A Whelton; J L Lefkowith; C R West; K M Verburg 

Kidney Int.  2006;70(8):1495-1502.  ?2006 Nature Publishing Group

Abstract and Introduction

Abstract

The cardiorenal safety database from the Celecoxib Long-term Arthritis Safety Study (CLASS) was analyzed to examine whether supratherapeutic doses of celecoxib are associated with decreased renal function and blood pressure (BP) effects compared with standard doses of diclofenac and ibuprofen in osteoarthritis (OA) and rheumatoid arthritis (RA) patients.

In total, 8059 patients were enrolled; 7968 received at least one dose of study drug (RA: N = 2183; OA: N = 5785). Patients received celecoxib, 400 mg twice a day (b.i.d.). (N = 3987); ibuprofen, 800 mg three times a day. (N = 1985); or diclofenac, 75 mg b.i.d. (N = 1996). Effects measured included: investigator-reported hypertension, edema or congestive heart failure, clinically important BP elevations, incidence of patients starting new antihypertensive medication, and increases in serum creatinine or reductions in creatinine clearance. Celecoxib was associated with a similar incidence of hypertension or edema to diclofenac but significantly lower than ibuprofen. The celecoxib group had significantly fewer initiations of antihypertensives versus ibuprofen. Systolic BP increases of >20 mmHg and above 140 mmHg occurred significantly less often with celecoxib compared with ibuprofen or diclofenac. Changes in serum creatinine or estimated creatinine clearance occurred in a similar percentage of patients taking celecoxib or ibuprofen; modest differences were evident against diclofenac. In patients with mild prerenal azotemia, significantly fewer patients taking celecoxib exhibited clinically important reductions in renal function (3.7%), compared with diclofenac (7.3%; P < 0.05) and ibuprofen (7.3%; P < 0.05). A supratherapeutic dose of celecoxib was associated with an improved cardiorenal safety profile compared with standard doses of either ibuprofen or diclofenac.

Introduction

Nonselective nonsteroidal anti-inflammatory drugs (NSAIDs) are commonly used to manage the pain and inflammation associated with arthritis. However, these agents inhibit both isoenzymes of cyclooxygenase (COX): COX-1 and COX-2.^[1] Therefore, the use of nonselective NSAIDs can result in a spectrum of mild-to-severe and occasionally life-threatening mechanism-based adverse events, particularly within the gastrointestinal tract, effects on platelets, and the cardiorenal axis, which may limit their therapeutic benefit.^{[2, 3, 4, 5, 6]} Celecoxib was the first of the COX-2 selective inhibitors (COXIBs) to be developed. At full therapeutic doses, celecoxib possesses comparable analgesic and anti-inflammatory efficacy to nonselective NSAIDs in the treatment of rheumatoid arthritis (RA) and osteoarthritis (OA), although sparing COX-1 activity, thereby avoiding many of the gastrointestinal- and platelet-related adverse events.^{[7, 8, 9, 10, 11]}

Nonselective inhibition of COX-1 and COX-2 attenuates the prostaglandin-mediated mechanisms responsible for modulating renal function, including renal vascular tone and electrolyte and water excretion, particularly during renal stress, as manifested by reduced renal perfusion.^{[12, 13]} The most common side effect is edema, which is followed by, and frequently linked with, destabilization of blood pressure (BP) control in treated hypertensive individuals. Retention of salt and water is also linked to the blunting of clinical response to diuretic treatment and may subsequently induce congestive heart failure in susceptible individuals. In the setting of reduced renal perfusion and, hence, dependence of renal perfusion upon the compensatory production of prostaglandin synthesis, the use of a nonselective NSAID may result in acute, hemodynamically mediated deterioration in renal function. This is fully reversible following discontinuation of the offending drug. Other renal syndromes are rare to very rare. They include other electrolyte complications, such as hyperkalemia, hyponatremia, nephrotic syndrome with interstitial nephritis, and papillary necrosis. The mechanisms of the induction of these rare events via inhibition of arachidonic acid processing by COX enzymes are reviewed in detail elsewhere.^{[14, 15]} It is remarkable to note that, following exposure to several days' treatment with nonselective NSAIDs, approximately one in five subjects will manifest a renal function abnormality.^[16]

COX-1 is expressed abundantly in the kidney. In the early development of celecoxib, it was assumed that COX-2 was not constitutively expressed in the healthy human kidney as was the case in most other tissues with the exception of the brain and the female reproductive system. However, subsequent studies have demonstrated that the COX-2 isoenzyme is indeed constitutively expressed in discrete areas of the adult mammalian kidney.^[17] For example, expression of COX-2 in the macula densa suggests a role in the production of prostaglandins regulating the release of renin. COX-2 expressed in the cortical thick ascending limb and medullary interstitial cells may be involved in the regulation of medullary blood flow. In the medulla, COX-2 expression decreases with salt depletion and increases with a high-salt diet and with dehydration.^{[18, 19]} Thus, as information accumulated, it became increasingly clear that COX-2 may serve an important role in the regulation of renal hemodynamic and/or renal excretory function under physiologic or pathophysiologic conditions.^{[20, 21]}

Typically, the use of a selective COX-2 inhibitor in otherwise healthy individuals is associated with mild edema, hypertension, and modest sodium retention for the first 1 or 2 days of therapy followed by a return to baseline excretion levels and elimination of retained salt and water.

As the kidney and vasculature are capable of expressing both COX-1 and COX-2, it may be hypothesized that a selective COX-2 inhibitor such as celecoxib may have decreased renal adverse effects relative to nonselective inhibitors. The Celecoxib Long-term Arthritis Safety Study (CLASS) was undertaken to compare the long-term upper gastrointestinal safety of celecoxib 400 mg twice a day (b.i.d.), two to four times the full therapeutic doses for RA and OA, respectively, with normal full therapeutic doses of diclofenac (75 mg b.i.d.) and ibuprofen (800 mg three times a day). A large cardiorenal safety database was accumulated in the CLASS trial, which was determined to be useful for studying cardiorenal-related adverse effects associated with nonselective NSAIDs and celecoxib. A separate publication has detailed the cardiovascular thrombotic complications observed during the CLASS trial.^[22]

Results

Patient disposition is outlined in Figure 1. A total of 8059 patients were randomized, of whom 7968 received at least one dose of study drug, and were therefore included in the intent-to-treat (ITT) population (RA: N = 2183; OA: N = 5785). In the ITT population, 3987 patients received celecoxib and 3981 were treated with either ibuprofen or diclofenac.

     

Baseline demographics, exposure to treatment, mean BP levels, and renal function are shown in Table 1 . Mean patient age was 60 years in each treatment group, most (69-71%) patients were female and mean BP was 133/80 mmHg. As determined by medical history, slightly less than 40% of patients in each treatment group had pre-existing hypertension. Overall BP levels and renal function were normal at baseline and comparable between treatment groups.

A summary of antihypertensive and diuretic medication use at baseline is summarized in Table 2 . Over 40% of the patients in each treatment group were receiving one or more of these agents, with nearly 20% receiving diuretics. Overall, the frequency and distribution of use of these drugs were similar across the treatment groups.

Blood Pressure

Celecoxib was associated with an incidence of hypertension (both new-onset and aggravated) that was comparable to standard doses of diclofenac and significantly lower than standard doses of ibuprofen (P<0.05) ( Table 3 ). In patients taking ibuprofen, the increased incidence of hypertension versus celecoxib was accompanied by a corresponding trend towards a mean increase in systolic BP (P = 0.09). Both ibuprofen and diclofenac treatments were associated with a significantly (P<0.05) greater proportion of patients with systolic BP elevations >20 mmHg and above 140 mmHg as compared with celecoxib (treatment differences 1.6-2.0%). No treatment differences were observed for diastolic BP ( Table 3 ). Withdrawals from study treatment owing to hypertension-related adverse events were uncommon and occurred with similar frequency among the three treatment groups ( Table 3 ).

The initiation of new antihypertensive and diuretic therapy was significantly higher with ibuprofen treatment as compared with celecoxib treatment and specifically for initiation of angiotensin II receptor antagonists, calcium channel blockers, and diuretic therapies ( Table 4 ). No significant differences were observed in the use of antihypertensives between the celecoxib and diclofenac treatment groups. Upward adjustments of antihypertensive or diuretic therapies already being administered to patients in order to maintain BP control were not considered in this analysis. Of the patients summarized in Table 4 , 59% (229/391) celecoxib-treated patients, 58% (107/184) diclofenac-treated patients, and 56% (133/238) ibuprofen-treated patients were already receiving antihypertensive or diuretic therapy at baseline and, therefore, were being initiated to a second agent concurrently or being switched from one treatment to another. For the remaining approximately 40% of patients in this cohort across the treatment groups, the initiation of antihypertensive or diuretic therapy occurred in the absence of any baseline treatment.

Fluid Retention

Celecoxib treatment was associated with a similar incidence of investigator-reported edema (including either peripheral or generalized edema) compared with diclofenac and a significantly lower incidence compared with ibuprofen (P<0.05) ( Table 5 ). This higher incidence of edema in patients taking ibuprofen paralleled a significantly greater use of new diuretic use compared with celecoxib-treated patients (P<0.05) ( Table 4 ). No significant differences were noted in diuretic use between patients assigned to celecoxib and diclofenac.

Adverse events of congestive heart failure were uncommonly reported by investigators (approximately one-tenth the rate of overall edema) and no significant differences were detected between treatments ( Table 5 ). In addition, no significant treatment differences were observed with respect to body weight gains of ≥3%.

Withdrawals from study treatment owing to either edema or congestive heart failure were uncommon overall and no significant treatment-related differences were detected.

Renal function

Renal function of enrolled patients was normal, as reflected by baseline serum creatinine levels or estimated creatinine clearance ( Table 1 ). Nearly all patients (>99%) had a serum creatinine ≤1.5 mg/dl, which reflects normal renal function required per protocol.

Renal function, as reflected by serum creatinine or estimated creatinine clearance, decreased significantly in patients assigned to diclofenac compared with those receiving celecoxib (P<0.05) ( Table 6 ). A nonsignificant trend towards decreased renal function was observed in patients treated with ibuprofen relative to those patients receiving celecoxib ( Table 6 ).

Clinically important renal dysfunction, defined as (1) increases in serum creatinine >0.5 mg/dl relative to baseline or serum creatinine >1.5 mg/dl at any post-baseline assessment or (2) a decrease in estimated creatinine clearance ≥30% relative to baseline are provided in Figure 2. For the entire treatment cohort, one significant treatment difference was detected: the incidence of ≥30% reductions in estimated creatinine clearance from baseline was significantly lower in patients treated with celecoxib as compared with diclofenac. An evaluation of the time course of these changes (data not shown) revealed that the highest frequency occurred early after initiation of treatment (week 4 and 13 clinic visits) and declined thereafter, rather than a progressively increasing risk as a function of duration of treatment in all three treatment groups. It should be noted, however, that a substantial proportion of patients exhibited these changes 6 months or longer after treatment was initiated, suggesting that the hazard, although perhaps diminishing with time, does continue throughout the therapy period. Increases in serum creatinine of >1.0 mg/dl were uncommon, occurring in nine celecoxib-treated patients (0.2%), two diclofenac-treated patients (0.1%), and two patients taking ibuprofen (0.2%). These treatment differences were not significantly different.

     

For those patients characterized by mild prerenal azotemia at baseline (serum blood urea nitrogen >20 mg/dl) but relatively normal renal function ( Table 7 ), the incidence of clinically important changes in renal function was at least twofold lower (P<0.05) in patients receiving celecoxib compared with either diclofenac- or ibuprofen-treated patients (Figure 2). Treatment differences in mean serum creatinine and estimated creatinine clearance were also observed in this cohort ( Table 7 ). No differences were observed between treatment groups for patients without baseline azotemia (i.e., with blood urea nitrogen ≤20 mg/dl) (Figure 2).

Patients with normal serum creatinine/prerenal azotemia at baseline were significantly more likely to withdraw from the study owing to an adverse event if their creatinine level increased to >1.5 mg/dl on treatment versus the overall cohort (relative risk, 2.0; P<0.05). When expressed as a percentage of the patients with normal serum creatinine/prerenal azotemia at baseline, the withdrawal rate in patients treated with celecoxib (1.5%) was significantly lower compared with those receiving diclofenac (3.7%; P<0.05) and ibuprofen (4.2%; P<0.05).

Electrolyte and Acid Base

Changes in mean serum levels of sodium, potassium, chloride, and bicarbonate were minimal between treatment groups and were not of clinical significance (data not shown).

Serious Cardiorenal-related Adverse Events

The incidence of serious cardiorenal-related adverse events during treatment with celecoxib, ibuprofen, or diclofenac was low. The study investigators reported one case of uremia in an ibuprofen-treated patient. Two cases of hyponatremia were reported in patients treated with celecoxib and one case was reported in a patient taking ibuprofen. No cases of serious adverse events related to nephrotic syndrome, interstitial nephritis, or papillary necrosis were observed.

Discussion

The large cardiorenal database from the CLASS trial has provided the opportunity to observe and compare the natural history of the onset and clinical management of cardiorenal-related adverse events that evolve when exposing a typical adult arthritic population to chronic treatment with a nonselective NSAID or supratherapeutic doses of the COXIB, celecoxib. Thus, the study provides a new and important clinical resource of therapeutic cardiorenal safety information. The data have permitted the description of a clinical paradigm wherein certain key patient characteristics may provide additional predictive clinical information with respect to future drug-specific susceptibility to impairment of cardiorenal function.

The present analysis supports the hypothesis that RA and OA patients receiving celecoxib 400 mg b.i.d. (two- to fourfold greater than the maximum effective doses for RA and OA) have an overall reduced risk of cardiorenal-related adverse events compared with standard therapeutic doses of the nonselective NSAIDs, diclofenac 75 mg b.i.d. and ibuprofen 800 mg three times a day. In particular, celecoxib treatment was superior to the nonselective NSAIDs with respect to renal function in patients with prerenal compromise. Some important differences were observed in the cardiorenal safety profiles of celecoxib versus nonselective NSAIDs and between the respective nonselective NSAIDs. Compared with celecoxib, the incidence of adverse events related to hypertension and edema were significantly higher with ibuprofen (but not diclofenac), and decline in renal function was more apparent with diclofenac than with ibuprofen. These data provide a greater understanding of the relative effects of these agents on cardiorenal function and homeostasis.

The increased incidence of hypertension in patients taking ibuprofen versus celecoxib appeared to be largely a function of effects on systolic, rather than diastolic, BP. Both diclofenac and ibuprofen were associated with significantly higher incidences of clinically meaningful elevations in systolic BP versus celecoxib. The importance of systolic BP has been highlighted in the Systolic Hypertension in Elderly Program trial,^[23] the Systolic hypertension in Europe trial^[24] and the Antihypertensive and Lipid-Lowering treatment to prevent Heart Attack Trial (ALLHAT).^[25] These studies showed that elevated systolic BP is associated with heart failure, stroke, myocardial infarction, and death. Evidence now indicates that destabilization of BP in treated hypertensive patients, by rofecoxib, is a significant risk marker for the development of acute myocardial infarction and stroke events.^[26] Much additional information with respect to the latter issues, and the emergence that some NSAIDs may also cause cardiotoxicity, is reviewed elsewhere.^[27]

Long-term treatment with supratherapeutic doses of celecoxib in the CLASS trial had little effect on BP levels in these arthritis patients. This finding is supported by previous studies demonstrating that celecoxib does not affect mean systolic BP changes in hypertensive patients treated with diuretics or antihypertensive agents.^{[28, 29, 30]} In contrast, many NSAIDs and some COXIBs, such as rofecoxib, antagonize the BP-lowering effects of certain antihypertensive agents, in particular, angiotensin-converting enzyme inhibitors and β-blockers.^{[31, 32, 33, 34]}

In the present analysis, celecoxib was superior to the nonselective NSAIDs studied with respect to renal dysfunction. Diclofenac, more so than ibuprofen, appeared to cause a significantly higher rate of decline in renal function than celecoxib. This may suggest a disproportionate effect of diclofenac on renal perfusion. Importantly, in patients with prerenal azotemia (but normal creatinine) at baseline, celecoxib was associated with a lower rate of deterioration of renal function compared with both diclofenac and ibuprofen. However, as some renal dysfunction was evident in the celecoxib treatment group, caution should still be used when treating patients at risk of adverse renal events. Changes in electrolyte levels were small and were not of clinical relevance.

Our results highlight that there are differential cardiorenal effects associated with celecoxib versus nonselective NSAIDs, and between the two NSAIDs, diclofenac and ibuprofen. Previous studies have also shown that cardiorenal safety profiles are not consistent across the class of COXIBs. For example, rofecoxib treatment is associated with dose-dependent increases in hypertension and edema.^{[28, 35]} In two recent, 6-week, randomized, double-blind trials of OA patients taking antihypertensive agents, the study participants taking celecoxib 200 mg daily were less likely to experience renal adverse events than the group receiving rofecoxib 25 mg daily.^[28] The differing renal safety profiles across the classes of nonselective NSAIDs and COXIBs suggest that some of the renal effects of the latter may be 'molecule-specific', rather than as a result of a 'class effect' of COX-2 inhibition.^[27]

Our results suggest that the clinical management of arthritis patients may be impacted by the agent selected to treat their pain and by patients' underlying health status with regard to cardiorenal side effects. Our results indicate that cardiorenal side effects are relatively common with NSAID treatment and require careful medical monitoring of patients together with other safety-related parameters (e.g., hepatotoxicity associated with some NSAIDs such as diclofenac). Anti-inflammatory drug-induced effects on BP or renal function should be cautiously monitored clinically during the use of such drugs. These adverse renal events may be selectively reduced or minimized by the selection of an anti-inflammatory drug, such as celecoxib, that appears to have the least detrimental effects upon renal function, as demonstrated in our study.

The present analysis has some potential limitations. Many OA and RA patients are elderly and have significant comorbidity, including cardiovascular disease, hypertension, and abnormal renal function. However, the CLASS trial did not include patients with abnormal renal function at baseline, and results cannot be applied to these patients. Thus, further work is needed to study the renal safety profile of nonspecific NSAIDs and COXIBs in patients with abnormal renal function. Another limitation in the present analysis is that patients in this study were not stratified by baseline body weight, which may impact significantly on baseline creatinine levels.

In conclusion, in a large adult arthritic population, supratherapeutic doses of celecoxib (400 mg b.i.d.) were associated with a reduced incidence of abnormal cardiorenal function, as assessed by investigator reports of hypertension, edema formation, or congestive heart failure, clinical measures of glomerular filtration, and various measures of destabilization of BP control, compared with standard doses of diclofenac 75 mg b.i.d. or ibuprofen 800 mg three times a day. In patients with mildly compromised renal function at baseline, celecoxib was associated with a lower rate of deterioration of renal function compared with both diclofenac and ibuprofen. These data suggest that celecoxib may frequently be a more suitable treatment of chronic pain and inflammation than nonselective NSAIDs in patients with compromised renal function.

Materialsand Methods

Study Population and Protocol

Details of the CLASS study population and protocol have been published previously.^{[22, 36]} In brief, outpatients ≥18 years of age were eligible to participate in the study if, upon screening, they were diagnosed with RA or OA evident for ≥3 months that required continuous treatment with an NSAID for the duration of the trial. Other inclusion and exclusion criteria were derived from the product label in order to mirror clinical practice.

Patients with significant renal disease that was considered by the investigator to be clinically significant were excluded from the CLASS population. The upper limit for serum creatinine was 1.5 mg/dl, which reflected the upper limit for normal renal function. Clinically important renal dysfunction was defined in the CLASS protocol as an increase in serum creatinine >0.5 mg/dl from baseline or a serum creatinine value above 1.5 mg/dl during double-blind treatment.

The CLASS trial was a randomized, double-blind, parallel-group study based on two protocols, the results of which were prospectively combined for analysis. The two separate protocols represented a comparison between celecoxib and a nonselective NSAID, ibuprofen or diclofenac. The study involved 386 centers in the United States and Canada from December 1998 to January 2000. Patients were randomized 1:1 to receive celecoxib 400 mg and a nonselective NSAID, ibuprofen 800 mg three times a day or diclofenac 75 mg b.i.d.

The concomitant use of stable doses of aspirin up to 325 mg daily was permitted during the trial. Patients who began taking aspirin during the course of the trial were not removed from the study and were included in all analyses. After a baseline visit, follow-up clinic visits took place at weeks 4, 13, 26, and every 13 weeks thereafter, following the initial dose of study medication until study termination. Some patients took study medication for up to 15 months. All patients were provided an opportunity to complete a minimum of 6 months of treatment ( Table 1 ).

Clinical Assessments

Adverse event data were collected by investigators at each clinic visit using the following question: 'Since your last visit, have you experienced or do you currently have any symptoms that are not associated with your arthritis?' All positive responses were recorded regardless of severity or relationship to study drug. Adverse events were judged to be 'serious' if they resulted in a fatality or hospitalization, or if the principal investigator felt that the event was life threatening or otherwise medically significant. Withdrawal rates owing to each adverse event were also recorded.

In the present analysis, all study data generated over a median exposure time of 9 months are presented for investigator-reported adverse events. These adverse events included new-onset or aggravated pre-existing hypertension, peripheral and generalized edema, and congestive heart failure. In addition, we analyzed the incidence of elevations in BP using the criteria for systolic BP of a >20 mmHg increase from baseline and absolute value >140 mmHg and for diastolic BP a >15 mmHg increase from baseline and absolute value >90 mmHg, consistent with a previous study.^[29] The introduction of new antihypertensive and diuretic medications was analyzed to assess the clinical relevance of observed effects on BP and edema. Body weight changes of at least ≥3% increase were also evaluated. The incidence of serum creatinine elevations was determined using the criteria >0.5 mg/dl increase over baseline or any value >1.5 mg/dl during treatment. The incidence of reductions in estimated creatinine clearance (as calculated by the Cockcroft-Gault formula: 140-age ?wt/[serum creatinine] ?72) of ≥30% from baseline during treatment was analyzed as well. The latter two analyses were also performed as a contingency analysis based on the presence of mild prerenal azotemia, defined as a baseline blood urea nitrogen level of >20 mg/dl.

Statistical Analysis

Statistical analyses were conducted on the ITT population, which was defined a priori in the original CLASS protocol as all patients who received at least one dose of assigned study medication. Baseline demographic characteristics and risk factors were compared using two-way analysis of variance or Pearson's χ². Incidences of treatment-emergent renovascular adverse events or clinical laboratory changes during the entire study were tabulated and compared using the Fisher exact test. Analyses of BP, serum creatinine and creatinine clearance were performed on changes from baseline to the final clinic visit for each patient.

	Celecoxib 400 mg b.i.d. (N=3987)	Diclofenac 75 mg b.i.d. (N=1996)	Ibuprofen 800 mg t.i.d. (N=1985)
Exposure to treatment
Median (days)	273	257	186
Total (patient-years)	2320	1081	1122
Age
Mean (years)	60.6	60.1	59.5
65?74 years (%)	26.9	26.4	25.5
≥75 years (%)	12.2	11.8	10.9
Female (%)	68.5	67.4	70.8
History (%)
Hypertension	39.0	38.9	38.1
Diabetes	8.8	7.5	8.1
Blood pressure (mmHg)
Systolic	132.7	133.0	132.6
Diastolic	79.4	79.5	79.9
Creatinine
Serum level (mg/dl)	0.79	0.79	0.77
Clearance (ml/mina)	112.2	112.0	116.4

b.i.d., twice a day; t.i.d., three times a day.
^aEstimated by the Cockcroft?Gault formula.

 

Drug N (% of patients)	Celecoxib 400 mg b.i.d. (N=3987)	Diclofenac 75 mg b.i.d. (N=1996)	Ibuprofen 800 mg t.i.d. (N=1985)
Any antihypertensive or diuretic medication	1688 (42.3)	859 (43.0)	845 (42.6)
ACE inhibitors	528 (13.2)	266 (13.3)	240 (12.1)
Alpha adrenergic agonists	209 (5.2)	95 (4.8)	87 (4.4)
Angiotensin II receptor antagonists	128 (3.2)	64 (3.2)	70 (3.5)
Beta-blockers	473 (11.9)	238 (11.9)	260 (13.1)
Calcium channel blockers	511 (12.8)	283 (14.2)	274 (13.8)
Diuretics	764 (19.2)	412 (20.6)	393 (19.8)

ACE, Angiotensin-converting enzyme; b.i.d., twice a day; t.i.d., three times a day.

 

Event	Celecoxib 400 mg b.i.d. (N=3987)	Diclofenac 75 mg b.i.d. (N=1996)	Ibuprofen 800 mg t.i.d. (N=1985)
Any hypertension adverse event (% of patients)a	2.7	2.6	4.2*
New-onset hypertension (% of patients)	2.0	2.0	3.1*
Aggravated hypertension (% of patients)	0.8	0.6	1.2
Withdrawals for hypertension-related adverse events (% of patients)a	0.3	0.2	0.3
Mean change in blood pressure (mmHg)
Systolic	?0.6	?0.8	0.3
Diastolic	?0.7	?1.1	?0.6
Increases in blood pressure (% of patients)b
Systolic >20 mmHg from baseline and absolute value >140 mmHg	5.0	6.6*	7.0*
Diastolic >15 mmHg from baseline and absolute value >90 mmHg	1.9	1.2	2.2

b.i.d., twice a day; t.i.d., three times a day.
^aNew-onset or aggravated.
^bObserved at final clinic visit.
^* P<0.05 versus celecoxib by Fisher exact test.

 

Drug N (% of patients)	Celecoxib 400 mg b.i.d. (N=3987)	Diclofenac 75 mg b.i.d. (N=1996)	Ibuprofen 800 mg t.i.d. (N=1985)
Any new antihypertensive or diuretic medicationa	391 (9.8)	184 (9.2)	238 (12.0)*
ACE inhibitors	108 (2.7)	63 (3.2)	62 (3.1)
Alpha adrenergic agonists	45 (1.1)	15 (0.8)	12 (0.6)
Angiotensin II receptor antagonists	34 (0.9)	17 (0.9)	39 (2.0)**
Beta-blockers	74 (1.9)	36 (1.8)	38 (1.9)
Calcium channel blockers	79 (2.0)	34 (1.7)	62 (3.1)**
Diuretics	153 (3.8)	84 (4.2)	103 (5.2)*

ACE, Angiotensin-converting enzyme; b.i.d., twice a day; t.i.d., three times a day.
^aIf a patient received more than one antihypertensive/diuretic medication, that patient is counted once in the overall incidence for any drug.
^*P<0.05 versus celecoxib by Fisher exact test.
^**P<0.01 versus celecoxib by Fisher exact test.

 

	Celecoxib 400 mg b.i.d. (N=3987)	Diclofenac 75 mg b.i.d. (N=1996)	Ibuprofen 800 mg t.i.d. (N=1985)
Any edema-related adverse event (% of patients)a	4.1	4.1	6.2*
Withdrawals for edema-related adverse events (% of patients)a	0.7	0.4	1.0
Congestive heart failure (% of patients)	0.3	0.2	0.5
Withdrawals due to congestive heart failure (% of patients)	0.1	<0.1	0.3
Increase in body weight of ≤3% (% of patients)	20.7	17.6	21.1

b.i.d., twice a day; t.i.d., three times a day.
^aIncludes investigator reports of edema, generalized edema or peripheral edema.
^*P<0.05 versus celecoxib by Fisher exact test.

 

	Celecoxib 400 mg b.i.d. (N=3987)	Diclofenac 75 mg b.i.d. (N=1996)	Ibuprofen 800 mg t.i.d. (N=1985)
Mean change in serum creatinine (mg/dl)	(n=3692) 0.009 ? 0.002	(n=1850) 0.027 ? 0.004*	(n=1786) 0.017 ? 0.004
Mean change in estimated creatinine clearance (ml/min)	(n=3668) 0.08 ? 0.37	(n=1846) -2.82 ? 0.51*	(n=1775) -0.96 ? 0.56

b.i.d., twice a day; t.i.d., three times a day.
^*P<0.05 versus celecoxib by analysis of covariance or by Fisher exact test.

 

	Celecoxib 400 mg b.i.d. (N=592)	Diclofenac 75 mg b.i.d. (N=300)	Ibuprofen 800 mg t.i.d. (N=260)
Baseline serum creatinine (mg/dl)	0.93	0.94	0.93
Estimated baseline creatinine clearance (ml/min)	91.1	91.3	94.3
Mean change in serum creatinine (mg/dl)	0.003 ? 0.007	0.049 ? 0.012*	0.033 ? 0.015*
Mean change in estimated creatinine clearance (ml/min)	1.27 ? 0.69	-1.65 ? 1.03*	-1.34 ? 1.28

b.i.d., twice a day; BUN, blood urea nitrogen; t.i.d., three times a day.
^*P<0.05 by analysis of covariance or by Fisher exact test.

 

Gierse JK, Hauser SD, Creely DP et al. Expression and selective inhibition of the constitutive and inducible forms of human cyclo-oxygenase. Biochem J 1995; 305(Part 2): 479-484. Verburg KM, Maziasz TJ, Weiner E et al. Cox-2-specific inhibitors: definition of a new therapeutic concept. Am J Ther 2001; 8: 49-64. Gabriel SE, Jaakkimainen L, Bombardier C. Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs. A meta-analysis. Ann Intern Med 1991; 115: 787-796. Henry D, Lim LL, Garcia Rodriguez LA et al. Variability in risk of gastrointestinal complications with individual non-steroidal anti-inflammatory drugs: results of a collaborative meta-analysis. BMJ 1996; 312: 1563-1566. Griffin MR, Piper JM, Daugherty JR et al. Nonsteroidal anti-inflammatory drug use and increased risk for peptic ulcer disease in elderly persons. Ann Intern Med 1991; 114: 257-263. Langman MJ, Weil J, Wainwright P et al. Risks of bleeding peptic ulcer associated with individual non-steroidal anti-inflammatory drugs. Lancet 1994; 343: 1075-1078. Bensen WG, Fiechtner JJ, McMillen JI et al. Treatment of osteoarthritis with celecoxib, a cyclooxygenase-2 inhibitor: a randomized controlled trial. Mayo Clin Proc 1999; 74: 1095-1105. Geis GS. Update on clinical developments with celecoxib, a new specific COX-2 inhibitor: what can we expect? J Rheumatol 1999; 26(Suppl 56): 31-36. Leese PT, Hubbard RC, Karim A et al. Effects of celecoxib, a novel cyclooxygenase-2 inhibitor, on platelet function in healthy adults: a randomized, controlled trial. J Clin Pharmacol 2000; 40: 124-132. Goldenberg MM. Celecoxib, a selective cyclooxygenase-2 inhibitor for the treatment of rheumatoid arthritis and osteoarthritis. Clin Ther 1999; 21: 1497-1513. Simon LS. Nonsteroidal anti-inflammatory drugs and their effects: the importance of COX selectivity. J Clin Rheumatol 1996; 2: 135-140. 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. 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. Whelton A. Nephrotoxicity of nonsteroidal anti-inflammatory drugs: physiologic foundations and clinical implications. Am J Med 1999; 106: 13S-24S. Atta MG, Whelton A. Acute renal papillary necrosis induced by ibuprofen. Am J Ther 1997; 4: 55-60. Whelton A, Hamilton CW. Nonsteroidal anti-inflammatory drugs: effects on kidney function. J Clin Pharmacol 1991; 31: 588-598. Harris RC, Breyer MD. Physiological regulation of cyclooxygenase-2 in the kidney. Am J Physiol Renal Physiol 2001; 281: F1-F11. Hao CM, Yull F, Blackwell T et al. Dehydration activates an NF-kappaB-driven COX2-dependent survival mechanism in renal medullary interstitial cells. J Clin Invest 2000; 106: 973-982. 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. 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. Harris Jr RC. Cyclooxygenase-2 inhibition and renal physiology. Am J Cardiol 2002; 89: 10D-17D. White WB, Faich G, Whelton A et al. Comparison of thromboembolic events in patients treated with celecoxib, a cyclooxygenase-2 specific inhibitor, versus ibuprofen or diclofenac. Am J Cardiol 2002; 89: 425-430. Kostis JB, Davis BR, Cutler J et al. Prevention of heart failure by antihypertensive drug treatment in older persons with isolated systolic hypertension. SHEP Cooperative Research Group. JAMA 1997; 278: 212-216. Staessen JA, Fagard R, Thijs L et al. Randomised double-blind comparison of placebo and active treatment for older patients with isolated systolic hypertension. The systolic hypertension in Europe (Syst-Eur) trial investigators. Lancet 1997; 350: 757-764. ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: The antihypertensive and lipid-lowering treatment to prevent heart attack trial (ALLHAT). JAMA 2002; 288: 2981-2997. Whelton A, Spalding W, White WB et al. Rofecoxib increases cardiovascular events in arthritis patients but celecoxib and nonspecific nonsteroidal anti-inflammatory drugs do not. J Am Coll Cardiol 2004; 5(Suppl A): 415A-4143. Whelton A. The coxib conundrum: lessons from the rise and fall of rofecoxib. Am J Ther 2004; 11: 417-421. 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. Whelton A, White WB, Bello AE et al. Effects of celecoxib and rofecoxib on blood pressure and edema in patients > or = 65 years of age with systemic hypertension and osteoarthritis. Am J Cardiol 2002; 90: 959-963. White WB, Kent J, Taylor A et al. Effects of celecoxib on ambulatory blood pressure in hypertensive patients on ACE inhibitors. Hypertension 2002; 39: 929-934. Oates JA. Antagonism of antihypertensive drug therapy by nonsteroidal anti-inflammatory drugs. Hypertension 1988; 11: II4-II6. Houston MC. Nonsteroidal anti-inflammatory drugs and antihypertensives. Am J Med 1991; 90: 42S-47S. Johnson AG, Nguyen TV, Day RO. Do nonsteroidal anti-inflammatory drugs affect blood pressure? A meta-analysis. Ann Intern Med 1994; 121: 289-300. de Leeuw PW. Nonsteroidal anti-inflammatory drugs and hypertension. The risks in perspective. Drugs 1996; 51: 179-187. VIOXX?[package insert]. Whitehouse Station, NJ: MERCK & Co., Inc., 2002. 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.

Acknowledgements

During the preparation of this manuscript our friend and colleague Dr Lefkowith died unexpectedly. We miss him greatly and we pray that he may rest in peace. Andrew Whelton, MD, serves as a consultant to Pfizer Inc.

Reprint Address

A Whelton, The Universal Clinical Research Center Inc., 1737 Beaver Brook Lane, Hunt Valley, Maryland 21030-1603, USA. E-mail: huntvalley@aol.com