Linda Laio

NCI, DCEG, OEEB

PLCO (Learn more about this study)

2009-0040

Feb 13, 2009

Examining Questionnaire-Based Risk Factors for Colorectal Cancer in the PLCO Cancer Screening Trial: Smoking behavior

In a coordinated series from PLCO intramural and extramural investigators of 19 published manuscripts on risk of prevalent colorectal adenoma in the PLCO Trial, risk evaluations were reported for dietary fiber, carbohydrate intake and glycemic load, meat cooking practices and the meat mutagen PhIP, dietary calcium, a gene variant in the calcium-sensing receptor, serum selenium, hormone replacement therapy, insulin-like growth factors, tobacco use, and selected genetics factors. Analyses are underway to investigate etiology of incident adenoma and recurrent adenoma in the PLCO Trial. As an ongoing trial, data security is important for primary and secondary trial endpoints. Colorectal cancer mortality is not under consideration in the current proposal. Following strategies worked out for prostate cancer analyses and multiple outcome analyses, we propose that analyses for colorectal cancer in the screening arm first be conducted by the study investigators at NCI/DCEG, with analyses including data from the control arm to be carried out in a secure data enclave (Information Management Systems, Inc., Rockville, MD). We will work closely with the Trial lead statistician, Dr. Phil Prorok, on all data analyses and presentations to ensure that the integrity of the Trial data is not compromised. In addition, we plan continual collaborations with the PLCO Colon Committee and other interested investigators to conduct coordinated analyses of selected risk factors for colorectal cancer. Factors to be evaluated include smoking behavior, use of non-steroidal anti-inflammatory drugs (NSAIDs), body mass index (BMI), energy balance, physical activity, family history, and dietary intake of meat, meat mutagens, heme iron, fruits, vegetables, flavonoids, dietary sources of flavonoids, coffee, tea, caffeine, dairy products, calcium, fatty acids, and fiber. We will use Cox proportional hazards regression to estimate relative risks (RRs) and 95% confidence intervals (95% CIs), adjusting for potential confounders. Because an active, broad and productive research program on colorectal tumors is underway in the PLCO Trial, there will be added-value opportunities to compare risk factor profiles for colorectal cancer with the profiles for other colorectal tumor categories in the same study population. No other study of colorectal tumors has had this capacity. This proposal in its entirety was approved by the PLCO in 2006 (Study ID 2006-317). Herein we submit the individual proposal for each hypothesis to facilitate progress tracking. A working group is formed among the lead investigators (see A.5. Investigator Team) to carry out all 11 proposed analyses in a coordinated fashion. A summary of the literature and the coordinated study design is provided from C.2. to C.5.; specifics for the Smoking behavior analysis are detailed in C.6.