This study examined the ability of concomitant use of aspirin to offset the risk of acute myocardial infarction (MI) in patients being treated with cyclooxygenase-2 (COX-2) selective and non-selective NSAIDs. The protocol was based on the increased risk of MI reported with the use of these agents. The hypothesis was based on studies that suggest that inhibition of thromboxane by aspirin may reverse the imbalance resulting from selective inhibition of COX-2-mediated prostacyclin formation.

Data was derived from a California Medicaid database of patients diagnosed with OA or RA. Patients were matched as to age, gender, and length of drug exposure. From this database, 15,343 cases of acute MI (8% fatal) were identified. Adjusted risk ratios (95% CI) of acute MI were 1.31 (1.20 − 1.43) with rofecoxib, 1.13 (1.04 − 1.19) with celecoxib, 1.08 (0.97 − 1.19) with ibuprofen, 1.65 (1.27 − 2.15) with indomethacin, 1.52 (1.14 − 2.03) with meloxicam, and 1.47 (1.03 − 2.11) with sulindac. Concomitant use of aspirin reversed MI risk with rofecoxib to 1.03, with celecoxib to 0.88, with meloxicam to 0.53, and with sulindac to 0.77. The risk was partially reversed with indomethacin to 1.21, but unchanged with ibuprofen (1.20).

The authors concluded that COX-2 selective and NSAIDS contribute to an increased risk of MI probably due to the inhibition of prostacyclin formation. This risk can be reduced by concomitant aspirin use. It was suggested that the incomplete reversal of risk with NSAIDS may be due to the pharmacodynamic interference of these NSAIDS with the binding of aspirin to platelet COX-1. It was also suggested that further evaluation of the potentially-increased GI toxicity of aspirin-NSAID therapy was needed.