DESCRIPTION (provided by applicant): We propose to continue Case-Control Surveillance (CCS) to assess the effects of prescription and nonprescription medications and other factors on the risk of cancers. Multiple case-control studies are conducted within a single data collection system. Nurse-interviewers in participating hospitals administer standard questionnaires to patients admitted for recently diagnosed cancers and other illnesses to obtain information on prescription and nonprescription medication use and on potential confounders or modifiers of exposure-disease associations. Data collection includes information on the newly popular class of nonprescription medications, herbals, and we have modified our drug dictionary in order to code these preparations. DNA is collected (from buccal cell samples) to allow for assessment of modification of drug disease associations by inherited genotype. We will continue to interview new cases and controls and to collect buccal samples. Enrollment of new patients is necessary because new drugs are constantly being introduced to the market, prescription drugs continue to be switched to nonprescription, and nonprescription drugs including herbals which have only recently become widely used are not otherwise monitored. Some herbals have been shown to have carcinogenic effects, affect estrogen levels, and influence the cytochrome p450 system involved with drug metabolism. One focus of CCS analyses will be the relation of the widely used drug classes, statins, histamine H2 antagonists, and selective serotonin uptake inhibitors, to risk of common cancers. We will also assess the relation of use of herbals to these cancers because the effects of these substances on cancer risk are largely unknown. We will assess modification by specific genetic polymorphisms of the relation of colon and breast cancer risk to use of nonsteroidal anti-inflammatory drugs. The usefulness of CCS for quantifying positive and inverse associations with cancer risk, documenting safety, and discovering associations has been extensively documented. Other systems do not have the capacity to assess a range of drug/genotype interactions or to systematically assess nonprescription drugs. CCS results have repeatedly been confirmed in other studies, indicating the success of efforts to minimize bias. CCS is a unique resource for cancer epidemiology.