How to mechanistically explain the CONDOR study data

Results of the CONDOR study suggest that in osteoarthritis and rheumatoid arthritis patients at elevated risk of gastrointestinal (GI) events, treatment with celecoxib, a cyclooxygenase (COX)-2 selective non-steroidal anti-inflammatory drug (NSAID), demonstrated significantly lower toxicity in the upper and lower (GI) tract when compared to the non-selective NSAID diclofenac plus a proton-pump-inhibitor (PPI), omeprazole. According to current knowledge, traditional NSAIDs (tNSAIDs) as non-selective COX-inhibitors exert their damaging effects on the upper GI tract, largely by reduction of the COX-1 related synthesis of gastro-protective prostaglandins. Thus, the question arises, how NSAIDs do exert their damaging effects especially in the lower GI tract and how to explain the reduced risk of a COX-2 selective inhibitor, celecoxib. Here we hypothesize, that the toxicity of celecoxib on enteral mucosa cells is lower than observed with other NSAIDs, and can be explained COX-independently by typical physicochemical properties of the NSAID substances (e.g., acidic, lipophilic, amphiphilic, surfactant properties). As a consequence these features account for differences in (1) uncoupling effects on mitochondria, (2) effects on cell membrane integrity, and/or (3) formation of "toxic micelles" with bile salts. The evidence for these differences is mainly based on experimental findings. However, several phenomena show differences in extent (e.g., uncoupling effects). The reduced toxicity appears to be rather a substance-specific characteristic. This is an unconditional reason to carry on investigating these phenomena in experimental and large-scale clinical trials.