Interstitial nephritis associated with celecoxib

Abstract

Jose Rocha and Jorge Fernández-Alonso (June 16, p 1946)1Rocha JL Fernandez-Alonso J Acute tubulointerstitial nephritis associated with the selective COX-2 enzyme inhibitor, rofecoxib.Lancet. 2001; 357: 1946-1947Summary Full Text Full Text PDF PubMed Scopus (82) Google Scholar referred to 11 cases of possible interstitial nephritis associated with celecoxib. No case of interstitial nephritis was reported in the preregistration clinical trials. From celecoxib's commercial availability in 1999, to July 16, 2001, we have received reports of 16 cases with a possible diagnosis of acute interstitial nephritis in patients treated with celecoxib. All have been reported to regulatory authorities worldwide. Two patients had alternative causes for their renal symptomatology (diabetic nephrosclerosis and glomerulonephritis). Only two cases had biopsy evidence, although the reports were not provided to us. In the remaining 12 cases there was no confirmation of the diagnosis. The incidence of tubulointerstitial nephritis in the general population is unknown, but it has been estimated from a large series of biopsy-proven cases that acute tubulointerstitial nephritis accounts for 3–10% of cases of acute renal failure.2Davison AM Jones CH Acute interstitial nephritis in the elderly: a report from the MRC glomerulonephritis register and a review of the literature.Nephrol Dial Transplant. 1998; 13: 12-16sCrossref PubMed Scopus (45) Google Scholar, 3Toto RD Review: acute tubulointerstitial nephritis.Am J Med Sci. 1990; 3299: 392-410Crossref Scopus (54) Google Scholar The incidence of acute renal failure in adults is 172 per 1000 000 persons-years.4Feest TG Round A Hamad S Incidence of severe acute renal failure in adults: results of a community based study.BMJ. 1993; 306: 481-483Crossref PubMed Scopus (235) Google Scholar If we assume 3–10% of these acute renal failure cases as acute tubulointerstitial nephritis, the incidence rate of tubulointerstitial nephritis can be estimated to be 5·2–17·2 per 1 000 000 person-years.4Feest TG Round A Hamad S Incidence of severe acute renal failure in adults: results of a community based study.BMJ. 1993; 306: 481-483Crossref PubMed Scopus (235) Google Scholar The worldwide exposure to celecoxib is estimated as 10·6 million patient-years, or around 21·5 million patients, assuming an average treatment period of 6 months. Therefore, the number of expected cases of acute tubulointerstitial nephritis cases ranges from 55 to 183. The receipt of cases is affected by an unknown degree of under-reporting inherent to the spontaneous reporting system. However, the 16 cases reported to us fall within what could be expected given the widespread use of the drug. This expected number is a conservative estimate because, by definition, it does not include those patients with interstitial nephritis who do not develop acute renal failure. An important proportion of acute tubulointerstitial nephritis is reportedly drug induced.2Davison AM Jones CH Acute interstitial nephritis in the elderly: a report from the MRC glomerulonephritis register and a review of the literature.Nephrol Dial Transplant. 1998; 13: 12-16sCrossref PubMed Scopus (45) Google Scholar, 3Toto RD Review: acute tubulointerstitial nephritis.Am J Med Sci. 1990; 3299: 392-410Crossref Scopus (54) Google Scholar We did an analysis of the WHO Collaborating Program for International Drug Monitoring database.5Bate A Lindquist M Edwards IR et al.A bayesian neural network method for adverse drug reaction signal generation.Eur J Clin Pharmacol. 1998; 954: 315-321Crossref Scopus (649) Google Scholar This database contains more than 2 million reports from 57 countries and is the largest database of its kind worldwide. The inhibitory concentration value (index used to measure the strength of the adverse drug reaction-drug pair association5Bate A Lindquist M Edwards IR et al.A bayesian neural network method for adverse drug reaction signal generation.Eur J Clin Pharmacol. 1998; 954: 315-321Crossref Scopus (649) Google Scholar) for interstitial nephritis for celecoxib does not differ significantly from background (expectation derived from all other drugs; p>0·05). However, the inhibitory concentrations of interstitial nephritis for rofecoxib, diclofenac, and ibuprofen are all significantly higher than background (p<0·05). The interstitial nephritis values for ibuprofen and diclofenac are also significantly higher than for celecoxib, but not that for rofecoxib. We cannot comment on the case report described by Rocha and Fernández-Alonso. We agree that the nephrotoxic potential of various cyclo-oxygenase-2 inhibitors needs to be studied further. There seems to be no evidence for a causal relation between tubulointerstitial nephritis and celecoxib, and differences between the same class of drugs need to be appreciated. Absolute numbers of spontaneous reports should not be overemphasised when making causal inferences without taking into consideration drug exposure. All authors work for Pharmacia Corp, who manufacture celecoxib.