PubMed1999
The effect of the NSAIDs, retinoids and DFMO on AOM-induced colon tumors, and ACF, cell proliferation, and apoptosis is summarized in Table 1. The ability to prevent AOM-induced ACF has been used as an assay to screen agents for chemoprevention. As discussed above, all six potential chemopreventive agents, aspirin, 2-CPR, DFMO, 4-HPR, piroxicam, and 9-cis-retinoic acid, decreased the level of AOM-induced ACF. However, two of the agents, aspirin (at doses that greatly reduced the yield of ACF) and 2-CPR did not prevent AOM-induced colon tumors. Hence, aspirin and 2-CPR would appear to be false positive in the ACF assay. Besides being a false positive in the ACF assay, 2-CPR actually had the opposite effect of doubling the yield of colon tumor. The false positive result for aspirin could be due to the lower sensitivity of the AOM-induced colon cancer assay compared to the ACF assay. However, aspirin [table: see text] significantly reduced the yield of ACF at a dose (600 mg/kg diet) one-third the dose (1800 mg/kg diet) that did not reproducibly reduce the yield of colon tumors. Thus, although there were no false negative results, two of the six agents gave false positive results in the AOM-induced ACF assay with respect to their ability to prevent colon cancer. Two other potential biomarkers for chemopreventive activity are the ability to reduce the level of cell proliferation and to enhance the level of apoptosis. All six of the agents including aspirin and 2-CPR reduced the level of cell proliferation in adenomas. Thus, similar to their ability to prevent ACF, the ability of aspirin and 2-CPR to decrease cell proliferation were also false positive responses with respect to prevention of colon cancer, but not with respect to the prevention of ACF. Piroxicam, the most potent of the six agents in preventing AOM-induced colon cancer, did not significantly affect the level of cell proliferation in adenomas which is a false negative response. Hence, only three of the six agents (50%) were correctly identified as potential chemopreventive agents by their ability to reduce the level of cell proliferation. In contrast, retinoids, including the three discussed here, demonstrated good correlation between the ability to prevent AOM-induced ACF and the ability to decrease cell proliferation in colonic mucosa or ACF. Thus, within some classes of agents such as the retinoids, the ability to prevent ACF and to reduce cell proliferation appear to correlate, while in other classes including the NSAIDs, the correlation appeared not to exist. The four agents that prevented colon cancer all enhanced the level of apoptosis, while the two agents that did not prevent colon cancer did not effect apoptosis. Three other chemopreventive agents, including phenylethyl-3-methylcaffeate and the NSAIDs, curcumin and sulindac, have been shown by Samaha et al. to enhance apoptosis in AOM-induced colon tumors. Thus, although a very limited number of chemopreventive agents have been evaluated for the ability to enhance apoptosis in the colon, there appears to be an association between the ability to enhance apoptosis and the ability to prevent colon cancer. The use of the AOM-induced ACF assay to screen agents for the ability to prevent colon tumors would appear to result in false positive responses including agents (2-CPR and quercetin) that actually promote colon cancer. However, our results suggest that false positive responders could be distinguished by their inability to enhance apoptosis while potential chemopreventive agents would enhance it. It is therefore proposed that a Two Step Procedure be used to screen agents for the ability to prevent colon cancer. Step 1 would be the determination of the ability to prevent ACF. Because the ACF assay appears to suffer more from false positive than from false negative responders, apparently few potent chemopreventive agents would be missed. Also the ACF assay could be the source of foci for evaluation of the effect