Risks of clinically significant upper gastrointestinal events with etodolac and naproxen: A historical cohort analysis
2004; Elsevier BV; Volume: 127; Issue: 5 Linguagem: Inglês
10.1053/j.gastro.2004.08.016
ISSN1528-0012
AutoresRick Weideman, Kevin C. Kelly, Salahuddin Kazi, Anh Cung, Kevin Roberts, Herbert J. Smith, George A. Sarosi, Bertis B. Little, Byron Cryer,
Tópico(s)Analytical Methods in Pharmaceuticals
ResumoBackground & Aims: Etodolac is a generic nonsteroidal anti-inflammatory drug (NSAID). Previous in vitro studies have shown that etodolac is a selective inhibitor of cyclooxygenase (COX)-2 with selectivity in between that of other COX-2 inhibitors such as celecoxib and rofecoxib. However, there are no outcomes data assessing clinically significant upper gastrointestinal (CSUGI) events with etodolac. Methods: A historical cohort study was performed at the Dallas Veterans Affairs Medical Center in which 16,286 veteran patients (5596 patient-years) received etodolac or naproxen during a 3-year period without concurrent use of other ulcerogenic drugs other than low-dose aspirin. The primary outcome was the CSUGI event rate of the etodolac and naproxen groups without concomitant low-dose aspirin. Results: The incidence of CSUGI events was .78% and .24% for naproxen and etodolac, respectively. In the NSAID-naive subset, the incidence of CSUGI events was .99% and .24% for naproxen and etodolac, respectively. Compared with naproxen, etodolac was associated with a reduction in upper gastrointestinal events, corresponding to an odds ratio of .39 (95% confidence interval, .20–.76; P = .006). Concomitantly used low-dose aspirin increased event rates with naproxen 2-fold and etodolac 9-fold. Hence, there was no significant difference in gastrointestinal event rates between etodolac and naproxen when low-dose aspirin was taken concomitantly. Conclusions: Etodolac is a generic COX-2 selective inhibitor that reduces CSUGI events compared with the nonselective NSAID naproxen. However, concomitant use of low-dose aspirin negates the gastrointestinal safety advantages of etodolac. Background & Aims: Etodolac is a generic nonsteroidal anti-inflammatory drug (NSAID). Previous in vitro studies have shown that etodolac is a selective inhibitor of cyclooxygenase (COX)-2 with selectivity in between that of other COX-2 inhibitors such as celecoxib and rofecoxib. However, there are no outcomes data assessing clinically significant upper gastrointestinal (CSUGI) events with etodolac. Methods: A historical cohort study was performed at the Dallas Veterans Affairs Medical Center in which 16,286 veteran patients (5596 patient-years) received etodolac or naproxen during a 3-year period without concurrent use of other ulcerogenic drugs other than low-dose aspirin. The primary outcome was the CSUGI event rate of the etodolac and naproxen groups without concomitant low-dose aspirin. Results: The incidence of CSUGI events was .78% and .24% for naproxen and etodolac, respectively. In the NSAID-naive subset, the incidence of CSUGI events was .99% and .24% for naproxen and etodolac, respectively. Compared with naproxen, etodolac was associated with a reduction in upper gastrointestinal events, corresponding to an odds ratio of .39 (95% confidence interval, .20–.76; P = .006). Concomitantly used low-dose aspirin increased event rates with naproxen 2-fold and etodolac 9-fold. Hence, there was no significant difference in gastrointestinal event rates between etodolac and naproxen when low-dose aspirin was taken concomitantly. Conclusions: Etodolac is a generic COX-2 selective inhibitor that reduces CSUGI events compared with the nonselective NSAID naproxen. However, concomitant use of low-dose aspirin negates the gastrointestinal safety advantages of etodolac. Traditional nonsteroidal anti-inflammatory drugs (NSAIDs) are associated with untoward gastrointestinal (GI) consequences such as upper GI bleeding. Across all drug classes, approximately 25% of reported adverse drug reactions are attributable to NSAIDs.1Wolfe M.M. Lichtenstein D.R. Singh G. Gastrointestinal toxicity of nonsteroidal anti-inflammatory drugs.N Engl J Med. 1999; 340: 1888-1899Crossref PubMed Scopus (1911) Google Scholar Cyclooxygenase (COX)-2-specific inhibitors were developed primarily to provide anti-inflammatory and analgesic therapy with improved GI safety profiles. Older, traditional NSAIDs (eg, naproxen and aspirin) inhibit COX-1 and COX-2. Because protective GI mucosal prostaglandins are derived mainly from the COX-1 isoform, it was reasoned that COX-2-specific inhibitors should be associated with a lower incidence of clinically significant upper GI (CSUGI) events. COX-2-specific inhibitors (eg, celecoxib and rofecoxib) were thus developed and introduced into clinical practice in the United States in early 1999.Celecoxib, when compared with nonselective NSAIDs, showed a numeric reduction in rates of clinically important GI events over 6 months, although the primary end point did not reach statistical significance.2Silverstein F.E. Faich G. Goldstein J.L. Simon L.S. Pincus T. Whelton A. Makuch R. Eisen G. Agrawal N.M. Stenson W.F. Burr A.M. Zhao W.W. Kent J.D. Lefkowith J.B. Verburg K.M. Geis G.S. Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis.JAMA. 2000; 284: 1247-1255Crossref PubMed Scopus (3076) Google Scholar Compared with naproxen, rofecoxib was associated with a significant reduction in upper GI events over a 13-month period.3Bombardier C. Laine L. Reicin A. Shapiro D. Burgos-Vargas R. Davis B. Day R. Bosi-Ferraz Hawkey C.J. Hochberg M. Kvien T.K. Schnitzer T.J. VIGOR Study GroupComparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.N Engl J Med. 2000; 343: 1520-1528Crossref PubMed Scopus (3700) Google Scholar Consequently, it is now accepted that NSAIDs with a high degree of COX-2 selectivity are associated with a decreased frequency of upper GI complications.Etodolac, another NSAID, was first approved by the US Food and Drug Administration for clinical use in 1991. In vitro analyses have shown that etodolac has a high degree of COX-2 selectivity, with COX-2 selectivity between that of celecoxib and rofecoxib.4Warner T.D. Giuliano F. Vojnovic I. Bukasa A. Mitchell J.A. Vane J.R. Nonsteroid drug selectivities for cyclo-oxygenase-1 rather than cyclo-oxygenase-2 are associated with human gastrointestinal toxicity a full in vitro analysis.Proc Natl Acad Sci U S A. 1999; 96: 7563-7568Crossref PubMed Scopus (1410) Google Scholar In a 4-week study of naproxen compared with etodolac, naproxen inhibited protective gastric prostaglandins by approximately 70% of baseline values and was associated with a much greater incidence of endoscopic gastric ulceration.5Laine L. Sloane R. Ferretti M. Cominelli F. A randomized, double-blind comparison of placebo, etodolac, and naproxen on gastrointestinal injury and prostaglandin production.Gastrointest Endosc. 1995; 42: 428-433Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar In contrast, etodolac was associated with significantly fewer gastric lesions and did not significantly affect gastric COX-1 or its prostaglandin derivatives.5Laine L. Sloane R. Ferretti M. Cominelli F. A randomized, double-blind comparison of placebo, etodolac, and naproxen on gastrointestinal injury and prostaglandin production.Gastrointest Endosc. 1995; 42: 428-433Abstract Full Text Full Text PDF PubMed Scopus (76) Google ScholarThese prior investigations suggested that etodolac might be safer than other NSAIDs such as naproxen in the upper GI tract. However, study end points (eg, short-term endoscopic ulceration) in prior investigations were only intermediate markers of NSAID-associated GI toxicity. No published investigations of CSUGI outcomes associated with etodolac are available. Therefore, a historical cohort outcomes study was designed to analyze rates of CSUGI events associated with etodolac compared with naproxen in our local Veterans Affairs health care system.Patients and methodsStudy design and populationThe institutional review board of the Dallas Veterans Affairs Medical Center approved this study. This historical cohort study was designed to parallel the methodology used in previously reported retrospective cohort GI outcome studies of COX-2 selective inhibitors.6Langman M. Jensen D. Watson D. Harper S.E. Zhao P.L. Quan H. Bolognese J.A. Simon T.J. Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs.JAMA. 1999; 282: 1929-1933Crossref PubMed Scopus (672) Google Scholar, 7Goldstein J.L. Silverstein F.E. Agrawal N.M. Hubbard R.C. Kaiser J. Maurath C.J. Verburg K.N. Geis G.S. Reduced risk of upper gastrointestinal ulcers with celecoxib, a COX-2 inhibitor.Am J Gastroenterol. 2000; 95: 1681-1690Crossref PubMed Google Scholar Study subjects were patients at the Dallas Veterans Affairs Medical Center who received naproxen or etodolac in doses typically used in clinical studies of NSAID-associated GI toxicity.8Laine L. Approaches to nonsteroidal anti-inflammatory drug use in the high-risk patient.Gastroenterology. 2001; 120: 594-606Abstract Full Text Full Text PDF PubMed Scopus (473) Google Scholar Eligible patients were those who (1) received outpatient prescriptions for etodolac ≥800 mg/day or naproxen ≥1000 mg/day between January 1, 1999, and December 31, 2001; (2) were not concurrently taking any NSAIDs other than low-dose aspirin (≤325 mg); (3) were not concurrently taking any confounding agent such as warfarin, ticlopidine, clopidogrel, extended-release dipyridamole/aspirin (Aggrenox; Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, CT), corticosteroids, proton pump inhibitors, misoprostol, or ≥40 mg famotidine twice daily. Eligibility for study evaluation was discontinued when a patient developed a CSUGI event (defined in the following text), discontinued the study drug, or died. Outcomes were assessed up to December 31, 2001.Assessment of use of low-dose aspirinAspirin is an important preventative medication for cardiovascular disease in our patient population, and its acquisition cost is negligible to the Dallas Veterans Affairs Medical Center. Consequently, aspirin is provided by prescription to our patients at no charge. Thus, there is a strong incentive for patients to obtain their aspirin as a prescription from the Dallas Veterans Affairs Medical Center pharmacy. Our computerized pharmacy records contain an extensive compilation of prescribed low-dose aspirin. Our ability to quantitatively assess use of aspirin in our computerized databases is as good as, or better than, our ability to track use of other prescription medications for which a co-payment is charged.Definition of study patient groups and study end pointsPatient-days of cohort membership were determined by totaling days of etodolac ≥800 mg/day or naproxen ≥1000 mg/day (date prescription filled through last day's supply9Ray W.A. Stein C.M. Hall K. Daugherty J.R. Griffin M.R. Non-steroidal anti-inflammatory drugs and risk of serious coronary heart disease an observational cohort study.Lancet. 2002; 359: 118-123Abstract Full Text Full Text PDF PubMed Scopus (322) Google Scholar or until December 31, 2001, whichever date came first). Also, days of concurrent use of a confounding agent were censored for the NSAID-exposure calculation. For example, if a patient started naproxen on June 1, 1999, and continued taking it until January 1, 2000, all days of naproxen exposure were counted. However, if that same patient received a prescription for a proton pump inhibitor (or any of the other confounding medications) from September 1, 1999, to December 31, 1999, all days in which the proton pump inhibitor was concomitantly taken with naproxen would have been censored from this patient's exposure.The primary prespecified study end point was the development of a CSUGI event during the study period among etodolac- or naproxen-exposed patients not concurrently taking low-dose aspirin using predetermined criteria (Table 1).2Silverstein F.E. Faich G. Goldstein J.L. Simon L.S. Pincus T. Whelton A. Makuch R. Eisen G. Agrawal N.M. Stenson W.F. Burr A.M. Zhao W.W. Kent J.D. Lefkowith J.B. Verburg K.M. Geis G.S. Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis.JAMA. 2000; 284: 1247-1255Crossref PubMed Scopus (3076) Google Scholar, 3Bombardier C. Laine L. Reicin A. Shapiro D. Burgos-Vargas R. Davis B. Day R. Bosi-Ferraz Hawkey C.J. Hochberg M. Kvien T.K. Schnitzer T.J. VIGOR Study GroupComparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.N Engl J Med. 2000; 343: 1520-1528Crossref PubMed Scopus (3700) Google Scholar To be adequately powered, our study would require 5000 patients in each arm to detect a 50% reduction in the incidence of CSUGI events.Table 1Diagnostic Criteria for GI EventsEventConfirmationGastric or duodenal perforationPerforated lesion requiring surgery. Could involve a laparoscopic repair, but only if evidence of the perforation was unequivocal, such as free air in the abdomen visible on radiography or direct visualization at surgeryGastric outlet obstructionGastric outlet obstruction requiring diagnosis by investigator; diagnosis was required to be supported by endoscopy (eg, ulcer with a tight edematous pyloric channel) or by radiographic results (eg, dilated stomach, delayed barium emptying with clinical evidence of outlet obstruction and with an ulcer in the channel, severe outlet narrowing, and edema)Upper GI bleeding (at least one criterion present)• Hematemesis with a lesion (ulcer or large erosionaUlcer = mucosal break ≥5 mm; erosion = mucosal break <3 mm; large erosion = mucosal break measuring 3 or 4 mm.) on endoscopy or radiography• Hematemesis or aspiration of bloody gastric fluid witnessed by health care provider; episode of melena witnessed by health care provider• Lesion (ulcer or large erosionaUlcer = mucosal break ≥5 mm; erosion = mucosal break <3 mm; large erosion = mucosal break measuring 3 or 4 mm.) on endoscopy with evidence of active bleeding or stigmata of recent hemorrhage (visible vessel or clot attached to the base of an ulcer)• Melena with a lesion (ulcer or large erosionaUlcer = mucosal break ≥5 mm; erosion = mucosal break <3 mm; large erosion = mucosal break measuring 3 or 4 mm.) on endoscopy or radiography• Occult blood-positive stool with a lesion (ulcer or large erosionaUlcer = mucosal break ≥5 mm; erosion = mucosal break 15 g/L 2. Postural vital sign changes (increase in heart rate of ≥20/min and/or decrease in systolic blood pressure of ≥20 mm Hg and/or in diastolic blood pressure of ≥10 mm Hg) 3. Transfusion of ≥2 units of blood 4. Blood in stomach on endoscopy or nasogastric aspirationSymptomatic gastric or duodenal ulcer• Evidence of gastric or duodenal ulceraUlcer = mucosal break ≥5 mm; erosion = mucosal break <3 mm; large erosion = mucosal break measuring 3 or 4 mm. on endoscopy, at surgery, on contrast-enhanced radiography of the upper gastrointestinal tract, or at autopsy, or• Abdominal pain or dyspepsia as patient indication followed by finding of gastric or duodenal ulcerNOTE. Table derived from CLASS2Silverstein F.E. Faich G. Goldstein J.L. Simon L.S. Pincus T. Whelton A. Makuch R. Eisen G. Agrawal N.M. Stenson W.F. Burr A.M. Zhao W.W. Kent J.D. Lefkowith J.B. Verburg K.M. Geis G.S. Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis.JAMA. 2000; 284: 1247-1255Crossref PubMed Scopus (3076) Google Scholar and VIGOR3Bombardier C. Laine L. Reicin A. Shapiro D. Burgos-Vargas R. Davis B. Day R. Bosi-Ferraz Hawkey C.J. Hochberg M. Kvien T.K. Schnitzer T.J. VIGOR Study GroupComparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.N Engl J Med. 2000; 343: 1520-1528Crossref PubMed Scopus (3700) Google Scholar studies.a Ulcer = mucosal break ≥5 mm; erosion = mucosal break <3 mm; large erosion = mucosal break measuring 3 or 4 mm. Open table in a new tab The risk of upper GI events differs between patients not recently exposed to an NSAID and those continuously receiving or recently switched from another NSAID.8Laine L. Approaches to nonsteroidal anti-inflammatory drug use in the high-risk patient.Gastroenterology. 2001; 120: 594-606Abstract Full Text Full Text PDF PubMed Scopus (473) Google Scholar, 10Gabriel S.E. Jaakkimainen L. Bombardier C. Risk for serious gastrointestinal complications related to the use of nonsteroidal anti-inflammatory drugs. A meta-analysis.Ann Intern Med. 1991; 115: 787-796Crossref PubMed Scopus (1226) Google Scholar Therefore, one secondary end point was CSUGI events among NSAID-naive patients, the subset of patients who had no exposure to other NSAIDs in the 120 days before starting treatment with etodolac or naproxen. The other prespecified secondary end point was CSUGI events in patients taking concomitant low-dose aspirin with the study drugs for both the NSAID-naive and the entire cohort (Figure 1).Monitoring for upper GI eventsPotential CSUGI events were identified and confirmed in blinded fashion. Identification of CSUGI events was accomplished by a review of each patient record in the electronic computerized patient record system, a computerized GI endoscopic database (cMore; ProVation Medical Inc, Minneapolis, MN), a computerized radiologic examinations database, and a computerized surgical database in the Dallas Veterans Affairs Medical Center. The study initiation date (January 1, 1999) was based on mandatory use of all of these electronic databases for patient care in our health care system as of that date. CSUGI events were attributed to etodolac or naproxen exposure only if the event, or the request for the procedure that documented the event, occurred within 30 days following the patient's final study drug prescription. All CSUGI events were validated blindly (without knowledge of NSAID exposure) by a gastroenterologist (B.C.), a radiologist (H.J.S.), and a general surgeon (G.A.S.) through a review, as appropriate, of the original endoscopic images, endoscopic reports, radiographic images, surgical operative reports, and computerized records.Data reporting and statistical analysesBaseline patient demographics in the etodolac and naproxen groups were analyzed using χ2 tests for categorical data and t tests for continuous data. Analyses of event rates by length of exposure were calculated based on number of events per patient-years of exposure to etodolac or naproxen. A patient-year was defined as 365 patient-days of exposure. A patient-day was defined as a day in which a patient received a prescription for at least one dose of study drug not confounded by the presence of other drugs. Event rates for the etodolac and naproxen groups were calculated using rates per 100 patient-years over 12 months and reported as cumulative incidence percentages. Time-to-event curves were constructed for etodolac and naproxen in patients not taking aspirin concurrently based on cumulative incidences of CSUGI events. This method is appropriate because it considers varying lengths of exposure to NSAIDs.3Bombardier C. Laine L. Reicin A. Shapiro D. Burgos-Vargas R. Davis B. Day R. Bosi-Ferraz Hawkey C.J. Hochberg M. Kvien T.K. Schnitzer T.J. VIGOR Study GroupComparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.N Engl J Med. 2000; 343: 1520-1528Crossref PubMed Scopus (3700) Google Scholar, 6Langman M. Jensen D. Watson D. Harper S.E. Zhao P.L. Quan H. Bolognese J.A. Simon T.J. Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs.JAMA. 1999; 282: 1929-1933Crossref PubMed Scopus (672) Google ScholarFisher exact test was used to compare the annualized incidence of CSUGI events in the etodolac and naproxen groups. The same comparisons and tests were used in the etodolac and naproxen groups with and without concurrent exposure to low-dose aspirin. Differences between groups were considered statistically significant for P values ≤.05. Comparative risks of CSUGI events are expressed as odds ratios and corresponding 95% confidence intervals (CIs) computed by logistic regression. Stepwise logistic regression analysis controlled for the effects of age, sex, history of ulcer, NSAID (etodolac or naproxen), NSAID exposure (patient-years), concurrent aspirin exposure, rheumatoid arthritis, congestive heart failure, diabetes, and prior NSAID exposure (prior NSAID exposure versus NSAID-naive patients). To obtain odds ratios and confidence limits for variables that were nonsignificant, a direct logistic regression was performed. SAS statistical analysis software version 8.2 (SAS Institute, Cary, NC) and Primer of Biostatistics version 4.02 (McGraw Hill, New York, NY) were used for all analyses.ResultsIdentification of study populationDuring the study interval (January 1, 1999, to December 31, 2001), a total of 26,556 patients received prescriptions for etodolac or naproxen. After exclusion of patients taking confounding drugs (eg, clopidogrel, misoprostol, proton pump inhibitors, and so on), our study group consisted of 16,286 patients. Of these, 9607 patients were treated with etodolac (2600 total patient-years of exposure), 8382 patients were treated with naproxen (2996 total patient-years of exposure), and 1703 patients received naproxen and etodolac at different times during the 3-year study interval. The average daily doses of etodolac and naproxen were 885 mg and 1054 mg, respectively.Demographics of study populationOur study population had a mean age of 56.4 years (range, 18–100 years) and was 89.5% male. Baseline characteristics of the 2 groups were similar (Table 2). However, due to the large sample sizes, statistical comparisons of the etodolac and naproxen groups showed significant differences in age, sex, use of low-dose aspirin, and history of ulcer. Thus, when potential differences in GI outcomes between etodolac and naproxen were assessed by logistic regression, baseline characteristics were variables that were entered into the adjusted analysis.Table 2Baseline Patient CharacteristicsCharacteristicsEtodolac (n = 9607)Naproxen (n = 8382)Patient years exposure26002996Age (y), mean ± SD57.6 ± 14.054.6 ± 13.8 50 years or older (%)72.263.7Men (%)91.088.0NSAID naive (%)59.558.7Rheumatoid arthritis (%)2.02.1Congestive heart failure (%)6.46.2Diabetes mellitus (%)14.814.7Concurrent low-dose aspirin (%)18.216.1History of ulcer (%)4.73.7Mean time in Dallas Veterans Affairs Medical Center system before study (y), mean ± SD4.4 ± 3.74.4 ± 3.5 Open table in a new tab Mean ages of aspirin users treated with etodolac or naproxen were 65.3 and 63.8 years, respectively. Mean ages of nonaspirin users treated with etodolac or naproxen were 56.7 and 53.7 years (P < .01 for age comparisons). Aspirin users were slightly more likely to be male: 95.6% in the etodolac group and 95.5% in the naproxen group compared with 88.5% and 87% in the etodolac and naproxen groups of nonaspirin users (P < .01 for etodolac/aspirin and naproxen/aspirin vs. nonaspirin groups).CSUGI eventsAfter adjudication, 41 CSUGI events were confirmed. Of these 41 events, 26 were complicated upper GI events (bleeding and perforations) and 15 were symptomatic gastroduodenal ulceration (Table 3). Time to development of a CSUGI event is plotted against cumulative event incidence for etodolac and naproxen in Figure 2. Among patients not taking low-dose aspirin, the annualized incidence of CSUGI events in etodolac-treated patients (.24%; 5 events/2115 patient-years) was significantly lower compared with naproxen-treated patients (.78%; 19 events/2451 patient-years) (P = .013; Figure 3A). In NSAID-naive patients not taking aspirin, the annualized incidence of CSUGI events was .24% (3 events/1248 patient-years) in patients taking etodolac and .99% (15 events/1518 patient-years) in patients taking naproxen (P = .02). Among patients not taking aspirin, the odds ratios of having a CSUGI event on etodolac was .24 (95% CI, .09–.63; P < .01) in the overall group and .18 (95% CI, .05–.61) in the NSAID-naive group (P < .01; Table 4).Table 3Description of CSUGI EventsCSUGI EventEtodolac (2600 patient-years)Naproxen (2996 patient-years)Symptomatic gastroduodenal ulcer69Upper GI bleedingaThe cause or source of upper GI bleeding was gastric ulcer in 7 patients, duodenal ulcers in 13 patients, and other GI sources in 4 patients.717Perforation02Total1328a The cause or source of upper GI bleeding was gastric ulcer in 7 patients, duodenal ulcers in 13 patients, and other GI sources in 4 patients. Open table in a new tab Figure 2Cumulative incidence of CSUGI events of all study patients not taking aspirin. Time to development of CSUGI events in the etodolac and naproxen groups; all patients not exposed to low-dose aspirin.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Figure 3Annualized incidences of CSUGI events. (A) Patients not taking aspirin. (B) Patients taking aspirin. The numbers above the bars are events per patient-years of exposure.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table 4Annualized Incidence Rates and Adjusted Odds Ratios of CSUGI Events for Etodolac Versus NaproxenAnnualized incidence rates (%) for etodolac/naproxenAdjusted odds ratio95% CIPNot taking aspirin All patients.24/.78.24.09–.63.01 NSAID-naive patients.24/.99.18.05–.61.01Taking aspirin All patients1.65/1.65.75.28–1.99NS NSAID-naive patients2.12/1.431.24.35–4.42NS Open table in a new tab Concurrent use of low-dose aspirin eliminated the significant difference in incidence of CSUGI events between etodolac and naproxen. The incidence of CSUGI events among naproxen-treated patients taking low-dose aspirin was approximately 2-fold higher than among naproxen-treated patients not taking aspirin (Figure 3B). Concomitant aspirin exposure increased event incidences among etodolac-treated patients by 7- to 9-fold (Figure 3B). Importantly, aspirin exposure concurrent with etodolac therapy was not associated with a lower incidence of CSUGI events when compared with the incidence in the naproxen group taking concurrent aspirin. In the overall patient groups taking aspirin (average dose of aspirin, 259 and 261 mg/day in the etodolac and naproxen groups, respectively [doses not statistically different]), the annualized incidence of CSUGI events was 1.65% for etodolac and 1.65% for naproxen (P = NS). Among NSAID-naive patients taking aspirin, the event incidence was 2.12% and 1.43% (Figure 3B) in the etodolac and naproxen groups, respectively (P = NS).Logistic regressionStepwise logistic regression analysis was performed to control for baseline demographics among groups. In the final analysis, only NSAID taken (etodolac or naproxen), concomitant use of aspirin, and congestive heart failure were significant independent predictors of a CSUGI event. Age of 50 years and older nears significance (P = .1). Compared with naproxen, use of etodolac was associated with a statistically significant decreased risk of CSUGI events (adjusted odds ratio, .39; 95% CI, .20–.76; P = .006). In addition, concurrent use of aspirin independently increased the likelihood of an event (adjusted odds ratio, 3.28; 95% CI, 1.7–6.32; P < .001), as did congestive heart failure (adjusted odds ratio, 2.46; 95% CI, 1.13–5.36; P = .02).DiscussionThe need for anti-inflammatory agents with safer GI profiles was the primary reason for the development of COX-2 selective inhibitors. While three COX-2 inhibitors are currently available by prescription in the U.S., among the currently available agents only rofecoxib and celecoxib have been subjected to large studies that document improved GI safety compared with nonselective NSAIDs. Our current study of upper GI outcomes with etodolac provides evidence that, based on its functional behavior, etodolac should also be classified as a COX-2 inhibitor.Our concept of the definition of drugs identified as COX-2 selective inhibitors is that such agents should not inhibit COX-1 in the GI tract and should be associated with a reduction in clinically relevant GI events while providing effective reductions in inflammation at clinically relevant doses of interest. Etodolac meets each of these objectives. In in vitro assays, etodolac is COX-2 selective.4Warner T.D. Giuliano F. Vojnovic I. Bukasa A. Mitchell J.A. Vane J.R. Nonsteroid drug selectivities for cyclo-oxygenase-1 rather than cyclo-oxygenase-2 are associated with human gastrointestinal toxicity a full in vitro analysis.Proc Natl Acad Sci U S A. 1999; 96: 7563-7568Crossref PubMed Scopus (1410) Google Scholar However, in vitro COX-2 selectivity does not always accurately predict reductions in ulcers and GI events in clinical practice. Using a dose of etodolac of 400 mg twice daily taken for 1 month, Laine et al observed no gastric mucosal prostaglandin inhibition in healthy humans.5Laine L. Sloane R. Ferretti M. Cominelli F. A randomized, double-blind comparison of placebo, etodolac, and naproxen on gastrointestinal injury and prostaglandin production.Gastrointest Endosc. 1995; 42: 428-433Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar Thus, etodolac at a dose of 400 mg twice daily in the short term does not inhibit gastric COX-1 (ie, is COX-2 selective). The short-term observations
Referência(s)