Analgesia and COX-2 inhibition.

While non-steroidal anti-inflammatory drugs (NSAIDs) are the mainstay of therapy for the management of acute pain and rheumatoid arthritis, toxicity associated with chronic administration limits their benefit-to-risk relationship in many patients. A series of studies is reviewed that assesses the relationship between cytokines released at the site of tissue injury and NSAID analgesia, and the in vivo selectivity of a selective cyclooxygenase (COX)-2 inhibitor (celecoxib) in comparison to a dual COX-1/COX-2 inhibitor (ketorolac). Three replicate studies in the oral surgery model of acute pain used submucosal microdialysis sample collection for the measurement of prostaglandin E2 (PGE2; a product of both COX-1 and COX-2) and thromboxane B2 (as a biomarker for COX-1 activity) with parallel assessments of pain. The time course of PGE2 production was consistent with early release due to COX-1 activity followed by increased production 2-3 hours after surgery, consistent with COX-2 expression. Ketorolac 30 mg at pain onset suppressed both pain and peripheral PGE2 levels. Ketorolac 1 mg either at the site of injury or intramuscularly also produced analgesia but without any effect on peripheral PGE2 levels. Celecoxib selectively suppressed PGE2 but not TxB2 at time points consistent with COX-2 activity, while producing analgesia. These studies demonstrate the ability to assess the time course and selective effects of COX-2 inhibitors in vivo and suggest that suppression of COX-2 mediated PGE2 is temporally related to NSAID analgesia.