Is COX-2 inhibition a panacea for cancer prevention?

The epidemiologic evidence and rodent studies suggest strongly that nonsele ctive inhibitors of cyclooxygenase (COX) enzymes such as aspirin, inhibitin g both COX-1 and COX-2 isoforms, reduce the incidence of and mortality from intestinal tumors. Genetically manipulated animals show that both Cox-1 an d Cox-2 disruptions decrease the tumor yield, both in genetically predispos ed and in carcinogen-treated mice. The mechanisms by which COX-1 and COX-2 deficiency decrease tumorigenesis are still unknown. Cox-2 overexpression i ncreased the tumor yield in mammary glands of the multiparous, but not virg inal female transgenic mice using the murine mammary tumor virus promoter. The Cox-2 protein was strongly induced during pregnancy and lactation. Thes e data suggest that Cox-2 overexpression may be an important target for can cer chemoprevention. This finding was supported by the observed cancer-prev entive effects of the COX-2-specific inhibitors in humans and in rodents. H owever, based on the available data, we cannot totally attribute the cancer preventive effects of nonsteroidal antiinflammatory drugs (NSAIDs) to COX- 2 alone even COX-1 may have an important role in cancer prevention as sugge sted by the Cox-l-deficient Min mice. It is likely that COX-1 plays a more important role in NSAID-induced toxicity in humans, such as in gastric ulce r formation-but inhibition of COX-2 may not be without toxic manifestations either, as suggested by the poor survival of the Cox-2-nulled mice. Combin ations of COX-2 inhibitors with other agents that target other pathways in carcinogenesis may be a more efficacious and a less toxic strategy in cance r chemoprevention. (C) 2001 Wiley-Liss, Inc.