Objective: Rofecoxib (Vioxx), the first COX-2 selective non-steroidal anti-inflammatory drug (NSAID), was recently withdrawn from the market due to the increased risk of acute myocardial infarction. The precise mechanism responsible for this phenomenon still remains unknown. Tumor necrosis factor alpha (TNF-alpha) is a cytokine, possibly most responsible for mortality in patients with acute myocardial infarction. However, this study was designed to study possible effects of rofecoxib on the level of TNF-alpha by using MSU crystal induced inflammation in the rat subcutaneous air pouch model.
Methods: Rat subcutaneous air pouches were produced and examinations commenced 6 days later. Control groups received only MSU crystals, or no crystals or drugs. To begin with, rofecoxib (30 mg/kg), indomethacin (20 mg/kg) or diclofenac (3 mg/kg) were administered to groups of 5 rats each. Thirty minutes later, MSU crystals were injected into air pouches, except for the negative control group. Twenty-four hours later, the rats were sacrificed for aspiration of fluid and for the dissection of pouch walls to determine leukocyte counts, pouch wall histology, and to assay IL-10 and TNF-alpha.
Results: Intra-pouch injection of MSU crystals, compared to non-injected pouches, caused an increase in white blood cell count (WBC) (30 +/- 44.7 versus 4508 +/- 792.3 cells/mm3), in the numbers of pouch wall vessels (vascular index) (4.8 +/- 0.3 versus 11.4 +/- 1.5 vessels/high-power field) and in TNF-alpha (50.0 +/- 13.4 versus 70.34 +/- 20.9 ng/mL), but not in interleukin-10 (IL-10) (60.6 +/- 63.0versus 61.48 +/- 7.1). WBC and vascular index were significantly reduced in all study groups compared to the control group (p < 0.05). Levels of TNF- in fluids were unexpectedly and significantly (p < 0.05) increased in all cases. The highest level of TNF-alpha was found in the rofecoxib group. In contrast to TNF-alpha, IL-10 levels were significantly (p < 0.05) decreased in all three drug groups. Indomethacin tended to suppress inflammation more effectively. However, there was no significant difference between the groups for IL-10 (p > 0.05).
Conclusion: All three NSAIDs exhibited anti-inflammatory activity against MSU crystal induced inflammation. The difference in anti-inflammatory effects of these three non-steroidal drugs is seen not only in the anti-inflammatory effect on MSU induced inflammation but also in the nature of the effects. Refocoxib tended to increase the TNF-alpha level. Whether increased TNF-alpha levels can help explain the side effect of COX-2 specific inhibitors still requires further studies.