- Email: [email protected]
DOES BODY MASS INDEX AFFECT SURVIVAL OF PATIENTS UNDERGOING RADICAL OR PARTIAL CYSTECTOMY FOR BLADDER CANCER?
0022-5347/05/1735-1513/0 THE JOURNAL OF UROLOGY® Copyright © 2005 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 173, 1513–1517, May 2005 Printed in U.S.A.
DOI: 10.1097/01.ju.0000154352.54965.14
DOES BODY MASS INDEX AFFECT SURVIVAL OF PATIENTS UNDERGOING RADICAL OR PARTIAL CYSTECTOMY FOR BLADDER CANCER? JASON HAFRON, NANDITA MITRA, GUIDO DALBAGNI, BERNARD BOCHNER, HARRY HERR AND S. MACHELE DONAT* From the Departments of Urology and Epidemiology and Biostatistics (NM), Memorial Sloan-Kettering Cancer Center, New York, New York
ABSTRACT
Purpose: Obesity is estimated to account for up to 20% of all cancer deaths. We evaluated the effect of body mass index (BMI) on survival in patients undergoing radical or partial cystectomy for bladder cancer. Materials and Methods: Pathological and medical records on 300 consecutive patients undergoing radical or partial cystectomy for invasive bladder cancer between January 1990 and December 1993 were reviewed. The standard WHO definition of BMI was used, that is normal weight—less than 25 kg/m2, overweight—25 to 29.9 kg/m2 and obesity—30 kg/m2 or greater. Sufficient data were available on 288 of 300 patients (96%) with cystectomy (radical in 264 and partial in 24) for analysis. Results: The BMI distribution was normal weight in 34% of patients, overweight in 41% and obesity in 25%. Mean followup was 53.4 months (median 39, range 1 to 168). Median overall survival was 43.0 months (95% CI 37.1 to 58.4), while median disease specific survival was 82.5 months (95% CI 50.0 to 127.5). Multivariate analysis revealed that age greater than 65 years, pathological stage, smoking history and soft tissue margin status as significant factors impacting overall survival (p ⬍0.05). Pathological stage was organ confined (less than pT3a) in 51% of cases. BMI was not associated with disease specific survival as a continuous (p ⫽ 0.17) or categorical (p ⫽ 0.51) variable. Although it was insignificant, unadjusted analysis showed lower disease specific mortality in patients with a BMI of less than 25 mg/kg2 and organ confined disease (p ⫽ 0.08). Conclusions: There was no significant association between BMI and overall or disease specific survival, although there may be a trend toward better disease specific survival in normal weight (BMI less than 25kg/m2) patients with organ confined disease (p ⫽ 0.08). KEY WORDS: bladder, bladder neoplasms, body mass index, survival, cystectomy
Obesity is reaching epidemic levels in the United States, increasing by 61% between 1991 and 2000, and steadily becoming the greatest health risk.1 National Health and Nutrition Examination survey data from 1999 to 2000 reported that the age adjusted prevalence of overweight Americans was 64.5% and the prevalence of obesity was 30.5%.2 Body mass index (BMI) is an indirect measure of adiposity that is determined by weight in kg divided by height in m2. Although BMI does not distinguish between adipose tissue and lean body mass, there is good correlation between BMI and the percent of body fat, as measured by underwater weight and other techniques.3 The WHO defines BMI as normal weight—less than 25 kg/m2, overweight—25 to 29.9 kg/m2 and obesity—30 kg/m2 or greater.4 The relationship between obesity and diabetes, cardiovascular disease, digestive disorders and various musculoskeletal disorders are well recognized, although the association between obesity and cancer, specifically bladder cancer, has not been conclusively established. However, a recent prospective study of more than 900,000 adults estimated that the proportion of all deaths from cancer attributable to overweight and obesity in
American adults 50 years or older may be as high as 14% in men and 20% in women.5 Furthermore, the group estimated that 90,000 deaths from cancer could be prevented each year in the United States if men and women could maintain an ideal body weight. Along with increased mortality rates excess body mass has been associated with an increased risk of renal cancer and to a lesser degree an increased risk of prostate cancer.6 –9 The relationship between obesity and bladder cancer is less well defined with conflicting conclusions in the literature with regard to its impact on cancer related mortality6 –9 and little attention has been given to its impact on survival relative to given treatments. Therefore, we examined the impact of adiposity or BMI in patients with locally advanced bladder cancer undergoing standard radical cystectomy as treatment. MATERIALS AND METHODS
After obtaining institutional review board approval the records of 300 consecutive patients undergoing radical or partial cystectomy for invasive bladder cancer at Memorial Hospital between January 1990 and December 1993 were reviewed. Sufficient data were available on 288 of 300 cystectomy cases (96%) (radical in 264 and partial in 24) for analysis. Case records were retrospectively reviewed for demographic, comorbid, pathological, surgical and clinical information on each case. Patients were followed after cystectomy until death. Preoperative clinical stage and pathological stage at surgery were as-
Submitted for publication August 9, 2004. Study received institutional review board approval. * Correspondence and requests for reprints: Department of Urology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, New York 10021 (telephone: 646-422-4398; FAX: 212-9880768; e-mail: [email protected]). 1513
1514
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
signed according to the 1997 American Joint Committee on Cancer-UICC TNM system.10 The Charlson index for age, a validated comorbidity index to classify comorbid condition and age impact on mortality, was assigned based on available clinical data.11, 12 BMI was calculated on the day of surgery as patient weight in kg divided by patient height in m2. The standard WHO definitions of BMI were used (normal—less than 25 kg/m2, overweight—25 to 29.9 kg/m2 and obesity—30 kg/m2 or greater). All survival data were calculated from the date of cystectomy. The primary study end point was overall survival in patients who underwent radical or partial cystectomy for bladder cancer. Kaplan-Meier estimates of overall survival time stratified by various prognostic factor categories were calculated and compared by the log rank test statistic. A multivariate Cox regression model was used to assess the association between survival and BMI, adjusting for important clinical and patient covariates.13 The probability of disease specific survival was also calculated using competing risk analysis accounting for death due to other causes as competing risks.14, 15 Analyses were done using SAS, version 9.0 (SAS Institute, Carey, North Carolina) and R, version 1.8.1 (Stanford University, Stanford, California). All p values are 2-sided with significance considered at p ⱕ0.05. RESULTS
Table 1 lists the demographic and clinical/pathological characteristics of the study population segregated by preoperative BMI groupings. Consistent with national trends in bladder cancer, there was a 3:1 male-to-female ratio with 226 of 288 male patients (79%) and 62 female patients (21%), and a median age of 67 years at cystectomy (mean 65.4, range 36 to 88). Mean BMI at surgery ⫾ SE in the entire patient cohort was 26.98 ⫾ 0.26 kg/m2 (range 18 to 43.6). The overall prevalence of overweight or obese patients in the cohort was 66%, which is consistent with reported national trends.2 There was a relatively equal distribution of patients in the 3 WHO BMI categories with 98 of 288 (34%) categorized as normal, 119 of 288 (41%) categorized as overweight and 71 of 288 (25%) categorized as obese. Median BMI remained constant during the 4 study years. Of 288 patients 202 (70%) were actively smoking or had quit smoking prior to surgery. The median
Charlson age index in all 3 groups was calculated to be 3, while the median American Society of Anesthesiologists (ASA) score was 2 in the normal and overweight groups, and 3 in the obese group. A small portion of the cohort underwent neoadjuvant chemotherapy and/or external beam radiation therapy (19% and 4.5%, respectively). Adjuvant chemotherapy was given in 12 of 288 patients (4.2%) at the discretion of the medical oncologist. Median overall survival in the cohort was 43 months (range 1 to 168, 95% CI 37.1 to 58.4). Median disease specific survival was 82.5 months (95% CI 50.0 to 127.5). Median followup in the cohort was 39 months (mean 53.4, range 1 to 168). Univariate analyses of clinical and demographic characteristics revealed that age at cystectomy, pathological stage, lymph node status, soft tissue margin status and smoking history were significant predictors of overall survival (table 2). Age greater than 65 years at cystectomy, smoking history, pathological stage and soft tissue margin status remained significant factors for predicting survival on multivariate analysis (p ⬍0.05, table 3). BMI was not a significant factor for predicting overall survival following radical or partial cystectomy on univariate or multivariate analysis (p ⫽ 0.33 and 0.16, respectively, tables 2 and 3). Additionally, Kaplan-Meier survival analyses did not reveal a significant association between BMI category and overall survival (p ⫽ 0.33, fig. 1) or disease specific survival (p ⫽ 0.55, fig. 2). Furthermore, when analyzed as a continuous variable, no significant correlation was demonstrated between BMI and disease specific (p ⫽ 0.17) or overall survival (p ⫽ 0.78). Although it was insignificant, on unadjusted analysis there may have been a trend toward better disease specific survival in patients with BMI less than 25 mg/kg2 and organ confined disease (p ⫽ 0.08). Of 288 patients in the series 146 (51%) had organ confined (less than T3a) disease at cystectomy with no significant correlation seen between BMI category and overall survival in patients with organ confined disease (p ⫽ 0.44, fig. 3). Due to the documented inverse relationship between smoking and BMI16, 17 subcohort analysis evaluating the impact of BMI on survival in nonsmokers was performed. Analysis again showed no significant relationship between BMI and overall survival (p ⫽ 0.51, fig. 4).
TABLE 1. Demographics and clinical/pathological features of cohort categorized by WHO BMI category BMI Cohort No. pts No. partial (%) No. radical cystectomy (%) Age at cystectomy: Median Mean ⫾ SD No. older than 65 (%) No. sex (%): Male Female Charlson age index ASA level No. ever smoked (%) No. neoadjuvant chemotherapy (%) No. neoadjuvant radiation (%) No. clinically under staged (%) No. pathological stage (%): Organ confined Nonorgan confined No. node pos (%) No. soft tissue margin pos (%) No. pelvic recurrence (%) Median mos to local recurrence No. alive at last followup (%) Median mos followup
288 24 264
(8) (92)
67 65.4 ⫾ 9.7 153 (53)
Less Than 25 98 7 (7) 91 (93) 67 66.1 ⫾ 10.3 51 (53) (68) (32)
25–29.9 119 9 110
(8) (92)
67 65.6 ⫾ 9.4 69 (58)
226 (79) 62 (21) 3 2 202 (70) 56 (19) 13 (4.5) 117 (41)
67 31 3 2 69 21 5 40
(70) (21) (5) (40)
103 (87) 16 (13) 3 2 85 (71) 20 (17) 4 (3.4) 47 (40)
146 (51) 142 (49) 45 (16) 25 (9) 18 (6) 16.5 85 (30) 39
47 (48) 51 (52) 24 (25) 11 (11) 7 (7) 15.7 26 (27) 32
66 (56) 53 (44) 18 (15) 7 (6) 5 (4) 9.1 33 (28) 40
Greater Than 29.9 71 8 (11) 63 (89) 64 64.2 ⫾ 9.5 33 (47) 56 15 3 3 48 15 4 30
(79) (21) (68) (21) (6) (42)
33 (47) 38 (53) 3 (4) 7 (10) 6 (9) 17.3 26 (37) 42
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
1515
TABLE 2. Overall survival by demographic and clinical categories Variable Category
No.
Median Survival (95% CI)
Sex: M 226 46.6 F 62 31.4 Age: 65 or Younger 132 70.4 Older than 65 156 38.9 Pathological stage: Organ confined 146 88.5 Nonorgan confined 142 39.6 Node status: Neg 243 50.5 Pos 45 27.0 Soft tissue margin status: Neg 263 47.7 Pos 25 13.9 Urothelial margin: Neg 268 43.0 Pos 20 60.2 Smoking history:* Never 64 31.9 Ever 202 52.2 2 BMI (kg/m ): Less than 25 98 38.7 25–29.9 119 42.8 Greater than 29.9 71 60.1 * Not available in 22 of 288 patients (8%).
p Value 0.88
(38.9, 59.7) (20.9, 65.4) (40.3, 92.1) (24.7, 47.1) (63.8, 111.6) (21.1, 47.1) (40.5, 65.4) (15.4, 35.5)
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
(40.3, 60.2) (6.3, 20.9)
FIG. 2. Kaplan-Meier analysis of disease specific survival based on BMI category.
0.90 (38.0, 58.4) (10.0, 147.8) 0.04 (18.8, 40.5) (42.8, 70.6) 0.33 (24.5, 70.6) (33.5, 58.8) (38.1, 88.5)
TABLE 3. Multivariate analysis of factors affecting overall survival Clinical Variable
HR (95% CI)
Sex 0.89 (0.61, 1.28) Age greater than 65 at cystectomy 2.09 (1.54, 2.83) Nonorgan confined disease 2.58 (1.87, 3.56) Pos node status 1.28 (0.86, 1.92) Pos soft tissue margin status 1.84 (1.15, 2.95) Pos urothelial margin 0.89 (0.48, 1.65) Smoking* 0.71 (0.51, 0.99) BMI (obese vs other) 0.87 (0.71, 1.06) * Not available in 22 of 288 patients (8%).
p Value 0.51 ⬍0.001 ⬍0.001 0.23 0.01 0.72 0.04 0.16
FIG. 1. Kaplan-Meier analysis of overall survival based on BMI category.
DISCUSSION
There are mounting data indicating that above normal BMI (overweight and obesity) not only contributes to the risk of comorbid conditions such as diabetes and cardiovascular disease, but also may portend an increased risk of death from cancer, especially in those with a BMI of greater than 40 kg/ m2.1, 5, 9 Four large contemporary epidemiological studies from
FIG. 3. Kaplan-Meier analysis of overall survival in patients with organ confined bladder cancer based on WHO BMI category.
FIG. 4. Kaplan-Meier analysis of overall survival in nonsmokers based on WHO BMI category.
Canada,8 Denmark,6 Sweden7 and the United States (veterans)9 assessed the relationship between excess body mass and the risk of cancer, concluding that excess body mass accounted for a 7% to 33% increased incidence in all forms of cancer. Evaluation of site specific cancers has shown a consistent increased risk for renal cancer associated with increasing BMI and a lesser, inconsistent increased risk for prostate and bladder cancer.6⫺9 However, a recent study of Calle et al failed to
1516
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
show a significant association between BMI and mortality from bladder cancer in men (p ⫽ 0.36) or women (p ⫽ 0.21) in the United States.5 Our study of patients with bladder cancer undergoing radical cystectomy also failed to show a relationship between BMI and survival in men or women, even in patients with a BMI of greater than 29.9 (p ⫽ 0.11). Consistent with prior studies of bladder cancer the extent of disease at cystectomy, as reflected by pathological stage, nodal status and soft tissue margin status, significantly impacted survival in our cohort of patients. One could argue that pathological stage, as reflected by organ confined status, may overshadow any adverse impact that BMI might have on survival in this cohort of patients. However, subcohort analysis of patients with organ confined disease separated by WHO BMI classifications did not demonstrate a significant difference in overall survival among the groups (p ⫽ 0.44, fig. 3). Disease specific survival was also unaffected by BMI category, although unadjusted analysis stratifying the population by organ confined status showed an insignificant trend toward better survival in normal weight patients (p ⫽ 0.08). Confounding risk factors for mortality, such as cancer, smoking and preexisting medical conditions, may distort the link between BMI and cancer survival. Smoking has definitively been shown to be inversely associated with body weight16, 17 and, therefore, it must be considered when examining the affect of BMI on cancer survival, especially in cancers with smoking related etiologies. Epidemiological studies have shown a causal association between smoking and an increased risk of bladder cancer.18 More importantly up to 50% of patients continue to smoke following the diagnosis with several studies now showing a worse disease related outcome compared with that in patients who quit smoking.19 The confounding effects of cigarette smoking on survival are most apparent in reports examining lung cancer, of which most show an inverse relationship between BMI and lung cancer mortality.20 Accordingly Calle et al found that in smoking related cancers the prospective effects of BMI on the risk of death could not be separated from those related to smoking and they concluded that estimates of relative risk of death based on those in the population who never smoked may offer the most valid effect of BMI on cancer death.5 However, when patients were stratified based on smoking history, in the nonsmoking group the inverse association disappeared. In our cohort 202 of the 288 patients (70%) had smoked at 1 time, while 62 of the 202 (31%) were still smoking at cystectomy and 44 of the 288 (15%) died of other smoking related causes. A history of past or current smoking significantly affected survival outcomes on univariate and multivariate analyses (p ⫽ 0.04 and 0.04, tables 2 and 3). However, unadjusted subset analysis in patients who had never smoked failed to show a significant effect of BMI on overall survival in that population (p ⫽ 0.51, fig. 4). The degree of comorbid illness has also been shown to be a strong predictor of outcome and, therefore, it is often used by surgeons to determine whether a patient is at acceptable risk for surgery.11, 12 This is especially relevant to the treatment of bladder cancer due to the advanced age and age associated comorbidities of most patients undergoing radical cystectomy, as demonstrated in this series, in which median age at cystectomy was 67 years with 153 of 288 patients (53%) older than 65 years and 51 of 288 (17%) 75 years or older. As a result, there have been multiple preoperative risk assessments tools developed to help determine the patient risk of mortality or significant morbidity with surgery.11, 12 The definition of a surgical candidate is not measurable but it includes evaluation and assessment of comorbidities or overall health. In our study the Charlson age index, a validated measure of comorbidity and age, along with the ASA score was used to determine operative risk. However, neither was found to be a predictor of survival in and of itself, most likely because the majority of patients in the study had a Charlson
index of 3 and were distributed evenly among the different categories of BMI, and the high surgical volume may have resulted in lower complication rates, which is a reflection of the referral pattern to this tertiary cancer center and a limitation of the study. CONCLUSIONS
We were unable to establish a significant association between increasing BMI and overall or disease specific survival in comorbid patients undergoing radial or partial cystectomy for invasive bladder cancer, although there may be a trend toward better disease specific survival in normal weight (BMI less than 25 kg/m2) patients with organ confined disease (p ⫽ 0.08). Pathological stage, soft tissue margin status, smoking history and age greater than 65 years at cystectomy significantly impacted survival with age greater than 65 years and nonorgan confined disease remaining the 2 most significant factors for survival on multivariate analysis (p ⬍0.001). The limitations of the study are its retrospective nature, small population size, high incidence of smoking, high rate of comorbidities and patient selection bias inherent in single institutional studies. We do not know the denominator of patients who did not undergo surgical resection or the effect that BMI may have had on their selection for therapy but the majority of the study population had a preoperative Charlson index of 3 and 66% were overweight or obese using WHO standards at surgery, indicating if anything that the cohort was weighted toward surgical patients at higher risk with multiple comorbidities. A separate analysis of the subset of nonsmokers in our cohort still failed to show a significant relationship between BMI category and survival, similar to the findings of Calle et al.5 While excess body mass may increase the risk of bladder cancer, it does not appear to have a significant role in the outcome of patients undergoing radical cystectomy for invasive disease and it should not be a limiting factor in surgical decision making. REFERENCES
1. Mokdad, A. H., Bowman, B. A., Ford, E. S., Vinicor, F., Marks, J. S. and Koplan, J. P.: The continuing epidemics of obesity and diabetes in the United States. JAMA, 286: 1195, 2001 2. Flegal, K. M., Carrol, M. D., Ogden, C. L. and Johnson, C. L.: Prevalence and trends in obesity among US adults, 1999 – 2000. JAMA, 288: 1723, 2002 3. Deurenberg, P., Westsrate, J. A. and Seidell, J. C.: Body mass index as a measure of body fatness: age- and sex-specific prediction formulas. Br J Nutr, 65: 105, 1991 4. Physical status: the use of interpretation of anthropometry. Report of a WHO expert committee. World Health Organ Tech Rep Ser, 854: 1, 1995 5. Calle, E. E., Rodriguez, C., Walker-Thurmond, K. and Thun, M. J.: Overweight, obesity and mortality from cancer in a prospectively studied cohort of US adults. N Engl J Med, 348: 1625, 2003 6. Moller, H., Mellemgaard, A., Lindvig K. and Olsen, J. H.: Obesity and cancer risk: a Danish record-linkage study. Eur J Cancer, 30A: 344, 1994 7. Wolk, A., Gridley, G., Svensson, M., Nyren, O., McLaughlin, J. K., Fraumeni, J. F. et al: A prospective study of obesity and cancer risk (Sweden). Cancer Causes Control, 12: 13, 2001 8. Pan, S. Y., Johnson, K. C., Ugnat, A. M., Wen, S. W., Mao, Y. and Canadian Cancer Registries Epidemiology Research Group: Association of obesity and cancer risk in Canada. Am J Epidemiol, 159: 259, 2004 9. Samanic, C., Gridley, G., Chow, W. H., Lubin, J., Hoover, R. N. and Fraumeni, J. F., Jr.: Obesity and cancer risk among white and black United States veterans. Cancer Causes Control, 15: 35, 2004 10. Greene, F. L., Page, D. L. Fleming, I. D., Fritz, A., Balch, C. M., Haller, D. G. et al: AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag, pp. 335–340, 2002 11. Charlson, M. E., Pompei P., Ales, K. and MacKenzie, C. R.: A
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
12. 13. 14. 15. 16.
new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 40: 373, 1987 Extermann, M.: Measuring comorbidity in older cancer patients. Eur J Cancer, 36: 453, 2000 Collet, D.: Modelling Survival Data in Medical Research. London: Chapman and Hall, 1994 Gray, R. J.: A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat, 16: 1141, 1988 Fine, J. P. and Gray, R. J.: A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc, 94: 496, 1999 Molarius, A., Seidell, J. C., Kuulasmaa, K., Dobson, A. J. and Sans, S.: Smoking and relative body weight: an international perspective from the WHO MONICA Project. J Epidemiol Community Health, 51: 252, 1997
1517
17. Henley, S. J., Flanders, W. D., Manatunga, A. and Thun, M. J.: Leanness and lung cancer risk: fact or artifact? Epidemiology, 13: 268, 2002 18. Burch, J. D., Rohan, T. E., Howe, G. R., Risch, H. A., Hill, G. B., Steele, R. et al: Risk of bladder cancer by source and type of tobacco exposure: a case-control study. Int J Cancer, 44: 622, 1989 19. Fleshner, N., Garland, J., Moadel, A., Herr, H., Ostroff, J., Trambert, R. et al: Influence of smoking status on the diseaserelated outcomes of patients with tobacco-associated superficial transitional cell carcinoma of the bladder. Cancer, 86: 2337, 1999 20. Miller, D. C., Taub, D. A., Dunn, R. L., Montie, J. E. and Wei, J. T.: The impact of co-morbid disease on cancer control and survival following radical cystectomy. J Urol, 169: 105, 2003
Vol. 173, 1513–1517, May 2005 Printed in U.S.A.
DOI: 10.1097/01.ju.0000154352.54965.14
DOES BODY MASS INDEX AFFECT SURVIVAL OF PATIENTS UNDERGOING RADICAL OR PARTIAL CYSTECTOMY FOR BLADDER CANCER? JASON HAFRON, NANDITA MITRA, GUIDO DALBAGNI, BERNARD BOCHNER, HARRY HERR AND S. MACHELE DONAT* From the Departments of Urology and Epidemiology and Biostatistics (NM), Memorial Sloan-Kettering Cancer Center, New York, New York
ABSTRACT
Purpose: Obesity is estimated to account for up to 20% of all cancer deaths. We evaluated the effect of body mass index (BMI) on survival in patients undergoing radical or partial cystectomy for bladder cancer. Materials and Methods: Pathological and medical records on 300 consecutive patients undergoing radical or partial cystectomy for invasive bladder cancer between January 1990 and December 1993 were reviewed. The standard WHO definition of BMI was used, that is normal weight—less than 25 kg/m2, overweight—25 to 29.9 kg/m2 and obesity—30 kg/m2 or greater. Sufficient data were available on 288 of 300 patients (96%) with cystectomy (radical in 264 and partial in 24) for analysis. Results: The BMI distribution was normal weight in 34% of patients, overweight in 41% and obesity in 25%. Mean followup was 53.4 months (median 39, range 1 to 168). Median overall survival was 43.0 months (95% CI 37.1 to 58.4), while median disease specific survival was 82.5 months (95% CI 50.0 to 127.5). Multivariate analysis revealed that age greater than 65 years, pathological stage, smoking history and soft tissue margin status as significant factors impacting overall survival (p ⬍0.05). Pathological stage was organ confined (less than pT3a) in 51% of cases. BMI was not associated with disease specific survival as a continuous (p ⫽ 0.17) or categorical (p ⫽ 0.51) variable. Although it was insignificant, unadjusted analysis showed lower disease specific mortality in patients with a BMI of less than 25 mg/kg2 and organ confined disease (p ⫽ 0.08). Conclusions: There was no significant association between BMI and overall or disease specific survival, although there may be a trend toward better disease specific survival in normal weight (BMI less than 25kg/m2) patients with organ confined disease (p ⫽ 0.08). KEY WORDS: bladder, bladder neoplasms, body mass index, survival, cystectomy
Obesity is reaching epidemic levels in the United States, increasing by 61% between 1991 and 2000, and steadily becoming the greatest health risk.1 National Health and Nutrition Examination survey data from 1999 to 2000 reported that the age adjusted prevalence of overweight Americans was 64.5% and the prevalence of obesity was 30.5%.2 Body mass index (BMI) is an indirect measure of adiposity that is determined by weight in kg divided by height in m2. Although BMI does not distinguish between adipose tissue and lean body mass, there is good correlation between BMI and the percent of body fat, as measured by underwater weight and other techniques.3 The WHO defines BMI as normal weight—less than 25 kg/m2, overweight—25 to 29.9 kg/m2 and obesity—30 kg/m2 or greater.4 The relationship between obesity and diabetes, cardiovascular disease, digestive disorders and various musculoskeletal disorders are well recognized, although the association between obesity and cancer, specifically bladder cancer, has not been conclusively established. However, a recent prospective study of more than 900,000 adults estimated that the proportion of all deaths from cancer attributable to overweight and obesity in
American adults 50 years or older may be as high as 14% in men and 20% in women.5 Furthermore, the group estimated that 90,000 deaths from cancer could be prevented each year in the United States if men and women could maintain an ideal body weight. Along with increased mortality rates excess body mass has been associated with an increased risk of renal cancer and to a lesser degree an increased risk of prostate cancer.6 –9 The relationship between obesity and bladder cancer is less well defined with conflicting conclusions in the literature with regard to its impact on cancer related mortality6 –9 and little attention has been given to its impact on survival relative to given treatments. Therefore, we examined the impact of adiposity or BMI in patients with locally advanced bladder cancer undergoing standard radical cystectomy as treatment. MATERIALS AND METHODS
After obtaining institutional review board approval the records of 300 consecutive patients undergoing radical or partial cystectomy for invasive bladder cancer at Memorial Hospital between January 1990 and December 1993 were reviewed. Sufficient data were available on 288 of 300 cystectomy cases (96%) (radical in 264 and partial in 24) for analysis. Case records were retrospectively reviewed for demographic, comorbid, pathological, surgical and clinical information on each case. Patients were followed after cystectomy until death. Preoperative clinical stage and pathological stage at surgery were as-
Submitted for publication August 9, 2004. Study received institutional review board approval. * Correspondence and requests for reprints: Department of Urology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, New York 10021 (telephone: 646-422-4398; FAX: 212-9880768; e-mail: [email protected]). 1513
1514
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
signed according to the 1997 American Joint Committee on Cancer-UICC TNM system.10 The Charlson index for age, a validated comorbidity index to classify comorbid condition and age impact on mortality, was assigned based on available clinical data.11, 12 BMI was calculated on the day of surgery as patient weight in kg divided by patient height in m2. The standard WHO definitions of BMI were used (normal—less than 25 kg/m2, overweight—25 to 29.9 kg/m2 and obesity—30 kg/m2 or greater). All survival data were calculated from the date of cystectomy. The primary study end point was overall survival in patients who underwent radical or partial cystectomy for bladder cancer. Kaplan-Meier estimates of overall survival time stratified by various prognostic factor categories were calculated and compared by the log rank test statistic. A multivariate Cox regression model was used to assess the association between survival and BMI, adjusting for important clinical and patient covariates.13 The probability of disease specific survival was also calculated using competing risk analysis accounting for death due to other causes as competing risks.14, 15 Analyses were done using SAS, version 9.0 (SAS Institute, Carey, North Carolina) and R, version 1.8.1 (Stanford University, Stanford, California). All p values are 2-sided with significance considered at p ⱕ0.05. RESULTS
Table 1 lists the demographic and clinical/pathological characteristics of the study population segregated by preoperative BMI groupings. Consistent with national trends in bladder cancer, there was a 3:1 male-to-female ratio with 226 of 288 male patients (79%) and 62 female patients (21%), and a median age of 67 years at cystectomy (mean 65.4, range 36 to 88). Mean BMI at surgery ⫾ SE in the entire patient cohort was 26.98 ⫾ 0.26 kg/m2 (range 18 to 43.6). The overall prevalence of overweight or obese patients in the cohort was 66%, which is consistent with reported national trends.2 There was a relatively equal distribution of patients in the 3 WHO BMI categories with 98 of 288 (34%) categorized as normal, 119 of 288 (41%) categorized as overweight and 71 of 288 (25%) categorized as obese. Median BMI remained constant during the 4 study years. Of 288 patients 202 (70%) were actively smoking or had quit smoking prior to surgery. The median
Charlson age index in all 3 groups was calculated to be 3, while the median American Society of Anesthesiologists (ASA) score was 2 in the normal and overweight groups, and 3 in the obese group. A small portion of the cohort underwent neoadjuvant chemotherapy and/or external beam radiation therapy (19% and 4.5%, respectively). Adjuvant chemotherapy was given in 12 of 288 patients (4.2%) at the discretion of the medical oncologist. Median overall survival in the cohort was 43 months (range 1 to 168, 95% CI 37.1 to 58.4). Median disease specific survival was 82.5 months (95% CI 50.0 to 127.5). Median followup in the cohort was 39 months (mean 53.4, range 1 to 168). Univariate analyses of clinical and demographic characteristics revealed that age at cystectomy, pathological stage, lymph node status, soft tissue margin status and smoking history were significant predictors of overall survival (table 2). Age greater than 65 years at cystectomy, smoking history, pathological stage and soft tissue margin status remained significant factors for predicting survival on multivariate analysis (p ⬍0.05, table 3). BMI was not a significant factor for predicting overall survival following radical or partial cystectomy on univariate or multivariate analysis (p ⫽ 0.33 and 0.16, respectively, tables 2 and 3). Additionally, Kaplan-Meier survival analyses did not reveal a significant association between BMI category and overall survival (p ⫽ 0.33, fig. 1) or disease specific survival (p ⫽ 0.55, fig. 2). Furthermore, when analyzed as a continuous variable, no significant correlation was demonstrated between BMI and disease specific (p ⫽ 0.17) or overall survival (p ⫽ 0.78). Although it was insignificant, on unadjusted analysis there may have been a trend toward better disease specific survival in patients with BMI less than 25 mg/kg2 and organ confined disease (p ⫽ 0.08). Of 288 patients in the series 146 (51%) had organ confined (less than T3a) disease at cystectomy with no significant correlation seen between BMI category and overall survival in patients with organ confined disease (p ⫽ 0.44, fig. 3). Due to the documented inverse relationship between smoking and BMI16, 17 subcohort analysis evaluating the impact of BMI on survival in nonsmokers was performed. Analysis again showed no significant relationship between BMI and overall survival (p ⫽ 0.51, fig. 4).
TABLE 1. Demographics and clinical/pathological features of cohort categorized by WHO BMI category BMI Cohort No. pts No. partial (%) No. radical cystectomy (%) Age at cystectomy: Median Mean ⫾ SD No. older than 65 (%) No. sex (%): Male Female Charlson age index ASA level No. ever smoked (%) No. neoadjuvant chemotherapy (%) No. neoadjuvant radiation (%) No. clinically under staged (%) No. pathological stage (%): Organ confined Nonorgan confined No. node pos (%) No. soft tissue margin pos (%) No. pelvic recurrence (%) Median mos to local recurrence No. alive at last followup (%) Median mos followup
288 24 264
(8) (92)
67 65.4 ⫾ 9.7 153 (53)
Less Than 25 98 7 (7) 91 (93) 67 66.1 ⫾ 10.3 51 (53) (68) (32)
25–29.9 119 9 110
(8) (92)
67 65.6 ⫾ 9.4 69 (58)
226 (79) 62 (21) 3 2 202 (70) 56 (19) 13 (4.5) 117 (41)
67 31 3 2 69 21 5 40
(70) (21) (5) (40)
103 (87) 16 (13) 3 2 85 (71) 20 (17) 4 (3.4) 47 (40)
146 (51) 142 (49) 45 (16) 25 (9) 18 (6) 16.5 85 (30) 39
47 (48) 51 (52) 24 (25) 11 (11) 7 (7) 15.7 26 (27) 32
66 (56) 53 (44) 18 (15) 7 (6) 5 (4) 9.1 33 (28) 40
Greater Than 29.9 71 8 (11) 63 (89) 64 64.2 ⫾ 9.5 33 (47) 56 15 3 3 48 15 4 30
(79) (21) (68) (21) (6) (42)
33 (47) 38 (53) 3 (4) 7 (10) 6 (9) 17.3 26 (37) 42
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
1515
TABLE 2. Overall survival by demographic and clinical categories Variable Category
No.
Median Survival (95% CI)
Sex: M 226 46.6 F 62 31.4 Age: 65 or Younger 132 70.4 Older than 65 156 38.9 Pathological stage: Organ confined 146 88.5 Nonorgan confined 142 39.6 Node status: Neg 243 50.5 Pos 45 27.0 Soft tissue margin status: Neg 263 47.7 Pos 25 13.9 Urothelial margin: Neg 268 43.0 Pos 20 60.2 Smoking history:* Never 64 31.9 Ever 202 52.2 2 BMI (kg/m ): Less than 25 98 38.7 25–29.9 119 42.8 Greater than 29.9 71 60.1 * Not available in 22 of 288 patients (8%).
p Value 0.88
(38.9, 59.7) (20.9, 65.4) (40.3, 92.1) (24.7, 47.1) (63.8, 111.6) (21.1, 47.1) (40.5, 65.4) (15.4, 35.5)
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
(40.3, 60.2) (6.3, 20.9)
FIG. 2. Kaplan-Meier analysis of disease specific survival based on BMI category.
0.90 (38.0, 58.4) (10.0, 147.8) 0.04 (18.8, 40.5) (42.8, 70.6) 0.33 (24.5, 70.6) (33.5, 58.8) (38.1, 88.5)
TABLE 3. Multivariate analysis of factors affecting overall survival Clinical Variable
HR (95% CI)
Sex 0.89 (0.61, 1.28) Age greater than 65 at cystectomy 2.09 (1.54, 2.83) Nonorgan confined disease 2.58 (1.87, 3.56) Pos node status 1.28 (0.86, 1.92) Pos soft tissue margin status 1.84 (1.15, 2.95) Pos urothelial margin 0.89 (0.48, 1.65) Smoking* 0.71 (0.51, 0.99) BMI (obese vs other) 0.87 (0.71, 1.06) * Not available in 22 of 288 patients (8%).
p Value 0.51 ⬍0.001 ⬍0.001 0.23 0.01 0.72 0.04 0.16
FIG. 1. Kaplan-Meier analysis of overall survival based on BMI category.
DISCUSSION
There are mounting data indicating that above normal BMI (overweight and obesity) not only contributes to the risk of comorbid conditions such as diabetes and cardiovascular disease, but also may portend an increased risk of death from cancer, especially in those with a BMI of greater than 40 kg/ m2.1, 5, 9 Four large contemporary epidemiological studies from
FIG. 3. Kaplan-Meier analysis of overall survival in patients with organ confined bladder cancer based on WHO BMI category.
FIG. 4. Kaplan-Meier analysis of overall survival in nonsmokers based on WHO BMI category.
Canada,8 Denmark,6 Sweden7 and the United States (veterans)9 assessed the relationship between excess body mass and the risk of cancer, concluding that excess body mass accounted for a 7% to 33% increased incidence in all forms of cancer. Evaluation of site specific cancers has shown a consistent increased risk for renal cancer associated with increasing BMI and a lesser, inconsistent increased risk for prostate and bladder cancer.6⫺9 However, a recent study of Calle et al failed to
1516
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
show a significant association between BMI and mortality from bladder cancer in men (p ⫽ 0.36) or women (p ⫽ 0.21) in the United States.5 Our study of patients with bladder cancer undergoing radical cystectomy also failed to show a relationship between BMI and survival in men or women, even in patients with a BMI of greater than 29.9 (p ⫽ 0.11). Consistent with prior studies of bladder cancer the extent of disease at cystectomy, as reflected by pathological stage, nodal status and soft tissue margin status, significantly impacted survival in our cohort of patients. One could argue that pathological stage, as reflected by organ confined status, may overshadow any adverse impact that BMI might have on survival in this cohort of patients. However, subcohort analysis of patients with organ confined disease separated by WHO BMI classifications did not demonstrate a significant difference in overall survival among the groups (p ⫽ 0.44, fig. 3). Disease specific survival was also unaffected by BMI category, although unadjusted analysis stratifying the population by organ confined status showed an insignificant trend toward better survival in normal weight patients (p ⫽ 0.08). Confounding risk factors for mortality, such as cancer, smoking and preexisting medical conditions, may distort the link between BMI and cancer survival. Smoking has definitively been shown to be inversely associated with body weight16, 17 and, therefore, it must be considered when examining the affect of BMI on cancer survival, especially in cancers with smoking related etiologies. Epidemiological studies have shown a causal association between smoking and an increased risk of bladder cancer.18 More importantly up to 50% of patients continue to smoke following the diagnosis with several studies now showing a worse disease related outcome compared with that in patients who quit smoking.19 The confounding effects of cigarette smoking on survival are most apparent in reports examining lung cancer, of which most show an inverse relationship between BMI and lung cancer mortality.20 Accordingly Calle et al found that in smoking related cancers the prospective effects of BMI on the risk of death could not be separated from those related to smoking and they concluded that estimates of relative risk of death based on those in the population who never smoked may offer the most valid effect of BMI on cancer death.5 However, when patients were stratified based on smoking history, in the nonsmoking group the inverse association disappeared. In our cohort 202 of the 288 patients (70%) had smoked at 1 time, while 62 of the 202 (31%) were still smoking at cystectomy and 44 of the 288 (15%) died of other smoking related causes. A history of past or current smoking significantly affected survival outcomes on univariate and multivariate analyses (p ⫽ 0.04 and 0.04, tables 2 and 3). However, unadjusted subset analysis in patients who had never smoked failed to show a significant effect of BMI on overall survival in that population (p ⫽ 0.51, fig. 4). The degree of comorbid illness has also been shown to be a strong predictor of outcome and, therefore, it is often used by surgeons to determine whether a patient is at acceptable risk for surgery.11, 12 This is especially relevant to the treatment of bladder cancer due to the advanced age and age associated comorbidities of most patients undergoing radical cystectomy, as demonstrated in this series, in which median age at cystectomy was 67 years with 153 of 288 patients (53%) older than 65 years and 51 of 288 (17%) 75 years or older. As a result, there have been multiple preoperative risk assessments tools developed to help determine the patient risk of mortality or significant morbidity with surgery.11, 12 The definition of a surgical candidate is not measurable but it includes evaluation and assessment of comorbidities or overall health. In our study the Charlson age index, a validated measure of comorbidity and age, along with the ASA score was used to determine operative risk. However, neither was found to be a predictor of survival in and of itself, most likely because the majority of patients in the study had a Charlson
index of 3 and were distributed evenly among the different categories of BMI, and the high surgical volume may have resulted in lower complication rates, which is a reflection of the referral pattern to this tertiary cancer center and a limitation of the study. CONCLUSIONS
We were unable to establish a significant association between increasing BMI and overall or disease specific survival in comorbid patients undergoing radial or partial cystectomy for invasive bladder cancer, although there may be a trend toward better disease specific survival in normal weight (BMI less than 25 kg/m2) patients with organ confined disease (p ⫽ 0.08). Pathological stage, soft tissue margin status, smoking history and age greater than 65 years at cystectomy significantly impacted survival with age greater than 65 years and nonorgan confined disease remaining the 2 most significant factors for survival on multivariate analysis (p ⬍0.001). The limitations of the study are its retrospective nature, small population size, high incidence of smoking, high rate of comorbidities and patient selection bias inherent in single institutional studies. We do not know the denominator of patients who did not undergo surgical resection or the effect that BMI may have had on their selection for therapy but the majority of the study population had a preoperative Charlson index of 3 and 66% were overweight or obese using WHO standards at surgery, indicating if anything that the cohort was weighted toward surgical patients at higher risk with multiple comorbidities. A separate analysis of the subset of nonsmokers in our cohort still failed to show a significant relationship between BMI category and survival, similar to the findings of Calle et al.5 While excess body mass may increase the risk of bladder cancer, it does not appear to have a significant role in the outcome of patients undergoing radical cystectomy for invasive disease and it should not be a limiting factor in surgical decision making. REFERENCES
1. Mokdad, A. H., Bowman, B. A., Ford, E. S., Vinicor, F., Marks, J. S. and Koplan, J. P.: The continuing epidemics of obesity and diabetes in the United States. JAMA, 286: 1195, 2001 2. Flegal, K. M., Carrol, M. D., Ogden, C. L. and Johnson, C. L.: Prevalence and trends in obesity among US adults, 1999 – 2000. JAMA, 288: 1723, 2002 3. Deurenberg, P., Westsrate, J. A. and Seidell, J. C.: Body mass index as a measure of body fatness: age- and sex-specific prediction formulas. Br J Nutr, 65: 105, 1991 4. Physical status: the use of interpretation of anthropometry. Report of a WHO expert committee. World Health Organ Tech Rep Ser, 854: 1, 1995 5. Calle, E. E., Rodriguez, C., Walker-Thurmond, K. and Thun, M. J.: Overweight, obesity and mortality from cancer in a prospectively studied cohort of US adults. N Engl J Med, 348: 1625, 2003 6. Moller, H., Mellemgaard, A., Lindvig K. and Olsen, J. H.: Obesity and cancer risk: a Danish record-linkage study. Eur J Cancer, 30A: 344, 1994 7. Wolk, A., Gridley, G., Svensson, M., Nyren, O., McLaughlin, J. K., Fraumeni, J. F. et al: A prospective study of obesity and cancer risk (Sweden). Cancer Causes Control, 12: 13, 2001 8. Pan, S. Y., Johnson, K. C., Ugnat, A. M., Wen, S. W., Mao, Y. and Canadian Cancer Registries Epidemiology Research Group: Association of obesity and cancer risk in Canada. Am J Epidemiol, 159: 259, 2004 9. Samanic, C., Gridley, G., Chow, W. H., Lubin, J., Hoover, R. N. and Fraumeni, J. F., Jr.: Obesity and cancer risk among white and black United States veterans. Cancer Causes Control, 15: 35, 2004 10. Greene, F. L., Page, D. L. Fleming, I. D., Fritz, A., Balch, C. M., Haller, D. G. et al: AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag, pp. 335–340, 2002 11. Charlson, M. E., Pompei P., Ales, K. and MacKenzie, C. R.: A
BODY MASS INDEX AND SURVIVAL AFTER RADICAL CYSTECTOMY
12. 13. 14. 15. 16.
new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 40: 373, 1987 Extermann, M.: Measuring comorbidity in older cancer patients. Eur J Cancer, 36: 453, 2000 Collet, D.: Modelling Survival Data in Medical Research. London: Chapman and Hall, 1994 Gray, R. J.: A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat, 16: 1141, 1988 Fine, J. P. and Gray, R. J.: A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc, 94: 496, 1999 Molarius, A., Seidell, J. C., Kuulasmaa, K., Dobson, A. J. and Sans, S.: Smoking and relative body weight: an international perspective from the WHO MONICA Project. J Epidemiol Community Health, 51: 252, 1997
1517
17. Henley, S. J., Flanders, W. D., Manatunga, A. and Thun, M. J.: Leanness and lung cancer risk: fact or artifact? Epidemiology, 13: 268, 2002 18. Burch, J. D., Rohan, T. E., Howe, G. R., Risch, H. A., Hill, G. B., Steele, R. et al: Risk of bladder cancer by source and type of tobacco exposure: a case-control study. Int J Cancer, 44: 622, 1989 19. Fleshner, N., Garland, J., Moadel, A., Herr, H., Ostroff, J., Trambert, R. et al: Influence of smoking status on the diseaserelated outcomes of patients with tobacco-associated superficial transitional cell carcinoma of the bladder. Cancer, 86: 2337, 1999 20. Miller, D. C., Taub, D. A., Dunn, R. L., Montie, J. E. and Wei, J. T.: The impact of co-morbid disease on cancer control and survival following radical cystectomy. J Urol, 169: 105, 2003