Predictive Factors for NSAIDs-related Gastrointestinal Toxicity: Can COX-2 Selective Inhibtor Prevent it?.

BACKGROUND/AIMS: To identify predictive factors for NSAIDs-related GI toxicity and to clarify whether cox-2 selective inhibitors can prevent it or not.

METHODOLOGY: We have surveyed all patients received esophagogastroduodenoscopy examined at our hospital between January 2008 and September 2011. We performed the following retrospective analyses of the clinical records of the 253 patients prescribed NSAIDs. The severity of gastro duodenal mucosal lesions was evaluated using the modified gastro duodenal LANZA score. The following scores for response were used: 0 = no lesions (LANZA score 0); 1 = erosion and redness (LANZA score 1-3); 2 = ulcer (LANZA score 4). Predictors evaluated were factors potentially related to pathogenesis of NSAIDs related GI toxicity. Ordered logistic regression analysis was performed to identify predictive factors for NSAIDs related. ulcer.

RESULTS: A multivariate logistic regression identified number of risk factors (odds ratio (OR) = 6.82, confidence interval (CI) = 5.31-8.76; P = 0.01), concomitant use of anti- cancer drugs (OR = 2.17, Cl = 1.02-4.62; P = 0.04) were found to be significant factors.

CONCLUSIONS: Number of risk factors and concomitant use of anticancer drugs were shown to be predictive factors for NSAID-related GI toxicity. Use of celecoxib or proton pump inhibitor was not identified as a protective factor.