in the analysis. This paper also reported that the incidence of cardiovascular events was 2.38 times more common with rofecoxib than with naproxen in the VIGOR study, while there was no significant difference between celecoxib and the comparator NSAIDs in the CLASS study. It is apparent that, though the incidence of myocardial infarction with rofecoxib and celecoxib was similar, it was substantially higher than naproxen. VIGOR Study: Consideration of Cardiovascular Findings Although Mukherjee et al. undertook a metaanalytical approach, the article did not fulfill all the criteria of a true meta-analysis. The rheumatoid arthritis sufferers recruited by the VIGOR study, and the osteoarthritis sufferers making up the majority of patients recruited to the CLASS study, represent very different study populations to the healthy subjects enrolled in the comparator aspirin studies (vide infra). As the authors of the article state, neither VIGOR nor CLASS was designed to assess the risk of cardiovascular events, and in the meta-analysis these studies were compared with trials in average risk placebo patients from aspirin primary prevention trials, in which cardiovascular events were the primary endpoint. As such, this investigation was not conducted with studies or study populations that were truly comparable. The apparent increased risk of myocardial infarction noted in rofecoxib-treated patients in the VIGOR study has been the focus of speculation regarding the cardiovascular effects of the coxibs. A number of hypotheses have been derived to explain the observed increase in relative risk with rofecoxib compared with naproxen. Chance The first consideration is that, given the low number of actual events observed, it is entirely possible that the results of the VIGOR study were related to a play of chance. In all, 66 cardiovascular events were observed in over 8,000 patients, 46 of which occurred in patients receiving rofecoxib. 1 Accordingly, a small shift in the number of events may well have altered the significance of these findings. Study Design A number of aspects of the design of the VIGOR study may have been responsible for the observed increase in cardiovascular events. The study recruited patients with rheumatoid arthritis, an inflammatory condition that has been linked with an increased risk of cardiovascular events. 32 A recent study of rheumatoid arthritis and osteoarthritis sufferers has demonstrated an increased cardiovascular risk of 2.0 in the rheumatoid arthritis patients after adjusting for age, gender and other covariants. 33 In addition, patients with rheumatoid arthritis also often have increased plasma levels of C-reactive protein, a marker that has clearly been correlated with an increased risk of cardiovascular events. 34 Consequently, it is reasonable to anticipate an increased risk of cardiovascular events in patients enrolled in the VIGOR study. Further, there is increasing evidence that inflammation plays an important role in the transformation of stable coronary disease to acute coronary events. 35,36 Inflammation within a stable plaque may result in weakening and subsequent fissure or rupture, leading to a superimposed thrombus and unstable angina or myocardial infarction. Given that patients with rheumatoid arthritis have a highly inflammatory state, they may be particularly vulnerable to plaque rupture. If patients recruited to the VIGOR study were at increased risk of cardiovascular events due to rheumatoid arthritis, the potential prophylactic efficacy of the comparator agent naproxen would be of particular relevance. There is evidence of considerable variability between different NSAIDs regarding their effects on platelet inhibition, and naproxen seems to have both a high degree and a long duration of inhibiting platelet aggregation. 37 Indeed, recent reports strongly suggest that patients treated with naproxen have a decreased risk of myocardial infarction compared with patients receiving non-naproxen NSAIDs or those receiving no NSAID therapy. 38,39,40 It is possible that the reason for the increased incidence of MI observed with rofecoxib in the VIGOR trial may be due to a protective effect of naproxen (mediated through the inhibition of platelet aggregation) rather than a thrombolytic effect of rofecoxib. 41

S46 DeMaria and Weir Vol. 25 No. 2S February 2003 Exclusion of Aspirin Patients enrolled in the VIGOR study were not permitted to use aspirin. Given the documented benefits of aspirin in patients with known cardiovascular disease, this design aspect clearly held the potential to impact on the study findings. The degree of benefit resulting from the antiplatelet effect of naproxen would be influenced by the concomitant use of aspirin, and would be expected to be of greater consequence in the absence of aspirin. The International Study of Infarct Survival (ISIS-2) reported that aspirin was associated with a 23% reduction in mortality and a 49% reduction in reinfarction. 42 The Antiplatelet Trialist Collaboration also revealed a 25% reduction in non-fatal myocardial infarction, stroke and vascular death when aspirin was used for secondary prevention. 43 However, the benefit of aspirin in the primary prevention of cardiovascular events has not been established, and a general recommendation for the use of aspirin for primary prevention has not been forthcoming. As such, in interpreting the data of COX-2 inhibition in both VIGOR and CLASS vs. placebo, it is important to be aware of the difference in the event rates associated with primary and secondary prevention. The potential role of the decision to exclude aspirin in the VIGOR study can be suspected by analyzing the occurrence of cardiovascular events in enrolled patients who were candidates for secondary prevention. In VIGOR, only 4% of patients met clearly established criteria for the use of aspirin for secondary cardiovascular prophylaxis, but 34% of myocardial infarctions occurred in this small subgroup. 1 As such, the absence of aspirin in these patients appears to be associated with a Fig. 3. Vascular events with rofecoxib: meta-analysis of studies of (6 months duration or more). 41 Relative risk (95% CI) of the APTC (Antiplatelet Trialist Collaboration) end point for rofecoxib relative to placebo, nonnaproxen NSAIDs, and naproxen in studies 6 months in duration. Triangles represent relative risk, and size of triangles represents patient-years of exposure. substantial incidence of myocardial infarction in those subjects treated with rofecoxib. It is possible that the antiplatelet effects of naproxen may have offset the risk of excluding aspirin in the patient population assigned to this nonspecific NSAID. Further Clinical Experience with Rofecoxib The incidence of cardiovascular events reported in the Mukherjee analysis differs from the general experience with rofecoxib derived from a meta-analysis of 23 studies that recruited over 28,000 patients, representing more than 14,000 patient-years at risk. 44 A significant proportion of these patient-years came from trials in Alzheimer s disease, in which the majority of subjects were older males and were consequently at increased cardiovascular risk. This analysis found that the risk of having a strictly adjudicated cardiovascular event was 0.84 when comparing rofecoxib with placebo, 0.79 when comparing rofecoxib with non-naproxen NSAIDs and 1.69 when comparing rofecoxib with naproxen. Thus, this large meta-analysis of comparable studies found no evidence of an increased risk of cardiovascular events relative to placebo or non-naproxen NSAIDs. When the meta-analysis was limited to studies of at least 6 months duration, these findings were unchanged (Fig. 3). As such, the VIGOR study stands apart from other data. Biology of COX-2 Inhibition Finally, the possibility that the theoretical imbalance between thromboxane and prostacyclin induced by selective COX-2 inhibition may have facilitated thrombotic events, such as infarction, cannot be completely dismissed. Nonspecific NSAIDs inhibit the effects of both COX-1 and COX-2, reducing the vasoconstrictor and thrombotic effects of thromboxane as well as the vasodilator and antithrombotic effects of prostacyclin. However, selective COX-2 inhibition reduces the favorable effects of prostacyclin while leaving the thrombotic actions of thromboxane unopposed, an imbalance that may be responsible for an increased risk of thrombosis. Subjects receiving coxibs demonstrate substantially reduced levels of urinary prostacyclin compared with placebo, though less than would be expected with deletion or knockout protocols in animals. 45 Recent data from animal studies suggest that deletion of the prostacyclin receptor leads to an increased

Vol. 25 No. 2S February 2003 Coxibs: Renal and Cardiovascular Issues S47 response to thrombogenic stimuli. 46 The clinical relevance of these experimental findings remains uncertain, and the hypothesis that an imbalance between thromboxane and prostacyclin may have played a role in the cardiovascular events seen in the VIGOR study remains to be tested fully. Conclusions Coxibs are anti-inflammatory agents designed to lower the risk of GI ulceration and bleeding. Over 90% inhibition of platelet thromboxane production is necessary to prevent platelet aggregation and produce a vascular protective effect, 47 and aspirin and naproxen are documented to achieve these results. 37,44 Evidence has been reported that coxibs and non-naproxen NSAIDs have similar myocardial infarction event rates in clinical trials, and that these rates are similar to those seen with placebo. 1,2,44 These trends indicate that the cardiovascular findings of the VIGOR study differed relative to the general experience with rofecoxib, and raise the possibility that the antithrombotic effect of naproxen may have been responsible for the difference in relative risk between the two treatment groups. It is important to recognize that the coxibs have been documented to give a risk reduction of around 50% in clinical upper GI events compared with non-selective NSAIDs. 1,2 The magnitude of this benefit can be best appreciated by considering that around 16,000 deaths result from the GI complications of NSAID use each year, almost the same number of deaths as resulted from AIDS. 48 A number of theories have been advanced to explain the cardiovascular events documented in the VIGOR study. Although it is possible that an imbalance between thromboxane and prostacyclin induced by selective COX-2 inhibition may have favored a thrombotic state, important aspects of study design may have been responsible for the results obtained in this trial. Specifically, the exclusion of aspirin in a group of patients with rheumatoid arthritis in whom cardiovascular risk was likely increased, coupled with the substantial antiplatelet effects of naproxen, probably played important roles in influencing the relative risk for cardiovascular events. Regardless of the explanation for these findings, when selecting anti-inflammatory therapy it is vital to balance the hypothetical cardiovascular risk with the documented GI benefit produced by selective COX-2 inhibition. Coxibs offer a definite advantage in patients at risk for GI events; it seems clear that patients with an established indication for aspirin therapy as a secondary prevention for cardiovascular events should receive low-dose aspirin in conjunction with coxib therapy. An important, and as yet unanswered, question is whether the addition of low-dose aspirin to coxib therapy will reduce or eliminate the beneficial, GI-sparing effects of these agents. A recent meta-analysis of 24 studies investigating long-term aspirin therapy reported that 2.3% of around 66,000 patients taking low-dose aspirin ( 163 mg/day) experienced a GI hemorrhage compared with 1.4% taking placebo. 49 Although not powered to detect such differences, the CLASS study found that subjects receiving celecoxib plus aspirin had a lower risk of GI complications than those receiving an NSAID plus aspirin. 2 In addition, the SUCCESS-I study reported a significant reduction in the risk of GI adverse events with celecoxib compared with non-selective NSAIDs when used concomitantly with lowdose aspirin. 50 Although the data are limited, they suggest that the coxibs may still offer an advantage in reducing upper GI events compared with NSAIDs when combined with aspirin. Based upon clinical and experimental evidence so far, what conclusions can be drawn to provide the basis for the current use of coxibs with respect to renal and cardiovascular events? Due to their similar effects on the kidney, the same precautions used for NSAID use in individuals at risk of nephrotoxic events should be used for coxibs. Despite the widespread use of NSAIDs, both prescription and over-the-counter, the occurrence of renal adverse events is uncommon. However, in patients at risk of nonsteroidal renal syndromes, careful monitoring is warranted, with amendments in NSAID and concurrent drug therapy made in cases where these syndromes occur. Coxibs are not well suited for the provision of cardiovascular prophylaxis, and in patients at risk of myocardial infarction the use of aspirin should always be considered. Until all the answers are available, physicians need to apply their best judgment in selecting the optimal anti-inflammatory agent for the individual patient.

S48 DeMaria and Weir Vol. 25 No. 2S February 2003 References 1. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:1520 1528. 2. 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. JAMA 2000;284:1247 55. 3. Laine L, Harper S, Simon T, et al. A randomized trial comparing the effect of rofecoxib, a cyclooxygenase 2-specific inhibitor, with that of ibuprofen on the gastroduodenal mucosa of patients with osteoarthritis. Gastroenterology 1999;117:776 783. 4. Langman MJ, Jensen DM, Watson DJ, et al. Adverse upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. JAMA 1999;282:1929 1933. 5. Mukherjee D, Nissen SE, Topol EJ. Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA 2001;286:954 959. 6. Carmichael J, Shankel SW. Effects of nonsteroidal anti-inflammatory drugs on prostaglandins and renal function. Am J Med 1985;78:992 1000. 7. Toto RD. The role of prostaglandins in NSAID-induced renal dysfunction. J Rheumatol 1991;28:22 25. 8. Belton O, Byrne D, Kearney D, Leahy A, Fitzgerald DJ. Cyclooxygenase-1 and -2 -dependent prostacyclin formation in patients with atherosclerosis. Circulation 2000;102:840 845. 9. Nantel F, Meadows E, Denis D, et al. Immunolocalization of cyclooxygenase-2 in the macula densa of human elderly. FEBS Lett 1999;457:475 477. 10. Schnermann J, Briggs JP. The macula densa is worth its salt. J Clin Invest 1999;104:1007 1009. 11. Weir MR, Froch L. Weighing the renal effects of NSAIDs and COX-2 inhibitors. Clinical Dilemmas 2000;1:3 12. 12. Brater DC. Effects of nonsteroidal anti-inflammatory drugs on renal function: focus on cyclooxygenase-2-selective inhibition. Am J Med 1999;107(Suppl. 6A):65S 71S. 13. Whelton A, Hamilton CW. Nonsteroidal anti-inflammatory drugs: effects on kidney function. J Clin Pharmacol 1991;31:588 598. 14. Rossat J, Maillard M, Nussberger J, Brunner HR, Burnier M. Renal effects of selective cyclooxygenase- 2 inhibition in normotensive salt-depleted subjects. Clin Pharmacol Ther 1999;66:76 84. 15. Whelton A, Schulman G, Wallemark C, et al. Effects of celecoxib and naproxen on renal function in the elderly. Arch Intern Med 2000;160:1465 1470. 16. 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. 17. Swan SK, Rudy DW, Lasseter KC, et al. Effects 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. 18. 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. 19. Johnson AG, Nguyen TV, Day RO. Do nonsteroidal anti-inflammatory drugs affect blood pressure? A meta-analysis. Ann Intern Med 1994;121: 289 300. 20. Gurwitz JH, Avorn J, Bohn RL, et al. Initiation of antihypertensive treatment during nonsteroidal antiinflammatory drug therapy. JAMA 1994;272:781 786. 21. Chrischilles EA, Wallace RB. Nonsteroidal antiinflammatory drugs and blood pressure in an elderly population. J Gerontol 1993;48:91 96. 22. Beckmann ML, Gerber JG, Byyny RL, Lo-Verde M, Nies AS. Propranolol increases prostacyclin synthesis in patients with essential hypertension. Hypertension 1988;12:582 588. 23. Polónia J. Interaction of antihypertensive drugs with anti-inflammatory drugs. Cardiology 1997;88 (Suppl. 3):47 51. 24. Houston MC, Weir M, Gray J, et al. The effects of nonsteroidal anti-inflammatory drugs on blood pressures of patients with hypertension controlled by verapamil. Arch Intern Med 1995;155:1049 1054. 25. Polónia J, Bonaventura I, Gama G, et al. Influence of nonsteroidal anti-inflammatory drugs on renal function and 24-hour ambulatory blood pressure-reducing effects of enalapril and nifedipine gastrointestinal therapeutic system in hypertensive patients. J Hypertens 1995;13:925 931. 26. Morgan TO, Anderson A, Bertram D. Effect of indomethacin on blood pressure in elderly people with essential hypertension well controlled on amlodipine or enalapril. Am J Hypertens 2000;13:1161 1167. 27. Conlin PR, Moore TJ, Swartz SL, et al. Effect of indomethacin on blood pressure lowering by captopril and losartan in hypertensive patients. Hypertension 2000;36:461 465. 28. Whelton A, Singh G, White W, et al. Celecoxib does not increease the risk of cardiac failure, edema, or hypertension compared to NSAIDs: Results from SUCCESS-1, a double blind, randomized trial in 13,274 OA patients. Ann Rheumat Dis 2001;60(Suppl. 1): 237 (abstract). 29. Cannon GW, Caldwell JR, Holt P, et al. Rofecoxib, a specific inhibitor of cyclooxygenase-2, with clinical efficacy comparable with that of diclofenac sodium. Arthritis Rheum 2000;43:978 987. 30. Geba GP, Polis AB, Dixon ME, et al. Comparative blood pressure effects of rofecoxib, celecoxib and placebo in patients with osteoarthritis (OA): a

Vol. 25 No. 2S February 2003 Coxibs: Renal and Cardiovascular Issues S49 randomized controlled trial. Ann Rheum Dis 2001; 60(Suppl. 1):238. 31. Whelton A, Fort JG, Puma JA, et al. Cyclooxygenase-2-specific inhibitors and cardiorenal function: a randomized, controlled trial of celecoxib and rofecoxib in older hypertensive osteoarthritis patients. Am J Ther 2001;8:85 95. 32. Wallberg-Jonsson S, Johansson H, Ohman ML, Rantapaa-Dahlqvist S. Extent of inflammation predicts cardiovascular disease and overall mortality in seropositive rheumatoid arthritis: a retrospective cohort study from disease onset. J Rheumatol 1999;26: 2562 2571. 33. Wolfe F, Mitchell DM, Sibley JT, et al. The mortality of rheumatoid arthritis. Arthritis Rheum 1994; 37(4):481 494. 34. Rosalki SB. C-reactive protein. Int J Clin Pract 2001;55(4):269 270. 35. Koenig W. Inflammation and coronary heart disease: an overview. Cardiol Rev 2001;9:31 35. 36. Libby P, Simon DI. Inflammation and thrombosis: the clot thickens. Circulation 2001;103:1718 1720. 37. Van Hecken A, Schwartz JI, Depré 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. 38. Solomon DH, Glynn RJ, Levin R, et al. Nonsteroidal anti-inflammatory drug use and acute myocardial infarction. Arch Intern Med 162:1099 1105. 39. Watson DJ, Rhodes T, Cai B, et al. Lower risk of thromboembolic cardiovascular events with naproxen among patients with rheumatoid arthritis. Arch Intern Med 162:1105 1110. 40. Rahme E, Pilote L, LeLorier J. Association between naproxen use and protection against acute myocardial infarction. Arch Intern Med 162:1111 1116. 41. Dalen JE. Selective COX-2 inhibitors, NSAIDs, aspirin and myocardial infarction. Arch Intern Med 162:1091 1092. 42. ISIS-2 (Second International Study of Infarct Survival) Collaborative Group. Randomized trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of suspected acute myocardial infarction. Lancet 1988 8607:349 360. 43. Collaborative overview of randomized trials of antiplatelet therapy. I: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ 1994; 308:81 106. 44. Konstam MA, Weir MR, Reicin A, et al. Cardiovascular thrombotic events in controlled, clinical trials of rofecoxib. Circulation 2001;104:15 23. 45. McAdam BF, Catella-Lawson F, Mardini IA, et al. Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: The human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA. 96:272 277. 46. Murata T, Ushikubi F, Matsuoka T, et al. Altered pain perception and inflammatory responses in mice lacking prostacyclin receptor. Nature 1997; 388(6643):678 682. 47. Reilly IA, FitzGerald GA. Inhibition of thromboxane formation in vivo and ex vivo: implications for therapy with platelet inhibitory drugs. Blood 1987;69:180 186. 48. Singh G, Triadafilopoulos G. Epidemiology of NSAID-induced gastrointestinal complications. J Rheumatol 1999;26 Suppl 56:18 24. 49. Derry S, Loke YK. Risk of gastrointestinal hemorrhage with long-term use of aspirin: meta-analysis. BMJ 2000;321:1183 1187. 50. Singh G, Goldstein J, Agrawal N, et al. COX-2 specific inhibitors is there any benefit of using these agents in patients on low dose aspirin (ASA)? Ann Rheum Dis 2001;60(Suppl. 1):239.