The effects of non-selective, preferential-selective and selective COX-inhibitors on the growth of experimental and human tumors in mice related to prostanoid receptors

Earlier observations on cyclo-oxygenase inhibitors (NSAIDs) restricting tumor growth were re-evaluated by comparing the effects of non-selective, preferential selective and selective derivatives of COX-inhibitors on tumor growth in mouse models with either prostaglandin-sensitive (MCG-101, human tumors) and -insensitive transplants (K1735-M2). Tumor growth, with and without provision of a classical cyclo-oxygenase inhibitor (indomethacin), was related to tumor content of COX-1/COX-2 protein as well as to EP1-EP4 and prostacyclin receptor expression. Mouse serum amyloid protein (SAP) was measured as an indicator of systemic inflammation, which relates to pro-inflammatory cytokines. Indomethacin inhibited tumor growth and prolonged the survival of mice bearing MCG-101 tumors, which display a high production of PGE2, while K1735-M2 tumors with insignificant amounts of PGE2 did not respond to indomethacin at all. However, the effects of various NSAIDs on tumor growth were highly variable in combination with the fact that most preferential selective and selective COX-2 inhibitors attenuated poorly systemic inflammation evaluated by plasma concentrations of mouse SAP. The ability of NSAIDs to attenuate tumor growth was not related to the tumor content of COX-2 protein as expected. Multivariate analysis suggests that significant COX-inhibition of tumor growth may be related to tumor expression of subtype EP2, EP3 (p<0.005) and perhaps EP4 (p<0.09) in complex interplay. The extent of tumor growth inhibition by COX-inhibitors is not simply related to drug specificity on COX-1 or COX-2 pathways. Such effects may instead be related to tumor expression of prostanoid receptors in tumor tissue.