Sugar-Sweetened Beverage Intake and Cancer Recurrence and Survival in CALGB 89803 (Alliance)

Background In colon cancer patients, obesity, sedentary lifestyle, and high dietary glycemic load have been associated with increased risk of cancer recurrence. High sugar-sweetened beverage intake has been associated with obesity, diabetes, and cardio-metabolic diseases, but the influence on colon cancer survival is unknown. Methods We assessed the association between sugar-sweetened beverage consumption on cancer recurrence and mortality in 1,011 stage III colon cancer patients who completed food frequency questionnaires as part of a U.S. National Cancer Institute-sponsored adjuvant chemotherapy trial. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated with Cox proportional hazard models. Results Patients consuming ≥2 servings of sugar-sweetened beverages per day experienced an adjusted HR for disease recurrence or mortality of 1.67 (95% CI, 1.04–2.68), compared with those consuming <2 servings per month (P trend = 0.02). The association of sugar-sweetened beverages on cancer recurrence or mortality appeared greater among patients who were both overweight (body mass index ≥25 kg/m2) and less physically active (metabolic equivalent task-hours per week <18) (HR = 2.22; 95% CI, 1.29–3.81, P trend = 0.0025). Conclusion Higher sugar-sweetened beverage intake was associated with a significantly increased risk of cancer recurrence and mortality in stage III colon cancer patients.


Introduction
Increasing evidence indicates that excess energy balance, including obesity, sedentary lifestyle, and high intake of a Western pattern diet, is associated with an increased risk of developing colon cancer. [1][2][3][4] One hypothesis explaining these observations has been the role of hyperinsulinemia in promoting cancer development and progression. [5,6] Prospective studies demonstrate associations between colon cancer risk among individuals with diabetes or elevated circulating levels of insulin or C-peptide. [6][7][8] Beyond cancer risk, observational studies of colon cancer patients suggest that these same host factors, including obesity, sedentary lifestyle, high intake of a Western pattern diet, and high dietary glycemic load confer an increased risk of colon cancer recurrence [9][10][11].
Across numerous studies, higher intake of sugar-sweetened beverages has been associated with an increased risk of obesity, type 2 diabetes, and related cardio-metabolic diseases. [12][13][14][15][16][17] Consumption of sugar-sweetened beverages has increased in the U.S. over the past 3 decades; per capita consumption in 2009 was estimated at 45 gallons per year, or nearly half of total beverage intake. [18] Worldwide, sugar-sweetened beverages have been associated with 180,000 deaths annually. [19] Nonetheless, the influence of sugar-sweetened beverage intake on colon cancer risk or colon cancer patient outcomes remains largely unknown.
In light of the emerging evidence supporting a link between excess energy balance, hyperinsulinemia and inferior survival in colon cancer patients, we prospectively examined the impact of sugar-sweetened beverage intake on cancer recurrence and mortality in a cohort of stage III colon cancer patients enrolled in a National Cancer Institute-sponsored adjuvant chemotherapy clinical trial. We prospectively collected extensive data on diet and lifestyle prior to any documentation of cancer recurrence.

Study Population
Patients in this prospective cohort participated in the National Cancer Institute-sponsored Cancer and Leukemia Group B (CALGB, now Alliance for Clinical Trials in Oncology) 89803 adjuvant therapy trial for stage III colon cancer, comparing therapy with weekly 5-fluorouracil and leucovorin to weekly irinotecan, 5-fluorouracil, and leucovorin (ClinicalTrials.gov NCT000038350). [20] Between May 1999 and May 2001, 1,264 patients were enrolled in the treatment trial. After 87 patients were enrolled, an amendment required patients to complete a selfadministered questionnaire that captured diet and lifestyle habits midway through their therapy (4 months after surgery; questionnaire 1 [Q1]) and again 6 months after completion of treatment (14 months after surgery; Q2).
Patients were eligible if they underwent a complete surgical resection of the primary tumor within 56 days of trial entry, had regional lymph node metastases but no evidence of distant metastases, had a baseline Eastern Cooperative Oncology Group performance status of 0 to 2, [21] and had adequate bone marrow, renal, and hepatic function. Figure 1 illustrates the derivation of the final sample size of 1,011 patients for this study. We previously demonstrated that there were no appreciable differences in baseline characteristics between the 1,011 patients who were eligible for this analysis and the remaining 253 patients enrolled in CALGB 89803 not included in the dietary analysis. [22] All patients signed informed consent, which was approved by the National Cancer Institute's Cancer Treatment Evaluation Program (CTEP) as well as each participating site's institutional review board (IRB), including the Dana-Farber/Harvard Cancer Center IRB.

Dietary Assessment
Patients in this analysis completed semi-quantitative food frequency questionnaires (FFQs) that included 131 food items, vitamin and mineral supplements, and open-ended sections for other supplements and foods not specifically listed. [23,24] Participants were asked how often, on average over the previous 3 months, they consumed a specific food portion size, with up to nine possible responses, which ranged from never to six or more times per day. We computed nutrient intakes by multiplying the frequency of consumption of each food by the nutrient content of the specified portions using composition values from Department of Agriculture sources supplemented with other data [25] All nutrient values were energy-adjusted using the residuals methods [26].
On each FFQ, patients reported on their usual daily or weekly intake of a standard 355-mL (12-ounce) serving (1 glass, can, or bottle) of sugar-sweetened caffeinated colas, caffeine-free colas, other carbonated sugar-sweetened beverages, and noncarbonated sugar-sweetened beverages (fruit punches, lemonades, or other fruit drinks). We summed the consumption of these beverages as total sugar-sweetened beverages. In other cohort studies, the reproducibility and validity of our FFQ in measuring sugarsweetened beverages was previously described: the Pearson correlation coefficient, corrected for within-person variation, between the FFQ and separate dietary records was 0.84 for cola and 0.36-0.40 for non-cola soda [27] In addition, we separately validated this FFQ among cancer patients receiving chemotherapy [28].
Patients who completed Q1 and whose cancer had not recurred prior to Q1 completion were included in these analyses. The median time from study entry to Q1 was 3.5 months (95% range, 2.5-5.0 months; full range, 0.2-9.9 months). To avoid potential biases due to declining health immediately before cancer recurrence or death, we excluded patients who experienced cancer recurrence or death within 90 days following completion of Q1 ( Figure 1). We calculated dietary exposures based on the results of the Q2 using cumulative averaging as previously described, [29] but weighted proportional to times between Q1 and Q2 and then between Q2 and disease-free survival time. For example, if a patient completed Q1 at 4 months, completed Q2 at 14 months, and had a cancer recurrence at 30 months, the total time between Q1 and cancer recurrence was 26 months and 38% of that time was between Q1 and Q2 and 62% of that time was between Q2 and the recurrence.

Study Endpoints
The primary end point of our study was disease-free survival, defined as time from the completion of the Q1 to tumor recurrence, occurrence of a new primary colon tumor, or death from any cause. We also assessed recurrence-free survival, defined as the time from the completion of Q1 to tumor recurrence or occurrence of a new primary colon tumor. For recurrence-free survival, patients who died without known tumor recurrence were censored at last documented evaluation by treating physician. Finally, overall survival was defined as the time from the completion of Q1 to death from any cause.

Statistical Analysis
Sugar-sweetened beverage intake was classified by frequency of 12-ounce servings: less than 2 per month (reference), 2 per month to 2 per week, 3 to 6 per week, 1 per day to less than 2 per day, and 2 or more per day. Cox proportional hazards regression was used to determine the simultaneous impact of other variables potentially associated with each outcome. [30] As previously described, [26] all analyses of sugar-sweetened beverage intake were adjusted for total caloric intake to assess the effect of sugarsweetened beverages independent of total energy intake. We further used time-varying covariates to adjust for total calories, physical activity (measured in metabolic equivalent tasks [MET]hours per week), [31] body mass index (BMI, calculated as weight in kilograms divided by height in meters squared), and consump-tion of Western and prudent pattern diets [10] with updating from the second questionnaire. Other covariates (including age at study entry, gender, number of positive lymph nodes, depth of invasion through bowel wall, baseline performance status, and treatment group) were also entered into the model as fixed covariates. In secondary multivariate analyses, we further adjusted for dietary glycemic load. [22] Covariates with missing variables were coded with indicator variables in adjusted models. We tested for linear trends across categories of sugar-sweetened beverages by assigning each participant the median value for each category and modeling this value as a continuous variable, consistent with prior studies. [10,22] In subgroup exploratory analyses, the five intake categories were collapsed (1 and 2; 3 and 4; 5) to conserve power to create 3 groupings with practical meaning -less than twice a week, several times per week to daily and more than 2 per day. The Cox regression models met the assumption of proportionality. A level of significance ,0.05 was considered statistically significant. All P values are 2-sided and were not adjusted for multiple comparisons. Patient registration and clinical data collection were managed and analyses were conducted by the Alliance Statistics and Data Center. All analyses were based on the study database frozen on November 9, 2009.

Results
Study participants were drawn from a multicenter study of adjuvant chemotherapy after surgery in patients with stage III colon cancer. Baseline characteristics by level of sugar-sweetened beverage consumption are shown in Table 1. Patients consuming greater quantities of sugar-sweetened beverages were younger and more likely to be male and had a higher intake of a Western pattern diet and dietary glycemic load.
The median follow-up from the time of completion of Q1 was 7.3 years. In total, 343 of the 1,011 patients included in this analysis had cancer recurrence; 262 of these 343 patients died. An additional 43 patients died without documented cancer recurrence.
The pre-defined primary endpoint of our analysis was diseasefree survival. Increasing consumption of sugar-sweetened beverages was associated with a greater risk of cancer recurrence or mortality after adjusting for other predictors of cancer recurrence ( Table 2). Compared to patients who drank less than two sugarsweetened beverage per month, consumers of 2 or more servings per day experienced an adjusted hazard ratio (HR) for disease recurrence or mortality of 1.67 (95% confidence interval [CI] = 1.04-2.68; P trend = 0.02). These results were largely unchanged when we further adjusted for dietary glycemic load in the multivariate model (P trend = 0.03).
To isolate the influence of sugar-sweetened beverage intake on cancer recurrence, we assessed the endpoint of recurrence-free survival and confirmed that higher intake of sugar-sweetened beverage conferred a significantly increased risk in cancer recurrence (P trend = 0.02; Table 2). Patients consuming 2 or more servings per day were 1.84 times more likely to recur than those who largely abstained. We observed a trend toward greater overall mortality with consumption of 2 or more sugar-sweetened beverages per day, although the results were not statistical significant (P trend = 0.12).
Taking into account the possibility that dietary changes could reflect occult cancer or imminent death, we excluded patients who developed cancer recurrence or died within 90 days of completing Q1 in our initial analyses. To further explore the potential influence of occult cancer or imminent death, we excluded 44 patients who developed cancer or died within 180 days of completing Q1; our results remained largely unchanged. Increasing sugar-sweetened beverage intake was associated with an inferior disease-free survival (P trend = 0.04) and an inferior recurrence-free survival (P trend = 0.04).
We explored whether the association between sugar-sweetened beverage intake and patient outcome differed between individual sugar-sweetened beverages subtypes (colas; other carbonated beverages; and non-carbonated fruit drink and punches). The association between consumption of specific sugar-sweetened beverages and survival endpoints did not differ between beverage subtypes; tests for heterogeneity between subtypes was not statistically significant for disease-free survival (P = 0.65), recurrence-free survival (P = 0.59), or overall survival (P = 0.71), though power for these analyses is limited.
We also examined the influence of sugar-sweetened beverage intake on disease-free survival across strata of other potential predictors of patient outcome (Figure 2), using three categories of intake (2 or less beverages per week, between 3 beverages per week and less than 2 per day and 2 or more beverages per week). In the forest plot, we demonstrate the hazard ratio comparing the two extreme categories. The association of sugar-sweetened beverage intake was not significantly modified by any potential predictors of patient outcome.
We further explored whether the association between sugarsweetened beverage intake and the risk for cancer recurrence or mortality was more pronounced among patients who were both overweight (BMI$25 kg/m 2 ) and less physically active (,18 MET-hours/week) ( Figure 3). Among patients who were both overweight and less active (n = 541), the adjusted HR for cancer recurrence or mortality was 2.22 (95% CI, 1.29-3.81) when comparing the highest and the lowest categories of sugarsweetened beverage intake (P trend = 0.0025). In contrast, among the 84 patients who were both normal weight (,25 kg/m 2 ) and physically active (MET-hours per week $18), sugar-sweetened beverage intake did not significantly influence the risk of cancer recurrence or mortality (P trend = 0.67). A test for statistical interaction between sugar-sweetened beverage intake and the combined assessment of BMI and physical activity was, however, not significant (P interaction = 0.40).

Discussion
In this cohort of stage III colon cancer patients enrolled in a clinical trial of postoperative adjuvant chemotherapy, increased intake of sugar-sweetened beverages was associated with a significantly increased risk of cancer recurrence or mortality, and the deleterious association of sugar-sweetened beverages was largely confined to patients who consumed 2 or more servings per day. The association between sugar-sweetened beverage consumption and patient outcome was independent of other measures of energy balance, including body-mass index, Western pattern diet, and dietary glycemic load.
To our knowledge, no previous study has examined the association of sugar-sweetened beverage intake on clinical outcome in patients with established colon cancer. Two studies have examined the relation between sugar-sweetened beverage intake and the risk of developing colon cancer, and the results have been inconsistent. [32,33] In otherwise healthy cohorts, sugarsweetened beverages have been linked with weight gain, diabetes, hypertension, hyperlipidemia, gout, and coronary artery disease. [13][14][15][16][17]34,35] In randomized clinical trials, replacement of sugarcontaining beverage with a sugar-free beverage significantly reduced weight gain and fat accumulation in children and adolescents. [36,37] In a recent observational study of healthy adults, the association between sugar-sweetened beverage and obesity was significantly greater in subjects with a genetic predisposition to obesity. [38] Interestingly, sugar-sweetened beverage intake was associated with significant increase in cancer recurrence and mortality among patients who were both overweight and less physically active, whereas there did not appear to be an association among patients who were normal weight and physically active.
Previous observational studies of colon cancer patients have observed increased risks of cancer recurrence or mortality in association with a history of type 2 diabetes, obesity, sedentary lifestyle, high Western pattern diet, and greater dietary glycemic load. [9][10][11] While the mechanisms accounting for these observations remain uncertain, chronic hyperinsulinemia, promoting increased cancer cell proliferation and reduced apoptosis, has been proposed as one putative explanation. [39][40][41][42] In support of the hypothesis, higher circulating C-peptide was associated with increased cancer mortality in a population of stage I-III colorectal cancer patients [11].     Sugar-sweetened beverages provide a high caloric content with low satiety and incomplete compensation for these liquid calories, resulting in an increased total energy intake. [43,44] In addition, because of the large amounts of rapidly absorbable carbohydrates in sugar-sweetened beverages, greater consumption may increase the risks of insulin resistance, beta-cell dysfunction, inflammation, visceral adiposity, and other metabolic disorders. [13,45] In a prior analysis of our colon cancer patient cohort, neither total sucrose nor fructose intake was significantly associated with patient outcome. [22] Nonetheless, sucrose or fructose from beverages and processed foods (e.g., sweets, cookies) are more rapidly absorbed and may have a greater detrimental effect than sugars from fruit or other natural sources [13,45].
Our analysis in patients treated within a U.S. National Cancer Institute-sponsored clinical trial provided several advantages. All patients had lymph node-positive, non-metastatic cancer at study entry, reducing the impact of heterogeneity by disease stage. Treatment and patient follow-up were carefully prescribed in the trial, and the date and nature of cancer recurrence were prospectively recorded through regular detailed medical examinations. Detailed information on other potentially confounding prognostic variables was also prospectively collected at study entry. Finally, we updated dietary data to reflect changes in diet that may have occurred after patients had completed adjuvant therapy and recovered from treatment effects. Repeated dietary intake data served to reduce random measurement error, and any residual random error would likely lead to an underestimate of the true effect of exposures with outcome.
Nonetheless, we do recognize limitations within our analysis. Patients who enroll in randomized trials may differ from the population at large. However, the distribution of dietary intake and lifestyle factors in our cohort was consistent with other U.S. cohort studies. [46] Moreover, since the study included patients from community and academic centers throughout North America, our findings should reflect the general population.
We cannot completely exclude the possibility that increased sugar-sweetened beverage consumption may simply reflect other predictors of poor cancer prognosis. However, the effect of sugarsweetened beverage was independent of other known or suspected prognostic factors, and our results remained largely unchanged when we further adjusted for other energy balance factors, including BMI, physical activity, Western and prudent pattern Figure 2. Multivariate hazard ratios and 95% CI for cancer recurrence and mortality across strata of various factors. Analyses utilized 3 categories (2 or less servings per week, between 3 servings per week and less than 2 servings per day and 2 or more serving per day. The forest plot represents the hazard ratios of the comparison of $2 servings per day of sugar-sweetened beverage intake to #2 serving per week. No. = Number; y = years; kg = kilograms; m = meters; wk = weeks; 5-FU/LV = 5-fluorouracil/leucovorin; IFL = bolus 5-fluorouracil, leucovorin, and irinotecan. doi:10.1371/journal.pone.0099816.g002 diets, and dietary glycemic load. Further, the detrimental effect of sugar-sweetened beverage intake persisted across strata of prognostic variables and dietary/lifestyle factors. Nevertheless, because of the observational nature of this study, we cannot exclude the possibility of residual and unmeasured confounding variable. Finally, while the original companion study for CALGB 89803 did pre-specify analyses of associations of dietary factors associated with energy balance and outcomes, caution should also be exercised when interpreting studies with multiple comparisons and P values.0.01. However, the point estimates are consistent across strata, and results are consistent with other analyses of energy balance in this cohort. [10,22,47,48] Nonetheless, these results require confirmation in other cohorts.
We considered the possibility that patients with either occult cancer recurrences or other statistically significant poor prognostic characteristics may have increased their sugar-sweetened beverage intake as an alternative source of needed calories. To minimize this bias, we excluded recurrences or deaths within 90 days of FFQ completion. When we extended this restriction to 6 months, we continued to observe a deleterious association of sugar-sweetened beverage consumption on patient outcomes. Finally, because patients on this clinical trial underwent comprehensive staging at study entry and were followed with prescribed follow-up visits and testing, we would expect few patients to have undetected recurrences over extended periods.
Given that patients who consume sugar-sweetened beverages after cancer diagnosis may have consumed a similar diet before diagnosis, we cannot exclude the possibility that consumers of sugar-sweetened beverages develop biologically more aggressive colon cancers. Nonetheless, we did not observe a meaningful association between sugar-sweetened beverage intake and tumorrelated characteristics associated with cancer recurrence.
In summary, this prospective analysis, imbedded in a clinical trial, suggests that increased sugar-sweetened beverage intake is associated with a significantly worse disease-free and recurrencefree survival for stage III colon cancer patients. Although our observational study does not provide conclusive evidence for causality, these findings support the potential role of sugarsweetened beverage intake in colon cancer progression and potentially offer meaningful recommendations for clinical care. Further analyses in other cohorts of colon cancer survivors are needed to confirm these findings.