B174

Although differences in people's dietary intake are thought to account for substantial variation in cancer incidence, there has been a consistent lack of effect of diet in cancer clinical trials. These inconsistencies may reflect the multi-factorial and complex nature of cancer and the complex and diverse dietary components that may alter one or more phases of the cancer process. Another potential reason for the failure to find an effect in dietary intervention trials is the variation in adherence among participants. In this analysis of the Polyp Prevention Trial (PPT), we examined the effect on adenoma recurrence of high adherence to a low fat, high fiber, high fruit and vegetable intervention. The most adherent participants were defined as "Super Compliers". These Super Compliers consistently met or exceeded each of the three dietary goals at all four annual visits during the trial. During the four years of the trial, these Super Compliers (n=210) reported consuming 17.2% of energy from fat, 22.9 g fiber/1000kc, and 5.32 servings of fruits & vegetables/1000kc. Dietary adherence was then modeled as a potential predictor of adenoma recurrence using logistic regressions. Super Compliers had a lower adenoma recurrence rate; 31.0% compared to a 39.5% for the control group. In a logistic regression model comparing adenoma recurrence in these Super Compliers to the control group, after adjusting for all significant baseline covariates, the OR= 0.65, 95% CI=0.47-0.92. Further adjustment for forty-two baseline variables known to be associated with adenoma recurrence or colorectal cancer including; NSAIDS use, number of adenomas, villous histology, did not alter the findings. These Super Compliers, compared to the control group, had a significantly reduced OR for the recurrence of either an advancedadenoma or 3 or more adenomas (OR = 0.41; 95% CI, 0.021-0.81). The reduced recurrence rate among super-compliers does suggest that a low-fat, high-fiber, fruit- and vegetable-enhanced eating pattern reduces colorectal adenoma occurrence. This was an 'according to protocol', as opposed to 'intention to treat' analysis, however, and it is possible that (unknown or unmeasured) factors associated with adherence, rather than diet per se, are responsible for the change in neoplastic recurrence.

[Fifth AACR International Conference on Frontiers in Cancer Prevention Research, Nov 12-15, 2006]