J Gastrointest Canc (2013) 44:203–210 DOI 10.1007/s12029-012-9472-4

ORIGINAL RESEARCH

Body Mass Index and Body Surface Area and Their Associations with Outcomes in Stage II and III Colon Cancer Sina Alipour & Hagen F. Kennecke & Ryan Woods & Howard J. Lim & Caroline Speers & Carl J. Brown & Sharlene Gill & Daniel J. Renouf & Winson Y. Cheung

Published online: 21 December 2012 # Springer Science+Business Media New York 2012

Abstract Background Our study aims were to measure the associations between body mass index (BMI) and body surface area (BSA) with outcomes for stage II and III colon cancer and to evaluate if the effect of obesity is modified by disease stage and receipt of adjuvant therapy. Methods Using a prospective cohort of stage II and III colon cancer patients who were referred between 2001 and 2005, we compared 3-year relapse-free survival (3-year RFS), 5year cancer-specific survival (5-year CSS), and 5-year overall survival (5-year OS) rates among different BMI and BSA categories. Cox proportional-hazards models were constructed to explore the relationships between different body compositions and outcomes while adjusting for confounders. Results Postoperative height and weight were used to classify 913 patients as normal weight (n0424, BMI <25 kg/ m2), overweight (n0319, BMI 25–30 kg/m2), and obese (n0 170, BMI >30 kg/m2). Using Mosteller formula, 684 subjects had normal BSA (≤2.0 m2) and 229 had high BSA (>2.0 m2). Obese subjects experienced similar 3-year RFS (61.9 vs. 66.5 vs. 63.6 %, p00.51), 5-year CSS (65.6 vs. S. Alipour : H. F. Kennecke : H. J. Lim : S. Gill : D. J. Renouf : W. Y. Cheung Department of Medicine, University of British Columbia, 2775 Laurel Street, Vancouver, BC, Canada V5Z 1M9 H. F. Kennecke : H. J. Lim : S. Gill : D. J. Renouf : W. Y. Cheung (*) Division of Medical Oncology, British Columbia Cancer Agency, Vancouver Clinic, 4415-600 West 10th Avenue, Vancouver, BC, Canada V5Z 4E6 e-mail: [email protected] R. Woods : C. Speers : C. J. Brown Cancer Surveillance and Outcomes, British Columbia Cancer Agency, 686 West Broadway, 8th Floor, Vancouver, BC, Canada V5Z 1G1

72.4 vs. 68.0 %, p00.22), and 5-year OS (60.8 vs. 64.0 vs. 62.2 %, p00.69) when compared to overweight subjects and those with normal BMIs, respectively. Likewise, individuals with high BSA had similar outcomes as those with normal BSA (66.2 vs. 63.6 %, p00.64 for 3-year RFS, 70.3 vs. 68.6 %, p00.62 for 5-year CSS, and 64.5 vs. 61.9 %, p0 0.48 for 5-year OS). In Cox models, advanced age, male gender, stage III disease, and poor performance status correlated with inferior RFS, CSS, and OS, but BMI and BSA did not. Conclusions Obesity as measured by either BMI or BSA was not associated with differences in outcomes in stage II and III colon cancer. Keywords Body mass index . Body surface area . Obesity . Colon cancer . Outcomes

Introduction Despite recent advances in diagnostic and therapeutic strategies, colon cancer continues to be a global health problem and remains one of the major causes of cancer-related morbidity and mortality in the developed world [1–3]. Large prospective studies have identified obesity as a risk factor for developing colon cancer [4–6]. In the Cancer Prevention Study II, researchers showed that a high body mass index (BMI) confers a 1.5 to 1.8 times increased risk of developing colon cancer when compared to a normal BMI [7]. Although the precise mechanisms underlying this association are unclear, altered regulation of hormones, such as insulin, insulin growth factor-1, and androgens, in the setting of obesity is hypothesized to potentiate mitogenesis and various anti-apoptotic pathways [8, 9]. The influence of obesity on disease recurrence and survival after colon cancer diagnosis and treatment is less clear.

204

In a retrospective analysis of clinical trial data, Meyerhardt et al. did not observe any statistically significant differences in mortality between colon cancer patients with a baseline BMI of >30 kg/m2 versus those with a BMI of 21–25 kg/m2 [10]. Conversely, among subjects enrolled in the National Surgical Adjuvant Breast and Bowel Project (NSABP) study, the group of clinical trial participants with a BMI >35 kg/m2 experienced a 27 % statistically significant increase in cancer recurrence and death when compared to normal-weight subjects [11]. Prior studies of obesity and colon cancer outcomes primarily included clinical trial subjects that may not reliably represent the general population. They also mainly focused on BMI as a measure of obesity; the effects of other measures of body composition, such as body surface area (BSA), have been less studied. Based on these observations, we conducted a population-based review to evaluate the association of obesity as measured by either BMI or BSA with relapse-free, cancer-specific and overall survival in a large cohort of patients with stage II and III colon cancer. We also assessed whether the effect of obesity on colon cancer outcomes was modified by the stage of disease and receipt of adjuvant chemotherapy.

Materials and Methods Characteristics of the Study Setting The British Columbia (BC) Cancer Agency is a provincewide agency that provides a population-based cancer control program for its residents. The Agency is comprised of five tertiary cancer centers that cover the entire province with a population of 4.4 million people. All of the centers offer ambulatory oncology clinics, chemotherapy suites, radiation facilities, and inpatient units for cancer patients. Annually, an estimated 15,000 to 20,000 new patients are referred to the BC Cancer Agency for management. All of the centers are actively involved in enrolling patients onto major oncology clinical trials. Description of the Study Population This study was conducted upon receiving full research ethics approval from the institutional review board at the BC Cancer Agency. Eligible subjects were identified from the prospective BC Cancer Agency Gastrointestinal Cancers Outcomes Unit (GICOU) database and included patients who were: (1) diagnosed with pathological stage II or III colon cancer in BC between 2001 and 2005; (2) residents of the province at the time of their cancer diagnosis; and (3) referred to the BC Cancer Agency for their cancer care. Patients must also have had their baseline height and weight

J Gastrointest Canc (2013) 44:203–210

prospectively recorded at their initial consultation visit prior to receipt of any systemic or radiation therapy, if any was given. Patients with stage I colon cancer were excluded since the primary treatment for this stage of disease is surgery, meaning that many are managed by surgeons alone without subsequent referrals to the BC Cancer Agency for evaluation of further treatment. Any patients who received neoadjuvant therapy were also removed from the study cohort. Of note, we included subjects who experienced a recurrence or death even if this occurred shortly after their original diagnosis of colon cancer so that we could explore the effect of body compositions on short- and long-term outcomes. Definitions of Dependent and Independent Variables The primary study endpoint was 3-year relapse-free survival (RFS), defined as the interval between the date of colon cancer diagnosis and the date of first tumor recurrence (local, regional, or distant), diagnosis of new primary tumor, or death from any cause. The secondary study endpoints were 5-year cancer-specific survival (CSS; time between date of colon cancer diagnosis and death from colon cancer) and 5-year overall survival (OS; time between date of colon cancer diagnosis and death from any cause). The main independent variables were BMI and BSA, which served as separate measures of obesity. BMI was calculated as the patient’s weight (in kilograms) divided by the patient’s height squared (in square meters). Subjects were subsequently assigned to three BMI categories: normal (<25 kg/m2), overweight (25–30 kg/m2), and obese (>30 kg/ m2). BSA was derived using Mosteller’s formula, which was defined as the square root of height (in centimeters) multiplied by weight (in kilograms) divided by 3,600 [12]. Subjects were classified into two BSA categories: ≤2.0 m2 (low) and >2.0 m2 (high) based on an approximate cutoff value that has been previously used and published [13]. In addition to the main dependent and independent variables, the following covariates, if available, were analyzed: patient demographics (e.g., age and gender), comorbidities (e.g., presence of ischemic heart disease, chronic kidney disease, liver disease), performance status [as defined by Eastern Cooperative Oncology Group (ECOG)], [14] disease characteristics (e.g., stage, grade, tumor location), receipt of systemic therapy or radiation, and type of regimen received. For tumor location, proximal was defined as cancers affecting the ascending and transverse colon, whereas distal included those in the descending and rectosigmoid colon. The type and duration of adjuvant chemotherapy are standardized across the province. All of these covariates as well as outcomes, such as tumor recurrence, diagnosis of new primary tumor, and death from colon cancer, and other causes, were captured prospectively by the GICOU.

J Gastrointest Canc (2013) 44:203–210

Statistical Analyses Baseline characteristics were summarized with descriptive statistics. Variations in baseline characteristics across the different BMI and BSA categories were evaluated using the Chi-square test for categorical variables and the Wilcoxon rank-sum test for continuous variables. Survival curves and 3-year RFS, 5-year CSS, and 5-year OS rates were estimated for groups based on BMI and BSA using the Kaplan–Meier method. Differences in survival outcomes were assessed with the log-rank test. Multivariate Cox proportional-hazards regression models were subsequently fitted to evaluate the relationships between different body compositions and outcomes while controlling for potentially confounding covariates, such as age, gender, stage, number of lymph nodes retrieved, and systemic therapy. We explored for potential effect modification by performing subset analyses based on stage of disease (stage II vs. III) and receipt of systemic therapy (yes vs. no). Furthermore, we carried out sensitivity analyses whereby (1) BMI and BSA were analyzed as continuous variables and (2) individuals who were underweight (BMI <20 kg/m2; n029) were excluded from our cohort since such low extremes of weight may have an impact on outcomes. These sensitivity analyses did not appreciably affect our results (data not shown), so only findings from the main analyses are presented. All tests were two-sided where a p value of <0.05 was considered statistically significant. SAS version 9.2 and the R statistical package were used to conduct all of the statistical analyses.

Results Characteristics of the Study Population A total of 913 stage II and III colon cancer patients were eligible and analyzed. Based on the BMI classification, 424 (46 %) patients were normal weight, 319 (35 %) were overweight, and 170 (19 %) were obese. In terms of BSA, 684 (75 %) patients had a BSA of ≤2.0 m2, while 229 (25 %) had a BSA of >2.0 m2. The median follow-up period was 6.9 years (IQR 5.2 to 8.5 years). At diagnosis, obese patients were more likely to be men (p00.002) and younger (p<0.001) when compared to over and normal-weight patients. They also had more advanced stage of disease (p 00.003) and received adjuvant chemotherapy more frequently (p < 0.001). The most frequently cited reasons for not offering or receiving adjuvant chemotherapy were patient preference, advanced age, and poor functional status. There were no statistically significant differences in the distribution of these reasons among the various BMI and BSA categories (p>0.05).

205

Obese patients were more likely to report hypertension (47 vs. 34 vs. 26 %, p<0.001), diabetes (22 vs. 16 vs. 10 %, p<0.001), and hyperlipidemia (20 vs. 15 vs. 8 %, p<0.001). There were similar differences in terms of age at diagnosis (p<0.001), gender (p<0.001), overall stage (p00.05), and receipt of chemotherapy (p<0.001) when comparing those with low versus high BSA. Likewise, those with a high BSA were more likely to have hypertension (40 vs. 30 %, p< 0.001), diabetes (23 vs. 12 %, p<0.001), and hyperlipidemia (21 vs. 10 %, p<0.001). Additional details regarding baseline characteristics across the different BMI and BSA categories are summarized in Tables 1 and 2, respectively. Relationship Between Obesity and Survival Rates We did not observe any significant differences among the three BMI categories for any of the outcomes (Table 3). Specifically, the 3-year RFS rate was 63.6 % in the normal, 66.5 % in the overweight, and 61.9 % in the obese (p00.51). Likewise, the 5-year CSS and 5-year OS rates were 68.0, 72.4, and 65.6 % (p00.22) and 62.2, 64.0, and 60.8 % (p0 0.69) in the normal, overweight, and obese groups, respectively. In terms of BSA, we also failed to detect any differences between groups for all 3 outcomes (Table 3). The 3year RFS rate was 66.2 % in the high BSA and 63.6 % in the low BSA group (p00.64). Similarly, the 5-year CSS and 5year OS rates in the high and low BSA categories were 70.3 and 68.6 % (p 00.62) and 64.5 and 61.9 % (p 00.48), respectively. Subgroup analyses were undertaken to examine outcomes based on groups defined by: those who did versus did not receive adjuvant systemic therapy; and those with stage II versus III colon cancer (Table 3). In these subgroup analyses, the impact of BMI on 3-year RFS rates was neither modified by stage of disease nor receipt of adjuvant chemotherapy. However, differences were observed for 5-year CSS and 5-year OS rates when we compared individuals who received adjuvant treatment versus those who did not. Specifically, obese patients had significantly worse 5-year CSS (68.1 vs. 76.6 vs. 57.1 %, p00.02) and 5-year OS (59.7 vs. 65.9 vs. 50.0 %, p00.05) than non-obese patients within the subset that did not receive adjuvant chemotherapy. In contrast, obese patients appeared to have comparable 5-year CSS (68.0 vs. 69.7 vs. 68.9 %, p00.99) and 5-year OS (64.4 vs. 62.7 vs. 65.1 %, p00.68) when adjuvant chemotherapy was used. Of note, the effect of BSA on all three outcomes was neither influenced by cancer stage or adjuvant therapy. Multivariate Associations Between Obesity and Outcomes In Cox proportional-hazards models, we adjusted for age, sex, tumor stage, ECOG performance status, and

206

J Gastrointest Canc (2013) 44:203–210

Table 1 Patient and tumor characteristics by body mass index categories By BMI Variable

Subgroup

All cases (N0913) Normal (N0424) Overweight (N0319) Obese (N0170) p value

Age at diagnosis (years)

Median (IQR) <70 70+ Female Male

68 (59–75) 55.1 % 44.9 % 44.8 % 55.2 %

68 (60–76) 53.5 % 46.5 % 50.9 % 49.1 %

70 (60–76) 49.8 % 50.2 % 38.9 % 61.1 %

64.5 68.8 31.2 40.6 59.4

II III T1/2 T3 T4 Unknown N0 N1 N2 Unknown Proximal Distal Colon, NOS None Yes Well Moderate Poor/undifferentiated

39.5 % 60.5 % 4.9 % 75.2 % 19.7 % 0.1 % 38.6 % 43.7 % 16.8 % 1% 53.8 % 45.9 % 0.3 % 41.6 % 58.4 % 10.5 % 71.6 % 16.3 %

44.1 % 55.9 % 4% 75.5 % 20.5 % 0% 42.9 % 42.2 % 13.7 % 1.2 % 58.7 % 41 % 0.2 % 48.3 % 51.7 % 11.3 % 70.5 % 17.5 %

39.2 % 60.8 % 6.6 % 75.9 % 17.2 % 0.3 % 38.9 % 41.4 % 19.4 % 0.3 % 54.9 % 44.8 % 0.3 % 39.8 % 60.2 % 11 % 70.8 % 15.7 %

28.8 % 71.2 % 4.1 % 73.5 % 22.4 % 0% 27.1 % 51.8 % 19.4 % 1.8 % 39.4 % 60 % 0.6 % 28.2 % 71.8 % 7.6 % 75.9 % 14.7 %

1.5 % 10 (6–14) 40.4 % 23.2 % 33.4 % 3%

0.7 % 10 (6–15) 38.2 % 24.1 % 35.6 % 2.1 %

2.5 % 9 (6–14) 42 % 23.2 % 31 % 3.8 %

1.8 % 9.5 (6–14) 42.9 % 21.2 % 32.4 % 3.5 %

Sex Overall stage (pathologic) T stage (pathologic)

N stage (pathologic)

Tumor location

Initial chemotherapy Differentiation

Unknown No. of regional nodes sampled Median (IQR) <8 9–12 >12 Unknown

(57–71) % % % %

<.01 <.01 <.01 <.01

0.34

<.01

<.01

<.01 0.59

0.42 0.63

Unknown/NX/TX and “Colon, NOS” are removed before performing statistical testing

administration of adjuvant therapy. As Table 4 illustrates, patients who were overweight and obese did not correlate with worse RFS [hazards ratio (HR) 0.89, 95 % confidence interval (CI) 0.72–1.10 and HR 1.03, 95 % CI 0.79–1.33, respectively, p00.47], CSS (HR 0.80, 95 % CI 0.61–1.05 and HR 1.05, 95 % CI 0.77–1.42, respectively, p00.20), and OS (HR 0.89, 95 % CI 0.71–1.11 and HR 1.02, 95 % CI 0.78–1.33, respectively, p00.51) when compared to patients who were normal weight. However, advanced age, male gender, stage III disease, poor performance status, and lack of administration of adjuvant therapy were each associated with poorer outcomes (all p<0.05). Similarly, Table 5 reveals that BSA did not correlate with outcomes in stage II and III colon cancer. Compared to individuals with a low BSA, those with a high BSA had a comparable RFS (HR 0.88, 95 % CI 0.69–1.11, p00.28),

CSS (HR 0.85, 95 % CI 0.64–1.14, p00.28), and OS (HR 0.89, 95 % CI 0.69–1.13, p00.33), even adjusting for confounders. As with the Cox models for BMI, advanced age, male gender, stage III disease, poor performance status, and lack of administration of adjuvant therapy were also independently associated with worse outcomes in the BSA models (all p<0.05).

Discussion Previous studies consistently demonstrate an association between obesity and an elevated risk of developing colon cancer [4–6]. Whether obesity also increases the risk of recurrence and death among patients who are already diagnosed with colon cancer is less clear. Using a large,

J Gastrointest Canc (2013) 44:203–210

207

Table 2 Patient and tumor characteristics by high and low body surface area Variable

Age at diagnosis (years)

Sex Overall stage (pathologic) T stage (pathologic)

N stage (pathologic)

Tumor location

Initial chemotherapy Differentiation

No. of regional nodes sampled

Subgroup

All cases (N0913)

By BSA ≤2.0 (N0684)

>2.0 (N0229)

p value

Median (IQR) <70 70+ Female Male

68 (59–75) 55.1 % 44.9 % 44.8 % 55.2 %

69.5 (61–76) 50.0 % 50.0 % 55.3 % 44.7 %

65 (57–71) 70.3 % 29.7 % 13.5 % 86.5 %

<.01

II III T1/2 T3 T4 Unknown N0 N1 N2 Unknown Proximal Distal Colon, NOS None Yes Well Moderate Poor/undifferentiated

39.5 % 60.5 % 4.9 % 75.2 % 19.7 % 0.1 % 38.6 % 43.7 % 16.8 % 1.0 % 53.8 % 45.9 % 0.3 % 41.6 % 58.4 % 10.5 % 71.6 % 16.3 %

41.4 % 58.6 % 4.7 % 74.6 % 20.6 % 0.1 % 40.4 % 43.4 % 15.2 % 1.0 % 58.6 % 40.9 % 0.4 % 44.4 % 55.6 % 10.4 % 70.3 % 17.5 %

34.1 % 65.9 % 5.7 % 77.3 % 17.0 % 0.0 % 33.2 % 44.5 % 21.4 % 0.9 % 39.3 % 60.7 % 0.0 % 33.2 % 66.8 % 10.9 % 75.5 % 12.7 %

0.05

Unknown Median (IQR) <8 9–12 >12 Unknown

1.5 % 10 (6–14) 40.4 % 23.2 % 33.4 % 3.0 %

1.8 % 10 (6–14) 40.5 % 23.5 % 33.2 % 2.8 %

0.9 % 10 (7–14) 40.2 % 22.3 % 34.1 % 3.5 %

<.01 <.01

0.44

0.04

<.01

<.01 0.20

0.72 0.93

Unknown/NX/TX and “Colon, NOS” are removed before performing statistical testing

prospective, population-based cohort, our study suggests that obesity as measured by either BMI or BSA does not appear to influence RFS, CSS, and OS in patients with stage II and III colon cancer. Subgroup analyses demonstrated that among those who did not receive adjuvant chemotherapy, however, obesity was associated with worse 5-year CSS and 5-year OS rates. The effect of BMI on 3-year RFS and the influence of BSA on all three of the outcomes do not appear to be modified by stage of disease or receipt of systemic therapy. One of the strengths of the current analysis is its reliable ascertainment of three important outcomes, including RFS, CSS, and OS. Given that the primary aim of surgery and adjuvant therapy in stage II and III colon cancer is to prolong survival by preventing or delaying recurrences, RFS is the most sensitive outcome measure. Unlike CSS

and OS, RFS is less affected by potential disparities in the treatment of recurrent disease, variations in the management of comorbid conditions, and differential rates of death from competing causes that are unrelated to colon cancer. This is relevant since obesity can be independently associated with other medical conditions, such as heart disease and diabetes, which can themselves lead to increased mortality. In general, our current study of non-clinical trial patients appears to support prior research that explored BMI as a marker of obesity. In a retrospective analysis of clinical trial participants with stage II and III colon cancer, for example, no significant differences were observed in CSS or overall mortality between obese patients and those who were normal weight [15]. In another secondary analysis of stage III colon cancer patients enrolled in a randomized trial of adjuvant chemotherapy, both BMI and weight were not

208

J Gastrointest Canc (2013) 44:203–210

Table 3 Survival rates by BMI and BSA categories and stratified by stage and adjuvant therapy Outcome

Subgroup

BMI

BSA

Normal BMI

High BMI

Obese

p value

BSA ≤2.0

BSA >2.0

p value

5-year OS

All patients No adjuvant systemic therapy Adjuvant systemic therapy Stage II Stage III

62 (57, 60 (53, 64 (58, 69 (62, 57 (51,

67) 67) 71) 76) 64)

64 (59, 66 (57, 63 (56, 71 (62, 60 (53,

70) 75) 70) 79) 67)

61 (53, 50 (35, 65 (56, 65 (51, 59 (50,

69) 65) 74) 79) 68)

0.69 0.05 0.68 0.30 0.59

62 (58, 60 (54, 64 (59, 68 (63, 57 (52,

66) 65) 69) 74) 62)

65 (58, 65 (53, 65 (57, 71 (60, 62 (53,

71) 76) 73) 82) 70)

0.48 0.34 0.90 0.74 0.38

5-year CSS

All patients No adjuvant systemic therapy Adjuvant systemic therapy Stage II Stage III All patients No adjuvant systemic therapy Adjuvant systemic therapy Stage II Stage III

68 (63, 68 (61, 68 (62, 77 (71, 61 (55, 64 (59, 62 (55, 65 (58, 71 (64, 58 (51,

73) 75) 75) 84) 68) 68) 69) 71) 78) 64)

72 (67, 77 (69, 70 (63, 82 (75, 66 (59, 67 (61, 73 (65, 62 (55, 77 (70, 60 (52,

78) 85) 77) 90) 73) 72) 81) 69) 85) 67)

66 (58, 57 (42, 69 (60, 71 (57, 64 (55, 62 (54, 64 (50, 61 (52, 68 (54, 60 (51,

73) 72) 78) 85) 73) 69) 78) 70) 82) 69)

0.22 0.02 0.99 0.07 0.60 0.51 0.09 0.85 0.22 0.73

69 (65, 68 (62, 69 (64, 77 (72, 63 (58, 64 (60, 64 (59, 63 (58, 72 (66, 58 (53,

72) 74) 74) 82) 68) 67) 70) 68) 77) 63)

70 (64, 76 (65, 68 (60, 82 (72, 65 (57, 66 (60, 74 (64, 63 (55, 77 (68, 61 (53,

77) 87) 76) 92) 73) 73) 84) 70) 87) 69)

0.62 0.19 0.70 0.37 0.68 0.64 0.35 0.74 0.81 0.55

3-year RFS

significantly associated with any differences in risk of cancer recurrence or death [10]. Of note, our study findings contrast the results from a pooled analysis of NSABP randomized trials of patients with Dukes B and C colon cancer where the very obese group (BMI >35 kg/m2) experienced greater risk of non-cancer-related deaths and worse OS [11]. The higher BMI cutoff value that was used to define the very obese group along with the longer follow-up period (11.2 years) in the NSABP study may explain the differences that were observed. Interestingly, our findings also contrast those of other studies that used different proxies for obesity. Tartter et al., for instance, investigated the Quetelet index, which was defined in their study as weight multiplied by 100 divided by height squared, and reported significant differences in recurrence rates based on this index among women with stage II or III colon cancer [16]. Likewise, Haydon et al. found that an increase of 10 cm in waist circumference was associated with a 20 % increase in the risk of recurrence or death from colon cancer [17]. These results highlight that other measures of obesity aside from BMI should be explored as potential prognostic factors in colon cancer. To this end, we investigated the use of BSA as a marker of obesity in the current analysis, but we failed to observe any differences in outcomes when we stratified our cohort based on low and high BSA. The prognostic value of body composition and obesity has been extensively studied in the context of other cancer types, and results have tended to be more consistent for these other cancers [18–21]. In a prospective evaluation of early stage breast cancer patients, for example, women

whose weights were above the 75th percentile value (>73 kg) and had a BMI >28 kg/m2 were at significantly higher risk of recurrence than others [18]. The higher levels of serum estrogen in obese women are proposed to pose a direct effect in stimulating the growth and recurrence of hormone-sensitive breast cancers [22, 23]. The role of this and other hormones in obesity, such as insulin and insulin growth factor-1, and their influence on the pathophysiology of colon cancer should be a focus of future research, which could more effectively elucidate potential links between obesity and colon cancer outcomes. A distinguishing feature of this study is that we explored for effect modification based on stage of disease and receipt of adjuvant chemotherapy. This revealed that obese patients had significantly worse 5-year CSS (68.1 vs. 76.6 vs. 57.1 %, p00.02) and 5-year OS (59.7 vs. 65.9 vs. 50.0 %, p00.05) than non-obese patients, but only among those that did not receive adjuvant chemotherapy, and not in the subset that received adjuvant treatment where 5-year CSS (68.0 vs. 69.7 vs. 68.9 %, p00.99) and 5-year OS (64.4 vs. 62.7 vs. 65.1 %, p00.68) were similar regardless of BMI. Since the influence of obesity on RFS was not similarly modified, the potential but likely explanation is that obese patients who were not offered chemotherapy were inherently less healthy and carried a worse overall prognosis than those who were able to tolerate treatment. This is one of the first studies that investigated both BMI and BSA as measures of obesity and evaluated their relationships with stage II and III colon cancer outcomes among non-clinical trial patients. However, our findings should be interpreted in the context of several limitations. First, this

J Gastrointest Canc (2013) 44:203–210

209

Table 4 Multivariate model for the effect of BMI on RFS, CSS, and OS

Table 5 Multivariate model for the effect of BSA on RFS, CSS, and OS

Variable

Variable

Relapse-free survival Age, <70 vs 70+ Sex, M vs F Stage, III vs II BMI, high vs normal BMI, obese vs normal Chemotherapy (Y vs N) ECOG, 2 vs 0–1 ECOG, 3–4 vs 0–1 ECOG, unknown vs 0-1 Cancer-specific survival Age, <70 vs 70+ Sex, M vs F Stage, III vs II BMI, high vs normal BMI, obese vs normal Chemotherapy (Y vs N) ECOG, 2 vs 0–1 ECOG, 3–4 vs 0–1 ECOG, unknown vs 0–1 Overall survival Age, <70 vs 70+ Sex, M vs F Stage, III vs II BMI, high vs normal BMI, obese vs normal Chemotherapy (Y vs N) ECOG, 2 vs 0–1 ECOG, 3–4 vs 0–1 ECOG, unknown vs 0–1

p value

HR estimate

HR 95 % CI

<0.01 <0.01 <0.01 0.47

1.34 1.50 1.59 0.89 1.03 0.72 1.40 2.05 1.13

(1.10, (1.23, (1.24, (0.72, (0.79, (0.55, (1.03, (1.35, (0.91,

0.04 <0.01

1.30 1.40

(1.02, 1.67) (1.10, 1.77)

<0.01 0.20

2.04 0.80 1.05 0.70 1.51 2.06 0.99

(1.49, (0.61, (0.77, (0.51, (1.06, (1.26, (0.75,

2.79) 1.05) 1.42) 0.95) 2.15) 3.37) 1.29)

1.55 1.56 1.61 0.89 1.02 0.69 1.42 2.04 1.05

(1.26, (1.27, (1.24, (0.71, (0.78, (0.53, (1.04, (1.34, (0.84,

1.91) 1.90) 2.08) 1.11) 1.33) 0.89) 1.95) 3.10) 1.31)

0.01 <0.01

0.02 <0.01

<0.01 <0.01 <0.01 0.51 <0.01 <0.01

1.65) 1.82) 2.05) 1.10) 1.33) 0.92) 1.91) 3.12) 1.39)

was a cohort study so we cannot definitively establish any causal relationships between obesity and survival since there may be residual confounding. Second, the study population only consisted of patients who were referred to the BC Cancer Agency for their cancer care so the results may not be generalizable to non-referred patients or other jurisdictions. It is possible that patients were non-referred because they were considered too frail for evaluation or further treatment. Finally, we used baseline height and weight at the patients’ initial consultation visit to calculate BMI and BSA, and thus, we were unable to characterize how changes in these measurements over time may affect outcomes. Not infrequently, patients may experience weight loss leading up to their diagnosis of colon cancer or in the postoperative setting. This could potentially underestimate the detrimental impact of obesity if previously overweight patients were classified as normal weight at the time of their initial consultation with the oncologist. Weight changes

Relapse-free survival Age, <70 vs 70+ Sex, M vs F Stage, III vs II BSA, high vs low Chemotherapy (Y vs N) ECOG, 2 vs 0–1 ECOG, 3–4 vs 0–1 ECOG, unknown vs 0–1 Cancer-specific survival Sex, M vs F Stage, III vs II BSA, high vs low Chemotherapy (Y vs N) ECOG, 2 vs 0-1 ECOG, 3–4 vs 0–1 ECOG, unknown vs 0–1 Overall survival Age, <70 vs 70+ Sex, M vs F Stage, III vs II BSA, high vs low Chemotherapy (Y vs N) ECOG, 2 vs 0–1 ECOG, 3–4 vs 0–1 ECOG, unknown vs 0–1

p value

HR estimate

HR 95 % CI

0.01 <0.01 <0.01 0.28 <0.01 <0.01

1.30 1.55 1.61 0.88 0.71 1.42 2.09 1.14

(1.06, (1.26, (1.25, (0.69, (0.55, (1.04, (1.38, (0.92,

<0.01 <0.01 0.28

1.44 2.19 0.85

(1.12, 1.86) (1.61, 2.99) (0.64, 1.14)

<0.01 <0.01

0.63 1.54 2.21 0.99

(0.47, (1.08, (1.35, (0.76,

0.84) 2.20v 3.60) 1.30)

<0.01 <0.01 <0.01 0.33 <0.01 <0.01

1.51 1.60 1.63 0.89 0.68 1.44 2.07 1.06

(1.22, (1.29, (1.26, (0.69, (0.53, (1.05, (1.36, (0.85,

1.86) 1.97) 2.10) 1.13) 0.89) 1.97) 3.14) 1.32)

1.59) 1.90) 2.08) 1.11) 0.92) 1.93) 3.17) 1.41)

Age effect was not significant in the model for CSS and thus was dropped from the model

have been shown to portend an increased risk of cancer recurrence and mortality in some studies [24, 25]. Therefore, future research should consider evaluating obesity longitudinally over the course of treatment and studying how such changes may influence colon cancer survival. In conclusion, obesity as measured by BMI or BSA does not appear to correlate with colon cancer recurrence or death. However, it is possible that alternate indicators of obesity, such as waist circumferences, may provide more prognostic value or that temporal changes in weight can better predict for outcomes. Considering the growing prevalence of obesity worldwide and the morbidity burden that can frequently result from obesity, this area warrants further investigation so that colon cancer care in this subset of the population can be optimized. Acknowledgments This project was supported by internal funds from the British Columbia Cancer Agency Gastrointestinal Cancers Outcomes Unit.

210 Financial/funding disclosure None

Conflict of Interest All authors of this manuscript do NOT have any actual or potential conflict of interest including any financial, personal, or other relationships with other people or organizations that would inappropriately influence our submitted work.

References 1. Siegel RL, Jemal A, Thun MJ, Hao Y, Ward EM. Trends in the incidence of colorectal cancer in relation to county-level poverty among blacks and whites. J Natl Med Assoc. 2008;100:1441–4. 2. Siegel RL, Jemal A, Ward EM. Increase in incidence of colorectal cancer among young men and women in the United States. Cancer Epidemiol Biomarkers Prev. 2009;18:1695–8. 3. Siegel RL, Ward EM, Jemal A. Trends in colorectal cancer incidence rates in the United States by tumor location and stage, 1992– 2008. Cancer Epidemiol Biomarkers Prev. 2012;21:411–6. 4. Moghaddam AA, Woodward M, Huxley R. Obesity and risk of colorectal cancer: a meta-analysis of 31 studies with 70,000 events. Cancer Epidemiol Biomarkers Prev. 2007;16:2533–47. 5. Renehan AG, Tyson M, Egger M, Heller RF, Zwahlen M. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet. 2008;371:569–78. 6. Ning Y, Wang L, Giovannucci EL. A quantitative analysis of body mass index and colorectal cancer: findings from 56 observational studies. Obes Rev. 2010;11:19–30. 7. Murphy TK, Calle EE, Rodriguez C, Kahn HS, Thun MJ. Body mass index and colon cancer mortality in a large prospective study. Am J Epidemiol. 2000;152:847–54. 8. Vincent AM, Feldman EL. Control of cell survival by IGF signaling pathways. Growth Horm IGF Res. 2002;12:193–7. 9. LeRoith D, Roberts Jr CT. The insulin-like growth factor system and cancer. Cancer Lett. 2003;195:127–37. 10. Meyerhardt JA, Niedzwiecki D, Hollis D, Saltz LB, Mayer RJ, Nelson H, et al. Impact of body mass index and weight change after treatment on cancer recurrence and survival in patients with stage III colon cancer: findings from Cancer and Leukemia Group B 89803. J Clin Oncol. 2008;26:4109–15. 11. Dignam JJ, Polite BN, Yothers G, Raich P, Colangelo L, O’Connell MJ, et al. Body mass index and outcomes in patients who receive adjuvant chemotherapy for colon cancer. J Natl Cancer Inst. 2006;98:1647–54.

J Gastrointest Canc (2013) 44:203–210 12. Mosteller RD. Simplified calculation of body-surface area. N Engl J Med. 1987;317:1098. 13. Tang PA, Heng DY, Choueiri TK. Impact of body composition on clinical outcomes in metastatic renal cell cancer. Oncologist. 2011;16:1484–6. 14. Oken MM, Creech RH, Tormey DC, Horton J, Davis TE, McFadden ET, et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol. 1982;5:649–55. 15. Meyerhardt JA, Tepper JE, Niedzwiecki D, Hollis DR, McCollum AD, Brady D, et al. Impact of body mass index on outcomes and treatment-related toxicity in patients with stage II and III rectal cancer: findings from Intergroup Trial 0114. J Clin Oncol. 2004;22:648–57. 16. Tartter PI, Slater G, Papatestas AE, Aufses Jr AH. Cholesterol, weight, height, Quetelet’s index, and colon cancer recurrence. J Surg Oncol. 1984;27:232–5. 17. Haydon AM, Macinnis RJ, English DR, Giles GG. Effect of physical activity and body size on survival after diagnosis with colorectal cancer. Gut. 2006;55:62–7. 18. Hebert JR, Augustine A, Barone J, Kabat GC, Kinne DW, Wynder EL. Weight, height and body mass index in the prognosis of breast cancer: early results of a prospective study. Int J Cancer. 1988;42:315–8. 19. Rodriguez C, Patel AV, Calle EE, Jacobs EJ, Chao A, Thun MJ. Body mass index, height, and prostate cancer mortality in two large cohorts of adult men in the United States. Cancer Epidemiol Biomarkers Prev. 2001;10:345–53. 20. Petrelli JM, Calle EE, Rodriguez C, Thun MJ. Body mass index, height, and postmenopausal breast cancer mortality in a prospective cohort of US women. Cancer Causes Control. 2002;13:325– 32. 21. Rodriguez C, Calle EE, Fakhrabadi-Shokoohi D, Jacobs EJ, Thun MJ. Body mass index, height, and the risk of ovarian cancer mortality in a prospective cohort of postmenopausal women. Cancer Epidemiol Biomarkers Prev. 2002;11:822–8. 22. Vona-Davis L, Rose DP. Adipokines as endocrine, paracrine, and autocrine factors in breast cancer risk and progression. Endocrinol Relat Cancer. 2007;14:189–206. 23. Stoll BA. Oestrogen/insulin-like growth factor-I receptor interaction in early breast cancer: clinical implications. Ann Oncol. 2002;13:191–6. 24. Kroenke CH, Chen WY, Rosner B, Holmes MD. Weight, weight gain, and survival after breast cancer diagnosis. J Clin Oncol. 2005;23:1370–8. 25. Cleveland RJ, Eng SM, Abrahamson PE, Britton JA, Teitelbaum SL, Neugut AI, et al. Weight gain prior to diagnosis and survival from breast cancer. Cancer Epidemiol Biomarkers Prev. 2007;16:1803–11.