Tumor promotion as a target of cancer prevention.

Tumor promotion is an essential process in multistage cancer development providing the conditions for clonal expansion and genetic instability of preneoplastic and premalignant cells. It is caused by a continuous disturbance of cellular signal transduction that results in an overstimulation of metabolic pathways along which mediators of cell proliferation and inflammation as well as genotoxic by-products are generated. Among such pathways the oxidative metabolism of arachidonic acid has turned out to be of utmost importance in tumor promotion. The aberrant overexpression of cyclooxygenase-2, an inducible enzyme of prostanoid synthesis and lipid peroxidation, is a characteristic feature of more than two-thirds of all human neoplasias, and the specific inhibition of this enzyme has been found to have a substantial chemopreventive effect in both animal models and man. The prostaglandins produced by COX-2 promote tumor development by stimulating cell proliferation and angiogenesis and by suppressing programmed cell death and immune defense. In mice, a COX-2 transgene fused with the keratin 5 promoter, which is constitutively active in the basal (proliferative) compartment of stratified and simple epithelia, causes a preneoplastic and premalignant phenotype in several organs. Among these organs, skin, mammary gland, urinary bladder, and pancreas have been investigated in more detail. Histologically and biochemically, the COX-2-dependent alterations resemble an autopromoted state that--as shown for skin and urinary bladder--strongly sensitizes the tissue for carcinogenesis. In transgenic animals COX-2 expression is not restricted to keratin 5-positive cells but is seen also in adjacent keratin 5-negative cells. This spreading of the COX-2 signal indicates a paracrine mechanism of autoamplification. While cancer chemoprevention by COX-2 inhibition is a rapidly developing field, much less is known about other pathways of unsaturated fatty acid metabolism, although some of them may play a role in carcinogenesis rivaling that of prostaglandin formation. Here an urgent demand for systematic research exists.