Screening agents for preventive efficacy in a bladder cancer model: study design, end points, and gefitinib and naproxen efficacy.

Purpose: We optimized agent testing in an in vivo bladder cancer model and determined the most sensitive, relevant protocol to test efficacy in clinical prevention trials.

Materials And Methods: Female Fischer-344 rats (Harlan) were treated with the bladder carcinogen OH-BBN (TCI America, Portland, Oregon) for 8 weeks. Rats were treated with naproxen (400 mg/kg diet), aspirin (Sigma(R)) (300 or 3,000 mg/kg diet), Iressa(R) (10 mg/kg gefitinib body weight daily) or resveratrol (1,000 mg/kg diet) using 1 of 3 protocols, including treatment beginning 1) 1 week after OH-BBN and continuing for 7 months, 2) 3 months after OH-BBN after microscopic lesions already existed and continuing for 3 months, and 3) 1 week after OH-BBN and continuing for 4 months. In protocols 1 and 2 bladder lesion weight and large tumors were primary end points, and in protocol 3 microscopic cancer was the end point.

Results: Using protocol 1 naproxen, Iressa, resveratrol, and low and high dose aspirin altered the formation of large bladder tumors by 87% (decreased), 90% (decreased), 3% (increased), 6% (decreased) and 60% (decreased), respectively. Using protocol 2 Iressa and naproxen were also highly effective. Protocol 3 evaluation revealed that only Iressa caused a significant decrease in microscopic bladder cancers (63%).

Conclusions: Initiating treatment after OH-BBN or when bladder lesions already existed showed naproxen and Iressa to be effective in preventing formation of large cancers. Low dose aspirin and resveratrol were ineffective. In protocol 3, in which microscopic lesions were the end point, only Iressa was effective. Thus, an established cancer end point appears preferable. Naproxen, which has an excellent cardiovascular profile, or epidermal growth factor receptor inhibitors may be effective in an adjuvant setting.