The chemopreventive efficacies of nonsteroidal anti-inflammatory drugs: the relationship of short-term biomarkers to long-term skin tumor outcome

Abstract

The ultraviolet B (UVB) component of sunlight, which causes DNA damage and inflammation, is the major cause of nonmelanoma skin cancer (NMSC), the most prevalent of all cancers. Nonsteroidal anti-inflammatory drugs (NSAID) and coxibs have been shown to be effective chemoprevention agents in multiple preclinical trials, including NMSC, colon, and urinary bladder cancer. NSAIDs, however, cause gastrointestinal irritation, which led to the recent development of nitric oxide (NO) derivatives that may partially ameliorate this toxicity. This study compared the efficacy of several NSAIDs and NO-NSAIDs on UV-induced NMSC in SKH-1 hairless mice and determined whether various short-term biomarkers were predictive of long-term tumor outcome with these agents. Naproxen at 100 (P = 0.05) and 400 ppm (P < 0.01) in the diet reduced tumor multiplicity by 26% and 63%, respectively. The NO-naproxen at slightly lower molar doses shows similar activities. Aspirin at 60 or 750 ppm in the diet reduced tumor multiplicity by 19% and 50%, whereas the equivalent doses (108 and 1,350 ppm) were slightly less effective. Sulindac at 25 and 150 ppm in the diet, doses far below the human equivalent dose was the most potent NSAID with reductions of 50% and 94%, respectively. In testing short-term biomarkers, we found that agents that reduce UV-induced prostaglandin E2 synthesis and/or inhibit UV-induced keratinocyte proliferation yielded long-term tumor efficacy.

Figure 1. Effect of NSAIDs on UV-induced COX-2 expression

Three groups of 3 mice each were fed their respective experimental diets for one week, exposed to 220 mJ/cm² UV (except the no UV group) and killed 6 hr later. The epidermis was chipped from frozen skin into RIPA buffer and subjected to Western blot analysis as described under Methods and Materials. A representative experiment is shown; densitometry was performed on each blot, the UV values set to 1.0 and the NSAID data expressed as the percentage of the UV control ± SE.

Figure 2. Effect of naproxen or NO-naproxen on UV-induced skin tumorigenesis

Groups of 20 female SKH-1 mice were placed on their experimental diets and exposed to UV irradiation thrice weekly. Tumors were counted weekly and the data calculated as the average number of tumors per mouse (panel A). Tumor multiplicity for 100 ppm naproxen was significantly (p= 0.05) different from the UV control group; 400 ppm naproxen and NO-naproxen were very significantly (p<0.001) different from the UV control. Tumor incidence (panel B) was calculated as the percentage of mice bearing tumors. Mice in the 400 ppm naproxen, 100 ppm and 400 ppm NO-naproxen groups developed tumors significantly (p<0.001 for all 3 groups) slower than the UV control.

Figure 3. Effect of aspirin or NO-aspirin on UV-induced skin tumorigenesis

Groups of 20 female SKH-1 mice were placed on their experimental diets and exposed to UV irradiation thrice weekly. Tumors were counted weekly and the data calculated as the average number of tumors per mouse (panel A). Tumor multiplicity for 60 ppm aspirin and 1350 ppm NO-aspirin were significantly (p<0.05) different from the UV control group; 750 ppm aspirin was very significantly (p<0.001) different from the UV control. Tumor incidence (panel B) was calculated as the percentage of mice bearing tumors. Although mice in the 1350 ppm aspirin group developed tumors more slowly than the other groups, this was not statistically different.

Figure 4. Effect of sulindac on UV-induced skin tumorigenesis

Groups of 20 female SKH-1 mice were placed on their experimental diets and exposed to UV irradiation thrice weekly. Tumors were counted weekly and the data calculated as the average number of tumors per mouse (panel A). Tumor multiplicity for 25 ppm and 150 ppm sulindac were very significantly (p<0.001) different from the UV control. Tumor incidence (panel B) was calculated as the percentage of mice bearing tumors. Mice in the 150 ppm sulindac group developed tumors significantly (p<0.001) slower than the UV control group.