Revised Abstract

1. The impact of dietary patterns on cancer recurrence and survival in patients with stage III colon cancer: findings from CALGB 89803 Jeffrey A. Meyerhardt1, Donna Niedzwiecki2, Donna Hollis2, Leonard B. Saltz3, Walter Willett4, Robert J. Mayer1, Charles S. Fuchs1 1Dana-Farber Cancer Institute, Boston, MA; 2CALGB Statistical Center, Durham, NC; 3Memorial Sloan-Kettering Cancer Center, New York, NY; 4 Harvard School of Public Health, Boston, MA

2. Revised Abstract Context: Dietary factors have been associated with the risk of developing colon cancer, however, the influence of diet on patients with established disease is unknown. Objective: Determine influence of dietary patterns on cancer recurrences & mortality of colon cancer survivors. Design, Setting, Patients: We conducted a prospective observational study of 1,009 patients with stage III colon cancer enrolled in a randomized adjuvant chemotherapy trial. Patients reported on dietary intake using a semi-quantitative food frequency questionnaire during and 6 months after adjuvant chemotherapy. We identified two major dietary patterns, prudent and Western,by factor analysis. The prudent pattern was characterized byhigh fruit and vegetable, poultry and fish intakes; the Western pattern was characterizedby high meat, fat, refined grains and dessert intakes. Patients were followed for cancer recurrence or death. Results: A higher intake of a Western pattern diet after cancer diagnosis was associated with a significantly increased risk of colon cancer recurrence or death (disease-free survival). Compared with patients in the lowest quintile of Western pattern diet, those in the highest quintile experienced an adjusted hazard ratio for recurrence or death of 3.91 (95% confidence interval [CI], 2.21 – 6.89; P for trend<0.0001), Western pattern diet was associated with a similar detriment in recurrence-free survival (adjusted P for trend<0.0001) and overall survival (adjusted P for trend<0.0001). The reduction in disease-free survival with a Western pattern diet was not significantly modified by gender, age, nodal stage, body mass index, physical activity level, baseline performance status, or treatment arm. In contrast, the prudent pattern diet was not significantly related to cancer recurrence or mortality. Conclusion: Higher intake of a Western pattern diet may increase the risk of recurrence and mortality among patients with stage III colon cancer patients treated with surgery and adjuvant chemotherapy. Further studies are needed to delineate which components of such a diet are most influential.

3. Background • Patients diagnosed with cancer are highly motivated to seek information about diet, physical activity, dietary supplement use, and nutritional complementary therapies.1-4 • Epidemiologic and scientific research indicates that diet and other lifestyle factors have a significant influence on the risk of developing colon cancer. 5,6 • The influence of diet on colon cancer recurrences is unknown • Factor analysis has been used to examine overall dietary patterns, beyond individual foods and nutrients, and the risk of several cancers.7-8 • Specific dietary patterns have been associated with the development of colorectal cancer in case-control and cohort studies.9-15

4. Methods We prospectively studied dietary patterns among 1,009 patients with stage III colon cancer enrolled in a randomized trial of post-operative adjuvant chemotherapy (CALGB 89803), which compared bolus 5-fluorouracil and leucovorin (5-FU/LV) to bolus irinotecan, 5-FU, and leucovorin (IFL) (Saltz et al. Proc ASCO 2004). Patients in these analyses completed semi-quantitative food frequency questionnaires (SFFQ) in the middle of their adjuvant chemotherapy (Q1) and approximately 6 months after the completion of adjuvant therapy (Q2). The questionnaire included 131 food items, vitamin and mineral supplements, and open‑ended sections for other supplements and foods not specifically listed. To identify dietary patterns, we applied factor analysis todata from the SFFQ in each cohort.Food items on the questionnaires were groupedinto 39 predefined food groups Patients who completed the first SFFQ were included in these analyses. Only patients who did not experiencerecurrence or die before Q1 were includedin these analyses. To avoid biases due to declining health immediately before recurrence or death, we also excluded from analysespatients who experienced either event within 90 days following thedietary assessment. We updated dietary exposures based on the results of the second SFFQ (Q2) using cumulative averaging, but weighted proportional to times between Q1 and Q2 and then Q2 and disease-free survival time.

5. Methods Factor analysis (principal component) was conducted using thefactor procedure in SAS software. Factor analysis aggregatescorrelated variables. We retained two factors and labeled these two factors as the"prudent" and "Western" patterns. The prudent patternwas characterized by high intakes of fruits, vegetables, wholegrains, legumes, poultry, and fish and the Western pattern was characterized by refined grains, processed and redmeats, desserts, high-fat dairy products, and french fries. These 2 factors are not correlated (correlation coefficient = 0.02). Factor scores were divided into quintiles for analyses. We computed Cox proportional hazards for recurrence-free (RFS), disease-free (DFS) and overall survival (OS). Time intervals were measured from completion of the 1st questionnaire to recurrence or death, excluding events within the first 90 days to minimize potential bias from decreased activity due to underlying disease. Median follow-up after the 1st questionnaire was 5.3 years. We used time-varying covariates to adjust for total calories, physical activity and body mass index with updating from Q2. Other covariates (including age at study entry, gender, number of positive lymph nodes, baseline performance status, presence of bowel perforation or obstruction at time of surgery, smoking history, treatment arm, and weight change between Q1 and Q2) were also entered into the model as fixed covariates. We tested for linear trends across quintiles of dietary pattern by assigning each participant the median value for the quintile and modeling this value as a continuous variable.

6. Derivation of Cohort Size Q1 = questionnaire 1 (midway through adjuvant therapy); Q2 = questionnaire 2 (6 months after completion of adjuvant therapy) * Calorie exclusion: Less than 600 calories or greater than 4,200 calories per day for men and less than 500 calories or greater than 3,500 calories per day for women.

7. Pearson correlation coefficients for the relationship between food intake and factors representing dietary patterns Food grouping Prudent Western Vegetables † 0.72 --- Leafy vegetables 0.71 --- Yellow vegetables 0.67 --- Cruciferous vegetables 0.65 --- Legumes 0.56 --- Fruit 0.55 -- Light salad dressing 0.48 --- Tomatoes 0.46 0.36 Garlic 0.39 --- Fish 0.46 --- Poultry 0.37 --- Fruit Juice 0.35 --- Whole grains 0.32 -- Low fat mayonnaise 0.31 -- Wine 0.19 --- Tea 0.16 --- Diet beverages --- -- High-fat dairy --- 0.67 Low-fat dairy --- 0.64 Food grouping Prudent Western Refined grains --- 0.60 Condiments --- 0.51 Red meat --- 0.53 Sweets and desserts --- 0.53 Margarine --- 0.50 Processed meat --- 0.45 Potatoes 0.17 0.45 Regular mayonnaise --- 0.35 Butter --- 0.33 French fries -0.16 0.37 Eggs --- 0.30 Snacks ‡ --- 0.36 Nuts --- 0.30 Coffee --- 0.29 Sugar beverages - 0.15 0.29 Beer --- 0.22 Cream soup or chowder 0.16 0.25 Pizza --- 0.26 Regular salad dressing 0.19 0.19 Liquor --- --- * values < 0.15 are not shown (---). † Vegetables other than yellow, cruciferous, or leafy-green vegetables. ‡ Potato, corn chips, crackers, or popcorn.

8. Table 2: Baseline characteristics * of 1,009 patients by Quintile of Dietary Pattern Baseline Characteristics by Quintile of Western Pattern Diet

9. Baseline Characteristics by Quintile of Prudent Pattern Diet

10. Impact of Western pattern diet on colon cancer recurrence and mortality

11. Impact of Prudent pattern diet on colon cancer recurrence and mortality

12. Subgroup analyses by quintile of Western pattern diet

13. In a cohort of patients with stage III colon cancer treated with surgery and adjuvant chemotherapy surviving without cancer recurrence 3 months after the completion of a FFQ, increasing consumption of a Western pattern diet after diagnosis was associated with an increased risk of cancer recurrence or death. Compared to patients in the lowest level, those in the highest level of Western pattern intake experienced a tripling in risk of recurrence or death. In contrast, prudent pattern diet did not influence the outcome after curative resection of stage III colon cancer. These data suggest that a diet characterized by higher intakes of red and processed meats, sweets and desserts, french fries, and refined grains increases the risk of cancer recurrence and decreases survival. Further analyses are underway to better delineate specific nutrients or food groupings that may be most influential. Conclusions

14. References