Endoscopic evidence of mucosal injury in patients taking ticlopidine compared with patients taking aspirin/nonsteroidal antiinflammatory drugs and controls

Background: Ticlopidine is a novel antiplatelet agent used alone or in combination with aspirin and anticoagulants in the treatment and prevention of various vascular diseases. Gastrointestinal side effects, including bleeding, have been reported with use of ticlopidine in most of the vascular prevention trials. We studied the endoscopic evidence of mucosal damage in patients taking ticlopidine compared with patients taking aspirin/nonsteroidal antiinflammatory drugs (NSAIDs) and matched controls. Study: We performed a longitudinal review of gastrointestinal endoscopy, pharmacy databases, and medical records of patients referred to our service over a period of 6 months for endoscopic evaluation of upper gastrointestinal bleeding, unexplained anemia, or abdominal pain. Data were collected and analyzed for 55 patients taking ticlopidine, 77 age- and gender-matched patients taking aspirin or NSAIDs, and 560 age- and gender-matched control patients not taking any of these medications. Results: The overall prevalence of ulcers was marginally higher in the aspirin/NSAID group than in the ticlopidine group (35% vs. 29%) and was significantly higher among patients taking aspirin, NSAIDs, or ticlopidine than among controls (15%). Gastritis was also noted more frequently in the aspirin/NSAID and ticlopidine groups than in the control group. Endoscopic evidence of esophagitis was significantly more frequent in the control group than in the aspirin/NSAID and ticlopidine groups. There was no significant difference across groups in the prevalence of ulcers, gastritis, or esophagitis. Conclusions: Patients taking ticlopidine are more likely to have endoscopic evidence of mucosal damage than matched control patients and are nearly as likely to have such damage as endoscopically evaluated patients taking aspirin or NSAIDs. However, these findings must be confirmed using prospective cohort data for patients in primary care settings, to avoid referral bias.