1. Small Bowel Toxicity of Nonselective NSAIDs Revealed by Capsule Endoscopy: Results From a Pivotal Clinical Trial Glenn M. Eisen, M.D., M.P.H. Associate Professor of Medicine Oregon Health and Science University, Portland, Oregon Glenn M. Eisen, M.D., M.P.H. Associate Professor of Medicine Oregon Health and Science University, Portland, Oregon

2. 2 NSAIDs and Small Intestinal Damage * 8.4% * 8.4% 0.6% 0 0 2 2 4 4 6 6 8 8 Nonspecific ulcers NSAID group (n = 249) Control group (n = 464) NSAID group (n = 249) Control group (n = 464) 10 *P < 0.001 Allison, et al. N Engl J Med. 1992;327:749–754. *P < 0.001 Allison, et al. N Engl J Med. 1992;327:749–754. % of Subjects With Small Intestinal Ulcers

3. 3 NSAID Use Is Associated With Both Upper and Lower GI Bleeding Case-controlled study of upper GI bleeding (UGIB) and lower GI bleeding (LGIB) Use of NSAIDs within 1 week of admission was documented Case-controlled study of upper GI bleeding (UGIB) and lower GI bleeding (LGIB) Use of NSAIDs within 1 week of admission was documented * P < 0.05 OR = odds ratio; AOR = OR adjusted for age, race, and gender. Wilcox CM, et al. Dig Dis Sci. 1997;42:990–997. * P < 0.05 OR = odds ratio; AOR = OR adjusted for age, race, and gender. Wilcox CM, et al. Dig Dis Sci. 1997;42:990–997. PatientsNSAID UseP ValueORAOR Controls 34% – – – UGIB 60%< 0.0013.0* 3.2 LGIB 60%< 0.0013.0* 2.6* PatientsNSAID UseP ValueORAOR Controls 34% – – – UGIB 60%< 0.0013.0* 3.2 LGIB 60%< 0.0013.0* 2.6*

4. 4 Rates of Serious Lower GI Events per 100 Patients-Years in Patients Receiving Rofecoxib vs Naproxen 0.41 0.89 0 0.2 0.4 0.6 0.8 1.0 % of Patients with Serious Lower GI Events per 100 Patient-Years Rofecoxib (n = 4,047) Naproxen (n = 4,029) Relative Risk (95% CI) 0.46(0.22-0.93) P = 0.03 Laine, et al. Gastroenterology. 2003;124:288-292.

5. 5 Celecoxib Diclofenac Ibuprofen Celecoxib Diclofenac Ibuprofen 10 % of Patients 0 0 2 2 4 4 6 6 8 8 All Patients Patients Without Complications *P < 0.05 vs celecoxib FDA Arthritis Advisory Committee Meeting. February 7, 2001. Gaithersburg, Maryland, USA. *P < 0.05 vs celecoxib FDA Arthritis Advisory Committee Meeting. February 7, 2001. Gaithersburg, Maryland, USA. 2.4 * 5.7 * 4.5 * 3.7 2.0 * 4.4 CLASS: Decreases in Hct ≥ 10% and/or Hgb > 2 g/dL CLASS: Decreases in Hct ≥ 10% and/or Hgb > 2 g/dL

6. 6 Hypothesis The combination of a non-specific NSAID + PPI will be associated with a rate of small bowel mucosal breaks that is significantly higher than the rate for placebo or for a COX-2 specific inhibitor

7. 7 Inside the M2A Capsule

8. 8 NSAID-Induced Lesions

9. 9 Celecoxib 200 mg BID Placebo Start study drug Study Design Naproxen 500 mg BID plus omeprazole 20 mg QD Naproxen 500 mg BID plus omeprazole 20 mg QD 2-week run-in period Screening Randomization Final M2A video capsule (baseline) M2A video capsule (baseline) M2A video capsule (final) M2A video capsule (final) 2-week Tx period Healthy Subjects

10. 10 Primary End Point (mITT) 0 0 1 1 2 2 3 3 Celecoxib (n = 115) Celecoxib (n = 115) Naproxen + PPI (n = 111) Naproxen + PPI (n = 111) Placebo (n = 113) Placebo (n = 113) Mean # of SB Mucosal Breaks (Grades 2, 3, 6, and 7) P < 0.001 Goldstein J, Eisen G, Gralnek I, Clin Gastroenterol and Hepatol Feb 2005 0.32 0.11 2.99 Mean Number of Small-Bowel Mucosal Breaks (Grades 2, 3, 6, and 7) P = 0.042

11. 11 Capsule Endoscopy Study Secondary End Point (mITT) Celecoxib (n = 115) Naproxen + omeprazole (n = 111) Placebo (n = 113) P value* Incidence of mucosal breaks, N (%) 18 (16%) 61 (55%) 8 (7%) < 0.001 Incidence of Small Bowel Mucosal Breaks *Across three treatments corresponding to general association of the Cochran-Mantel- Haenszel test having stratified by site

12. 12  As in the upper GI tract, inhibition of COX-1 by naproxen and not celecoxib translates into significantly different rates of mucosal injury in the small bowel  These findings extend the original COX-1 sparing hypothesis beyond the upper GI tract and into the small bowel  As in the upper GI tract, inhibition of COX-1 by naproxen and not celecoxib translates into significantly different rates of mucosal injury in the small bowel  These findings extend the original COX-1 sparing hypothesis beyond the upper GI tract and into the small bowel Conclusions

13. 13 Conclusions (2) Use of NS-NSAIDs can potentially lead to mucosal lesions, gastrointestinal bleeding beyond the ligament of Treitz PPI use does not protect against NSAID induced damage to the small/large intestine Capsule endoscopy provides a noninvasive assessment of the small bowel, which may be clinically useful in patients at risk for small bowel injury