EVIDENCE-BASED ANSWER
No, COX-2 inhibitors, as a class, do not worsen renal function for those without renal disease. Celecoxib is the only COX-2 inhibitor available, and it is associated with a lower risk of renal dysfunction and hypertension when compared with controls. Available data do not allow for adjusted risk assessment for patients with preexisting renal disease on COX-2 inhibitors (strength of recommendation [SOR]: A, based on meta-analysis).
Clinical commentary
Use celecoxib cautiously in patients at risk of serious complications
Vincent LO, MD
San Joaquin Family Medicine Residency, French Camp, Calif
Recent studies have raised concerns about the safety of this class of medication. For example, rofecoxib was linked with increased cardiovascular events, leading to it being pulled from the market.1 The claim of decreased gastrointestinal bleeding with long-term use of COX-2 inhibitors has also been questioned.2
Although this Clinical Inquiry concludes that celecoxib does not appear to worsen renal function, it should still be used with caution for patients who are elderly, hospitalized, or at risk of developing serious complications such as acute renal failure, heart failure, and gastrointestinal bleeding.
Evidence summary
A 2006 meta-analysis, including 114 trials and 116,094 patients randomized to either cyclooxygenase-2 (COX-2) inhibitor or control (placebo, nonsteroidal anti-inflammatory drug [NSAID], or mixed), indicated that the COX-2 inhibitors, as a class, had no effect on renal endpoints.3 Trials were reviewed for data on renal endpoints, including peripheral edema, hypertension, and renal dysfunction (defined as significant worsening of serum urea or creatinine, or clinical evidence of kidney disease and renal failure).
When viewed separately, rofecoxib (Vioxx) was associated with a composite relative risk (RR) of 1.53 (95% confidence interval [CI], 1.33–1.76) for all renal endpoints compared with controls. In contrast, the composite RR for the same endpoints among patients taking celecoxib (Celebrex) was 0.97 (95% CI, 0.84–1.12), indicating no effect on renal function. In fact, for the specific outcomes of hypertension and renal dysfunction, celecoxib was associated with a decreased risk compared with controls (TABLE).3
Stratified analysis by type of control (placebo, alternate NSAID, or mixed) yielded consistent results; rofecoxib was uniquely associated with adverse renal outcomes. No effect on renal function was noted for celecoxib compared with the same controls: the RR for adverse renal effects was 0.87 (95% CI, 0.55–1.38), 0.93 (95% CI, 0.70–1.23), and 1.26 (95% CI, 0.94–1.69) for celecoxib vs placebo, NSAID, and mixed controls, respectively. Statistical analysis for heterogeneity showed that the variation in effects on renal function among the COX-2 inhibitors was more likely due to actual differences than due to chance (heterogeneity [I2]=57%; P<.001).
Data were not available to assess the effect of COX-2 agents on patients with pre-existing renal disease, primarily because trials reporting abnormal renal function at baseline were excluded from this meta-analysis.
A recent randomized controlled trial compared standard dosing of diclofenac (75 mg twice daily) and ibuprofen (800 mg 3 times daily) with high-dose celecoxib (400 mg twice daily) for patients with normal kidney function being treated for osteoarthritis and rheumatoid arthritis.4 The mean increase in serum creatinine in the celecoxib arm was less than that noted in the diclofenac controls (0.009 mg/dL vs 0.027 mg/dL; P<.05; number needed to harm [NNH]=56). No difference in mean serum creatinine was seen among those patients using ibuprofen (800 mg 3 times daily) compared with those using high-dose celecoxib.
This evidence further supports the safety of celecoxib vs standard NSAIDs with respect to renal dysfunction.