These authors contributed equally.
Obesity is associated with worse oncological outcomes in patients treated with radical cystectomy
Article first published online: 21 JUN 2012
© 2012 BJU International
Volume 111, Issue 2, pages 249–255, February 2013
How to Cite
Chromecki, T. F., Cha, E. K., Fajkovic, H., Rink, M., Ehdaie, B., Svatek, R. S., Karakiewicz, P. I., Lotan, Y., Tilki, D., Bastian, P. J., Daneshmand, S., Kassouf, W., Durand, M., Novara, G., Fritsche, H.-M., Burger, M., Izawa, J. I., Brisuda, A., Babjuk, M., Pummer, K. and Shariat, S. F. (2013), Obesity is associated with worse oncological outcomes in patients treated with radical cystectomy. BJU International, 111: 249–255. doi: 10.1111/j.1464-410X.2012.11322.x
- Issue published online: 29 JAN 2013
- Article first published online: 21 JUN 2012
- bladder cancer;
- radical cystectomy;
- urothelial carcinoma
What's known on the subject? and What does the study add?
- Little is known on the association between obesity and urothelial carcinoma of the bladder (UCB). Most studies have shown that higher body mass index (BMI) is associated with higher rates of perioperative complications. Only one study specifically investigated obesity and bladder cancer-specific outcomes and reported no significant association between higher BMI and disease-specific survival in patients with UCB treated with radical cystectomy. However, that study was limited by its small sample size and a high rate of preoperative therapies.
- In contrast to the only previous study evaluating the association of BMI with oncological outcomes in UCB, we found that obesity (BMI ≥30 kg/m2) was associated with features of biologically aggressive UCB and clinical outcomes after radical cystectomy and, even when adjusting for the effects of standard clinicopathological features, obesity remained an independent predictor of cancer recurrence, cancer-specific mortality and overall mortality.
- To investigate the association between body mass index (BMI) and oncological outcomes in patients after radical cystectomy (RC) for urothelial carcinoma of the bladder (UCB) in a large multi-institutional series.
Patients and Methods
- Data were collected from 4118 patients treated with RC and pelvic lymphadenectomy for UCB. Patients receiving preoperative chemotherapy or radiotherapy were excluded.
- Univariable and multivariable models tested the effect of BMI on disease recurrence, cancer-specific mortality and overall mortality.
- BMI was analysed as a continuous and categorical variable (<25 vs 25–29 vs ≥30 kg/m2).
- Median BMI was 28.8 kg/m2 (interquartile range 7.9); 25.3% had a BMI <25 kg/m2, 32.5% had a BMI between 25 and 29.9 kg/m2, and 42.2% had a BMI ≥30 kg/m2.
- Patients with a higher BMI were older (P < 0.001), had higher tumour grade (P < 0.001), and were more likely to have positive soft tissue surgical margins (P = 0.006) compared with patients with lower BMI.
- In multivariable analyses that adjusted for the effects of standard clinicopathological features, BMI >30 was associated with higher risk of disease recurrence (hazard ratio (HR) 1.67, 95% confidence interval (CI) 1.46–1.91, P < 0.001), cancer-specific mortality (HR 1.43, 95% CI 1.24–1.66, P < 0.001), and overall mortality (HR 1.81, CI 1.60–2.05, P < 0.001). Themain limitation is the retrospective design of the study.
- Obesity is associated with worse cancer-specific outcomes in patients treated with RC for UCB.
- Focusing on patient-modifiable factors such as BMI may have significant individual and public health implications in patients with invasive UCB.
urothelial carcinoma of the bladder
body mass index
Obesity is a growing epidemic worldwide; it has been estimated that in 2015, 75% of American adults will be overweight, and 41% will be deemed obese . Unfortunately, these epidemiological trends continue to worsen and obesity has reached epidemic levels in parts of the USA and Europe . Obesity is strongly associated with heart disease, diabetes and various medical disorders. Moreover, researchers have noted an association between obesity and the incidence of several organ-specific cancers [3-6], such as cancer of the breast, colon, kidney and prostate [7-10]. Conversely, little is known on the association between obesity and urothelial carcinoma of the bladder (UCB). Most studies have shown that higher body mass index (BMI) is associated with higher rates of perioperative complications [11-13], but the only study specifically investigating obesity and bladder cancer-specific outcomes reported no significant association between higher BMI and disease-specific survival in patients with UCB treated with radical cystectomy (RC) . However, this study was limited by its small sample size and a high rate of preoperative therapies. We hypothesized that obesity would not be associated with worse cancer-specific outcomes but might impact other-cause mortality and tested this hypothesis in a large multi-institutional cohort of over 4000 patients treated with RC for UCB.
Patients and Methods
This was an institutional review board-approved study with all participating sites providing the necessary institutional data-sharing agreements before initiation. A total of 12 centres worldwide provided data. The records of 4118 patients treated with RC for UCB between 1979 and 2008 were reviewed. Patients receiving preoperative chemotherapy or radiotherapy were excluded. No patient had known metastatic disease at the time of surgery, as shown by radiographic or nuclear imaging. A total of 912 patients received adjuvant chemotherapy at the provider's discretion. Indication for RC was muscle-invasive disease or invasion into the prostatic stroma, or recurrent Ta, T1, or carcinoma in situ refractory to transurethral resection with or without intravesical chemotherapy or immunotherapy.
All surgical specimens were processed according to standard pathological procedures. Genitourinary pathologists assigned tumour grade according to the 1973 WHO grading system. Pathological stage was reassigned according to the 2002 American Joint Committee on Cancer tumour node metastasis (TNM) staging system. Positive soft tissue surgical margin status was defined as tumour at inked areas of soft tissue on the RC specimen; urethral or ureteral margin status was not considered positive in this analysis . Lymphovascular invasion was defined as the unequivocal presence of tumour cells in an endothelium-lined space without underlying muscular walls .
Multiple clinical variables were extracted from patient charts, including age and gender. The BMI was calculated based on individual weight and height, and defined as individual body weight (kg) divided by the square of the height (m2) and was evaluated at time of surgery. The BMI was analysed as continuous and categorical variables (<25 vs 25–29.9 vs ≥30 kg/m2).
Follow-up was performed according to institutional protocols. Patients were seen postoperatively at least every 3–4 months in year 1, semi-annually in year 2, and annually thereafter. Follow-up visits consisted of physical examination and serum chemistry evaluation, including liver function tests and alkaline phosphatase measurement. Diagnostic imaging of the upper tract and chest radiography were performed at least annually, or as clinically indicated. Detection of cancer in the ureter or urethra was coded as a second (metachronous) primary lesion, and not as local or distant recurrence. Patients identified as having died of UCB had progressive, disseminated and often symptomatic metastasis at death. Perioperative mortality (30 days after RC) was censored at death for UCB-specific survival analysis.
The Fisher's exact test and the chi-squared test were used to evaluate the association between categorical variables. Differences in variables with a continuous distribution across categories were assessed using the Mann–Whitney U-test (two categories) and Kruskal–Wallis test (three and more categories). Spearman's rank correlation coefficient was used to compare ordinal and continuous variables. Univariable recurrence and survival probabilities were estimated using the Kaplan–Meier method. All variables that were significant on univariable analyses were included in the multivariable analyses. Univariable and multivariable Cox regression models addressed time to recurrence and mortality. In all models, proportional hazards assumptions were verified using the Grambsch–Therneau residual-based test. All reported P values are two-sided and significance was set at 0.05. Analyses were performed with SPSS® 17 (SPSS Inc., IBM Corp., Somers, NY, USA).
Median patient BMI was 28.8 kg/m2 (interquartile range (IQR) 7.9); 25.3% had a BMI <25 kg/m2, 32.5% had a BMI between 25 and 29.9 kg/m2 and 42.2% had a BMI ≥30 kg/m2. The association of BMI with clinicopathological features is shown in Table 1. Patients with a higher BMI were older at the time of RC (P < 0.001), had higher grade disease (P < 0.001), and were more likely to have positive soft tissue surgical margins (P = 0.006). Patients with a higher BMI were also more likely to receive adjuvant chemotherapy (P < 0.001). However, BMI was not associated with lymphovascular invasion (P = 0.812), concomitant carcinoma in situ (P = 0.336), gender (P = 0.313) or positive lymph node status (P = 0.065). There was a significant difference in BMI between tumour stages, but there was no trend in any direction.
|BMI continuous||BMI categorical|
|N (%)||Median||IQR||P value||<25||25–29.9||>30||P value|
|All||4118 (100)||28.8||7.9||–||1041 (25.3)||1338 (32.5)||1739 (42.2)||–|
|Male||3279 (79.6)||28.9||7.8||816 (24.9)||1064 (32.4)||1399 (42.7)|
|Female||839 (20.4)||28.5||8.1||225 (26.8)||274 (32.7)||340 (40.5)|
|Cystectomy pT stage||<0.001||<0.001|
|T0||226 (5.5)||27.7||8.3||85 (37.6)||64 (28.3)||77 (34.1)|
|Ta||120 (2.9)||29.9||9.6||31 (25.8)||29 (24.2)||60 (50.0)|
|Tis||407 (9.9)||29.4||6.7||72 (17.7)||145 (35.6)||190 (46.1)|
|T1||560 (13.6)||28.9||8.1||161 (28.8)||164 (29.3)||235 (42.0)|
|T2||1000 (24.3)||28.9||7.9||262 (26.2)||309 (30.9)||429 (42.9)|
|T3||1279 (31.1)||28.7||7.7||309 (24.2)||434 (33.9)||536 (41.9)|
|T4||526 (12.8)||28.6||7.5||121 (23.0)||193 (36.7)||212 (40.3)|
|Cystectomy tumour grade||<0.001||<0.001|
|0||226 (5.5)||27.7||8.3||85 (37.6)||64 (28.3)||77 (34.1)|
|1||72 (1.7)||28.0||8.6||24 (33.3)||20 (27.8)||28 (38.9)|
|2||1859 (45.1)||28.1||7.8||557 (30.0)||601 (32.3)||701 (37.7)|
|3||1961 (47.6)||29.6||6.6||375 (19.1)||653 (33.3)||933 (47.6)|
|Absent||2718 (66.0)||28.9||7.9||694 (25.5)||869 (32.0)||1155 (42.5)|
|Present||1400 (34.0)||28.6||7.7||347 (24.8)||469 (33.5)||584 (41.7)|
|Lymph node status||0.065||0.073|
|Negative||3088 (75.0)||29.0||7.8||771 (25.0)||982 (31.8)||1335 (43.2)|
|Positive||1030 (25.0)||28.0||7.9||270 (26.2)||356 (34.6)||404 (39.2)|
|Soft tissue surgical margin||0.006||0.003|
|Negative||3887 (94.4)||28.7||7.8||997 (25.6)||1273 (32.8)||1617 (41.6)|
|Positive||231 (5.6)||30.9||7.8||44 (19.0)||65 (28.1)||122 (52.8)|
|Concomitant carcinoma in-situ||0.336||0.384|
|Absent||2048 (49.7)||28.8||8.1||537 (26.2)||658 (32.1)||853 (41.7)|
|Present||2070 (50.3)||28.8||7.8||504 (24.3)||680 (32.9)||886 (42.8)|
|None||3206 (77.9)||29.4||7.8||736 (23.0)||1008 (31.4)||1462 (45.6)|
|Received||912 (22.1)||26.3||7.5||305 (33.4)||330 (36.2)||277 (30.4)|
|Below 50||294 (7.1)||25.2||4.5||138 (46.9)||156 (53.1)||0 (0)|
|50–59.9||753 (18.3)||30.5||7.9||156 (20.7)||191 (25.4)||406 (53.9)|
|60–69.9||1504 (36.5)||28.6||8.1||424 (28.2)||451 (30.0)||629 (41.8)|
|70–79.9||1312 (31.9)||29.3||7.5||280 (21.3)||444 (33.8)||588 (44.8)|
|>80||255 (6.2)||29.5||5.8||43 (16.9)||96 (37.6)||116 (45.5)|
The median age and numbers of lymph nodes removed was 67 years (IQR 13.86) and 18 (IQR 20), respectively. Age and numbers of lymph nodes removed, when analysed as continuous variables, were weakly correlated with BMI (correlation coefficients 0.086, P < 0.001, and −0.052, P = 0.001, respectively). When BMI was analysed as a categorical variable, patients with a BMI ≥30 kg/m2 had significantly fewer (P < 0.001) lymph nodes removed (median 17, IQR 19), compared with patients with a BMI <25 kg/m2 (median 19, IQR 19) and those with a BMI between 25 and 29.9 kg/m2 (median 20, IQR 21).
Median follow-up in patients who were alive at last follow-up was 44 months (IQR 68.4). During follow-up, disease recurred in 1365 patients (33.1%), and 1121 (27.2%) died of UCB. Actuarial recurrence-free survival estimates at 3, 5 and 10 years were 63 ± 1% (±se), 60 ± 1% and 57 ± 1%, respectively. Actuarial cancer-specific survival estimates at 3, 5 and 10 years were 70 ± 1%, 66 ± 1% and 60 ± 1%, respectively. Actuarial overall survival estimates at 3, 5 and 10 years were 61 ± 1%, 53 ± 1% and 19 ± 1%, respectively.
In univariable Cox regression analyses, the risk of disease recurrence, cancer-specific mortality and overall mortality was highest in obese patients (≥30 kg/m2). Pairwise comparisons showed that patients with a BMI ≥30 kg/m2 were at higher risk for disease recurrence and cancer-specific mortality than patients with a BMI <25 kg/m2 and those with a BMI between 25 and 29.9 kg/m2 (each Mantel–Cox log-rank test P < 0.001, Figs 1, 2), but this was not true for patients with a BMI between 25 and 29.9 kg/m2 compared with patients of normal weight (P = 0.801 and P = 0.362, respectively). The rate of overall mortality was higher in patients with a BMI ≥30 kg/m2 compared with those with a BMI between 25–29.9 kg/m2, who in turn had a higher probability of overall death compared to those with a BMI <25 kg/m2 (each P < 0.001, Fig. 3).
In multivariable Cox proportional hazards regression analyses, BMI >30 coded as a categorical variable was an independent predictor of disease recurrence (HR 1.67, 95% CI 1.46–1.91, P < 0.001), cancer-specific mortality (HR 1.43, 95% CI 1.24–1.66, P < 0.001), and overall mortality (HR 1.81, 95% CI 1.60–2.05, P < 0.001) (Table 2). The same was true for BMI when analysed as a continuous variable (all P values <0.001). Subgroup analyses in patients who did not undergo adjuvant chemotherapy did not change the significance of these associations.
|Recurrence||Cancer-specific death||Overall death|
|HR||95% CI||P value||HR||95% CI||P value||HR||95% CI||P value|
|Cystectomy pT stage||<0.001||<0.001||<0.001|
|Cystectomy tumour gradea||0.410||0.019||0.190|
|Lymph node metastasis||2.01||1.76||2.29||<0.001||2.33||2.02||2.69||<0.001||2.05||1.83||2.30||<0.001|
|Soft tissue surgical margin positivity||1.67||1.40||1.99||<0.001||1.88||1.56||2.27||<0.001||1.54||1.30||1.82||<0.001|
|Concomitant carcinoma in situ||0.98||0.88||1.10||0.757||0.98||0.87||1.11||0.731||1.06||0.96||1.17||0.267|
Obesity is a growing worldwide epidemic and has a major impact on patients' health and their treatment modalities. Beside the relationships to known comorbidities, such as cardiovascular diseases and diabetes, obesity has also been shown to influence the course of various cancer types. Several epidemiological studies have assessed the relationship between obesity and the risk of cancer, most of them concluding that higher BMI increases the incidence of various malignancies [3-6]. Moreover, higher BMI has been shown to be associated with higher perioperative complication rates in patients treated with RC [11-13, 17]. The association of BMI with cancer-specific outcomes after RC remains poorly investigated.
In contrast to our initial hypothesis, we found that obesity (BMI ≥30 kg/m2) was associated with features of biologically aggressive UCB and clinical outcomes after RC and, even when adjusting for the effects of standard clinicopathological features, obesity remained an independent predictor of cancer recurrence, cancer-specific mortality and overall mortality. This is in contrast to the only previous study evaluating the association of BMI with oncological outcomes in UCB . The study of Hafron et al.  included only 288 patients and included patients with preoperative chemotherapy or radiation therapy (24%). Moreover, they had a higher proportion of overweight patients (41% vs 32.5%) and fewer obese patients (25% vs 42.2%). A large population-based study examining obesity and mortality from all types of cancer included 872 deaths from bladder cancer and found no association between BMI and cancer-specific mortality . Our study, as it was specifically designed to investigate the association between BMI and cancer-specific outcomes in UCB, included many more patients and was limited to those who underwent radical cystectomy.
The worse outcomes associated with obesity could be related to differences in time to diagnosis, differences in biological aggressiveness and differences in treatments including complications and response. One hypothesis for the worse oncological outcomes in obese patients could be that these patients have more aggressive disease based on an increased insulin level, which increases insulin-like growth factor-I . Insulin-like growth factor-I stimulates cell proliferation and suppresses apoptosis . Excess body fat is also associated with systemic inflammation, which may have a role in UCB outcomes, as suggested by studies that measured circulating levels of inflammatory markers [20-22]. In our study, patients with higher BMI were more likely to have worse tumour-specific characteristics such as higher tumour stage and grade. It is not possible to determine from this study whether the difference in stage of diagnosis is related to time to diagnosis or actual biological propensity for more aggressive behaviour. However, even after adjusting for tumour stage and grade, obesity was still associated with worse clinical outcomes.
Another hypothesis for the worse outcomes of obese patients could be greater perioperative complications after RC , leading to a lower rate of stage-adjusted chemotherapy, and more technical challenges during the procedure, leading to a higher rate of positive soft tissue surgical margins. This is in agreement with Butt et al. , who found that overweight and obese patients had significantly higher rates of positive surgical margins. Although we did not have data for perioperative complications, we found a higher rate of positive soft tissue surgical margins in obese patients but a high BMI was an independent predictor for recurrence even when controlling for positive soft tissue surgical margins. Evaluating response to chemotherapy is also difficult. Obese patients were less likely to receive adjuvant chemotherapy despite a more aggressive disease profile. The reason underlying this selection is difficult to uncover in a retrospective study; hypotheses include a worse general health status of obese patients precluding adjuvant chemotherapy in these patients or other selection bias. Finally, one explanation for the worse outcomes in obese patients could be the quality of care received. Indeed, obese patients had fewer lymph nodes removed than non-obese patients. The number of lymph nodes removed is a quality-of-care indicator and has been associated with outcomes [24, 25].
There are several important limitations to our study. First and foremost are the limitations inherent to retrospective analyses. For example, we found that administration of adjuvant chemotherapy was independently associated with improved overall survival, but not significantly associated with disease recurrence or cancer-specific mortality. Factors such as renal function not captured in the clinicopathological variables probably contributed to some selection of patients for adjuvant chemotherapy, resulting in these findings. Second, we are unable to comment on the extent to which BMI factored into the selection of RC among other available treatment options for patients, or how many patients were not treated with RC. Third, outcomes from multiple surgeons and surgical techniques were evaluated in this study. However, this can be interpreted as adding strength to this study, because it reflects real world practice and extends the generalizability of the results. Other study strengths include its large cohort size and the duration of follow-up. Fourth, the worse outcome in obese patients may be the result of delay in diagnosis or definitive therapy with curative intent .
In conclusion, in our study, obesity was associated with worse cancer-specific outcomes in patients treated with RC for UCB. This, together with the increasing evidence linking obesity to cancer-specific outcomes across multiple organ systems, suggests that metabolic syndrome is an important area of investigation and therapy in patients with UCB. Improving oncological outcomes by also focusing on patient-modifiable factors such as BMI could have significant individual and public health implications in patients with advanced UCB.
Conflict of Interest
- 8The effects of obesity and obesity-related conditions on colorectal cancer prognosis. Cancer Control 2010; 17: 52–57, , , , , .
- 9Obesity and oncological outcome after radical prostatectomy: impact of prostate-specific antigen-based prostate cancer screening: results from the Shared Equal Access Regional Cancer Hospital and Duke Prostate Center databases. BJU Int 2008; 102: 969–974, , et al.
- 13Age and body mass index are independent risk factors for the development of postoperative paralytic ileus after radical cystectomy. Urology 2010; 76: 1419–1424, , et al.
- 15Soft tissue surgical margin status is a powerful predictor of outcomes after radical cystectomy: a multicenter study of more than 4,400 patients. J Urol 2010; 183: 2165–2170, , et al.
- 23Impact of body mass index on robot-assisted radical cystectomy. JSLS 2008; 12: 241–245, , et al.
- 25Critical analysis and validation of lymph node density as prognostic variable in urothelial carcinoma of bladder. Urol Oncol 2011; Epub ahead of print, , et al.
- 26A delay in radical cystectomy of >3 months is not associated with a worse clinical outcome. BJU Int 2007; 100: 1015–1020, , et al.