Extreme body mass index is associated with poor survival outcomes after radical cystectomy: a retrospective cohort study in a Chinese population.

BACKGROUND: Body mass index (BMI) has been evidenced to be a significant prognostic factor in multiple cancers. This retrospective study aimed to investigate the association between BMI and survival outcomes after radical cystectomy (RC) in patients with bladder cancer (BCa).
METHODS: Clinical and pathological parameters of patients who were diagnosed with BCa and received RC between 2010 and 2018 were collected. The associations between BMI at surgery and clinicopathological features were examined. The prognostic value of BCa for overall survival (OS) and cancer-specific survival (CSS) was examined using the Kaplan-Meier method and Cox regression models.
RESULTS: Among the 217 patients enrolled in this study, 13 (6.0%), 121 (55.8%), 60 (27.6%), and 23 (10.6%) had a BMI value of <18.5 kg/m2 (underweight), 18.5-23.9 kg/m2 (normal), 24-27.9 kg/m2 (overweight), and ≥28 kg/m2 (obese), respectively. Underweight and obese patients tended to have poorer survival after RC than normal and overweight patients (P<0.05). Multivariable Cox regression revealed that extreme BMI was an independent predictor of both OS (BMI <18.5 vs. 18.5-27.9 kg/m2, OR =2.675, 95% CI: 1.131-6.327, P=0.025; BMI ≥28 vs. 18.5-27.9 kg/m2, OR =3.693, 95% CI: 1.589-8.583, P=0.002) and CSS (BMI <18.5 vs. 18.5-27.9 kg/m2, OR =3.012, 95% CI: 1.180-7.687, P=0.021; BMI ≥28 vs. 18.5-27.9 kg/m2, OR =3.801, 95% CI: 1.526-9.469, P=0.004), along with tumor stage and urinary diversion type.
CONCLUSIONS: Being underweight or obese is associated with a poor prognosis in patients with BCa undergoing RC. For patients who are preparing to undergo RC for BCa, controlling the BMI index through diet or exercise before surgery may contribute to the surgical curative effect and an improved prognosis. 2021 Translational Andrology and Urology. All rights reserved.

Introduction

Bladder cancer (BCa) is one of the most common cancers affecting the urinary system and has high morbidity and relapse rates (1). An estimated 573,278 new cases of BCa and 212,536 deaths from BCa occurred worldwide in 2020 (1,2). Over recent decades, much attention has been dedicated to establishing risk factors associated with the incidence and prognosis of BCa. Cigarette smoking, occupational exposure to aromatic amines, and a familial history of BCa have been well established as significant factors associated with BCa (3).

Body weight is a basic information for evaluating the physical condition of a patient, which is usually affected by living habits, hormone levels, and genetic factors. Excess body weight, which is commonly measured using body mass index (BMI), has been linked to the prognosis of various cancers, including esophageal, pancreatic, colorectal, endometrial, and prostate cancer (4-8). However, a conclusion on the association of BMI with BCa has yet to be reached, and the current guidelines do not consider BMI to be a risk factor for the disease (9-11). Although several studies have reported that obesity might be a poor prognostic factor for patients with advanced BCa following radical cystectomy (RC) (12,13), recent studies have found no such correlation (14,15).

To our knowledge, most studies investigating the association between body mass and oncologic outcomes after RC to date were performed in Western populations, and there is a dearth of research in Asian patients. Due to genetic and dietary differences, the Chinese cutoff points for overweight and obesity differ from those used in Western countries. Therefore, illuminating the association between BMI and oncologic outcomes based on a Chinese population would be extremely valuable. In this study, we retrospectively analyzed data from patients with BCa who underwent RC in a Chinese hospital and evaluated the association between BMI value and survival outcomes. We present the following article in accordance with the STROBE reporting checklist (available at https://dx.doi.org/10.21037/tau-21-871).

Methods

Study participants and design

Consecutive patients who were diagnosed with BCa and underwent RC in Shanghai Tenth People’s Hospital between January 2010 and June 2018 were enrolled in this study. According to the European Association of Urology (EAU) guidelines, indications for RC included muscle-invasive BCa or high-risk non-muscle invasive BCa. Patients who received neoadjuvant chemotherapy or radiotherapy were excluded.

Patients’ medical records were reviewed retrospectively for clinical and pathological information, including age, sex, tumor stage, size, tumor grade, histology, lymphovascular invasion, and Charlson Comorbidity Index (CCI), measured according to the standardized International Classification of Diseases, Ninth Revision (16).

Height and weight were measured at admission, and a patient’s BMI was calculated by dividing their weight in kilograms by the square of their height in meters (kg/m2). Since all the patients were from China, the BMI cut-offs were set according to the Chinese criteria (17), as follows: underweight (BMI <18.5 kg/m2), normal (BMI 18.5–23.9 kg/m2), overweight (BMI 24–27.9 kg/m2), and obese (BMI ≥28 kg/m2). This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by the Ethics Committee of Shanghai Tenth People’s Hospital (SHSY-IEC-4.1/20-24/01). Informed consent was taken from all individual participants.

Surgical procedure and pathological evaluation

Each patients underwent RC and pelvic lymph node dissection. Three urinary diversion methods were employed: cutaneous ureterostomy, ileal conduit, and orthotopic bladder substitution (Studer pouch). Tumor grade was evaluated according to the 2004 World Health Organization (WHO) classification by a single group of pathology doctors in Pathology Department of our hospital. Pathological stage was reassigned according to the 2010 American Joint Committee on Cancer TNM staging system. The definition of lymphovascular invasion was the presence of tumor cells in the lymphatic vessels and in the vascular walls.

Follow-up

Patients were followed up at 3-month intervals during the first 2 years after RC, and every 6 months thereafter. Follow-up consisted of medical history review, physical examination, blood laboratory tests, and urine sedimentation. Imaging examinations included computed tomography of the chest, abdomen and pelvis, and bone scanning, which were performed at 6 and 12 months postoperatively, and annually thereafter. Patients were followed until death or the cut-off date (November 30, 2018). The study endpoints were overall survival (OS) and cancer-specific survival (CSS). OS was defined as the length of time from the start of cystectomy to the date of any-cause death. CSS was defined as the length of time from the start of cystectomy to the date of death from BCa. The median follow-up period was 53 months (range, 4–112 months).

Statistical analysis

Clinicopathological variables were compared between four BMI groups (underweight, normal, overweight, and obese). Differences in clinical and pathological characteristics between groups were estimated using Student’s t-test, Wilcoxon signed-rank test, and Chi-square test. Then, the patients were divided into two groups: the underweight/obese group and normal/overweight group. Kaplan-Meier survival curves were generated for OS and CSS in these two BMI groups, and compared using the log–rank test. To identify predictive factors of OS and CSS, univariable and multivariate Cox regression models were performed using all the variables collected. All statistical tests were two-tailed, and P<0.05 was considered significant. Correlations of variables with outcomes were expressed as hazard ratio (HR) with 95% confidence interval (CI). All statistical analyses were performed with SPSS 20.0 (SPSS Inc., Chicago, IL, USA).

Results

Patients and procedure

This study enrolled 217 patients, including 189 males and 28 females. The flow chart for patient enrolment is shown in . Three patients were lost to follow-up. A total of 75 patients died, 64 of whom died due to BCa during follow-up.

Figure 1

Flow chart of the study population enrolment.

Baseline data

compares the patients’ clinical and pathological variables according to BMI category. The mean BMI was 23.3 kg/m2 and the BMI distribution is shown in . Of the 217 patients in this study, 13 (6.0%), 121 (55.8%), 60 (27.6%) and 23 (10.6%) had a BMI value of <18.5 kg/m2 (underweight), 18.5–23.9 kg/m2 (normal), 24–27.9 kg/m2 (overweight), and ≥28 kg/m2 (obese), respectively. Patients with low BMI values tended to be older than those with normal or high BMI values at the time of RC (P=0.009) and tended to receive cutaneous ureterostomy as opposed to another urinary diversion procedure (P=0.005). Analysis of other variables, including sex, stage, tumor size, tumor grade, histology, lymphovascular invasion, and CCI score, uncovered no significant differences between the different BMI categories.

Table 1

Comparison of clinical/pathological features of the patient cohort according to BMI category

Variables	Underweight (<18.5 kg/m2)	Normal (18.5–23.9 kg/m2)	Overweight (24–27.9 kg/m2)	Obese (≥28 kg/m2)	P value
Number	13	121	60	23
Age (years)					0.009
Mean	73.92	67.07	63.6	65.35
Median (range)	73 (55–82)	67 (37–87)	63.5 (32–83)	66 (29–80)
Sex, n (%)					0.486
Female	1 (7.7)	14 (11.6)	11 (18.3)	2 (8.7)
Male	12 (92.3)	107 (88.4)	49 (81.7)	21 (91.3)
Stage, n (%)					0.671
Tis	0 (0.0)	3 (2.5)	0 (0.0)	0 (0.0)
I	4 (30.8)	42 (34.7)	25 (41.7)	10 (43.5)
II	3 (23.1)	26 (21.5)	8 (13.3)	5 (21.7)
III	6 (46.2)	47 (38.8)	22 (36.7)	7 (30.4)
IV	0 (0.0)	3 (2.5)	5 (8.3)	1 (4.3)
Size, n (%)					0.552
<3 cm	6 (46.2)	56 (46.3)	35 (58.3)	13 (56.5)
≥3 cm	6 (46.2)	58 (47.9)	23 (38.3)	10 (43.5)
N/A	1 (7.7)	7 (5.8)	2 (3.3)	0 (0.0)
Grade, n (%)					0.445
Low	0 (0.0)	9 (7.4)	7 (11.7)	1 (4.5)
High	13 (100)	112 (92.6)	53 (88.3)	21 (95.5)
Histology*, n (%)					0.121
Pure UCC	9 (69.2)	107 (88.4)	49 (81.7)	17 (73.9)
Not pure UCC	4 (30.8)	14 (11.6)	11 (18.3)	6 (26.1)
LVI, n (%)					0.822
Yes	4 (33.3)	33 (27.5)	14 (23.3)	5 (21.7)
No	8 (66.7)	87 (72.5)	46 (76.7)	18 (78.3)
Surgery type, n (%)					0.005
Ureterostomy	10 (76.9)	56 (46.3)	16 (26.7)	6 (26.1)
Ileal conduit	2 (15.4)	37 (30.6)	25 (41.7)	13 (56.5)
Neobladder	1 (7.7)	28 (23.1)	19 (31.7)	4 (17.4)
CCI [median (P25, P75)]	3 (2, 4)	2 (2, 3)	2 (2, 3)	2 (2, 3)	0.130

*, pathology showing another histological component such as squamous cell carcinoma or adenocarcinoma was defined as “not pure UCC”. BMI, body mass index; LVI, lymphovascular invasion; UCC, urothelial cell carcinoma; CCI, Charlson Comorbidity Index.

Figure 2

Distribution of BMI values. BMI, body mass index.

Survival estimates after RC

The Kaplan-Meier survival curves for OS and CSS according to BMI category are displayed in . Patients who were underweight or obese tended to have worse survival than those who were of a normal weight or overweight. For underweight, normal, overweight, and obese patients, the 5-year OS rate was 33%, 70.2%, 62.7%, and 38%, respectively, and the 5-year CSS rate was 40%, 73%, 66.8%, and 46%, respectively. Subsequently, the patients were divided into three categories: BMI value <18.5 kg/m2, BMI value 18.5–27.9 kg/m2, and BMI value ≥28 kg/m2. The OS and CSS survival curves further illustrated that underweight and obese patients had significantly worse survival than patients who were of a normal weight or overweight ().

Figure 3

Kaplan-Meier analysis of patient survival after radical cystectomy for bladder cancer according to the four BMI categories. Left: OS; right: CSS. BMI, body mass index; OS, overall survival; CSS, cancer-specific survival.

Figure 4

Kaplan-Meier analysis of patient survival after radical cystectomy for bladder cancer according to three BMI categories. Left: OS; right: CSS. BMI, body mass index; OS, overall survival; CSS, cancer-specific survival.

Prognostic factors associated with OS after RC

Univariate analysis revealed tumor stage, lymphovascular invasion, urinary diversion type, and three-category variable BMI to be associated with OS. Multivariable analysis further identified extreme BMI as an independent predictor of OS [BMI <18.5 vs. 18.5–27.9 kg/m2, odds ratio (OR) =2.675, 95% CI: 1.131–6.327, P=0.025; BMI ≥28 vs. 18.5–27.9 kg/m2, OR =3.693, 95% CI: 1.589–8.583, P=0.002], along with tumor stage (stage IV vs. I, OR =9.247, 95% CI: 3.049–28.046, P<0.001) and urinary diversion type (Bricker vs. ureterostomy, OR =0.415, 95% CI: 0.213–0.807, P=0.010; orthotropic bladder vs. ureterostomy, OR =0.328, 95% CI: 0.151–0.711, P=0.005) ().

Table 2

Univariable and multivariable Cox regression models evaluating the OS of patients with bladder cancer after radical cystectomy

Variables	Univariate			Multivariate
	RR (95% CI)	P value		RR (95% CI)	P value
Age	1.021 (0.998–1.044)	0.073		0.989 (0.960–1.018)	0.443
Sex (male vs. female)	1.126 (0.594–2.136)	0.716		0.705 (0.339–1.467)	0.350
Stage*	Ref
Tis	–	0.966		–	0.976
II	3.473 (1.618–7.456)	0.001		3.656 (1.603–8.339)	0.002
III	5.409 (2.796–10.462)	<0.001		6.090 (2.853–12.998)	<0.001
IV	10.052 (3.899–25.916)	<0.001		9.247 (3.049–28.046)	<0.001
Size (<3 vs. ≥3 cm)	1.235 (0.780–1.956)	0.368		1.132 (0.668–1.918)	0.646
Grade (low vs. high)	1.657 (0.665–4.131)	0.279		1.316 (0.485–3.567)	0.590
Histology (UCC vs. not pure UCC)	1.342 (0.751–2.400)	0.321		0.690 (0.351–1.356)	0.281
LVI (yes vs. no)	0.525 (0.323–0.851)	0.009		0.844 (0.492–1.448)	0.537
Surgery type**	Ref
Bricker	0.333 (0.190–0.582)	<0.001		0.415 (0.213–0.807)	0.010
Orthotropic bladder	0.350 (0.191–0.642)	0.001		0.328 (0.151–0.711)	0.005
CCI	1.174 (0.919–1.499)	0.198		1.179 (0.836–1.664)	0.348
BMI***	Ref
<18.5 kg/m2	2.917 (1.386–6.141)	0.005		2.675 (1.131–6.327)	0.025
≥28 kg/m2	2.039 (1.002–4.148)	0.048		3.693 (1.589–8.583)	0.002

*, stage I was used as a reference value; **, ureterostomy was used as a reference value; ***, BMI value within 18.5–27.9 kg/m2 was used as a reference value. OS, overall survival; LVI, lymphovascular invasion; BMI, body mass index; CCI, Charlson Comorbidity index.

Prognostic factors associated with CSS after RC

Regarding CSS, univariate analysis identified tumor stage, lymphovascular invasion, urinary diversion type, and three-category variable BMI as significant predictors. Subsequent multivariate analysis showed that extreme BMI was still an independent predictor of CSS (BMI <18.5 vs. 18.5–27.9 kg/m2, OR =3.012, 95% CI: 1.180–7.687, P=0.021; BMI ≥28 vs. 18.5–27.9 kg/m2, OR =3.801, 95% CI: 1.526–9.469, P=0.004), along with tumor stage (stage IV vs. I, OR =20.436, 95% CI: 5.795–72.066, P<0.001) and urinary diversion type (Bricker vs. ureterostomy, OR =0.456, 95% CI: 0.222–0.934, P=0.032; orthotropic bladder vs. ureterostomy, OR =0.295, 95% CI: 0.127–0.684, P=0.004) ().

Table 3

Univariable and multivariable Cox regression models evaluating the CSS of patients with bladder cancer after radical cystectomy

Variables	Univariate			Multivariate
	RR (95% CI)	P value		RR (95% CI)	P value
Age	1.018 (0.994–1.043)	0.141		0.985 (0.954–1.017)	0.356
Sex (male vs. female)	1.076 (0.513–2.257)	0.847		0.951 (0.409–2.210)	0.907
Stage*	Ref
Tis	–	0.972		–	0.980
II	5.170 (2.054–13.018)	<0.001		5.903 (2.143–16.262)	0.001
III	8.108 (3.575–18.393)	<0.001		9.559 (3.702–24.683)	<0.001
IV	16.762 (5.812–48.339)	<0.001		20.436 (5.795–72.066)	<0.001
Size (<3 vs. ≥3 cm)	1.359 (0.825–2.238)	0.229		1.294 (0.727–2.302)	0.381
Grade (low vs. high)	1.717 (0.622–4.741)	0.297		1.154 (0.380–3.510)	0.800
Histology (UCC vs. not pure UCC)	1.619 (0.895–2.928)	0.111		0.808 (0.405–1.613)	0.545
LVI (yes vs. no)	0.498 (0.296–0.837)	0.008		0.841 (0.471–1.500)	0.558
Surgery type**	Ref
Bricker	0.328 (0.178–0.602)	<0.001		0.456 (0.222–0.934)	0.032
Orthotropic bladder	0.335 (0.174–0.646)	0.001		0.295 (0.127–0.684)	0.004
CCI	1.140 (0.871–1.493)	0.339		1.166 (0.794–1.710)	0.434
BMI***	Ref
<18.5 kg/m2	3.058 (1.378–6.784)	0.006		3.012 (1.180–7.687)	0.021
≥28 kg/m2	2.179 (1.023–4.641)	0.043		3.801 (1.526–9.469)	0.004

*, stage I was used as a reference value; **, ureterostomy was used as a reference value; ***, BMI value within 18.5–27.9 kg/m2 was used as a reference value. OS, overall survival; LVI, lymphovascular invasion; BMI, body mass index; CCI, Charlson Comorbidity index.

Discussion

Obesity has become prevalent worldwide and is a growing health problem even in developing countries. The WHO criteria define a BMI of 25 to 30 kg/m2 as overweight and a BMI of over 30 kg/m2 as obesity (18). Despite the growing influence of Westernized lifestyles and diets in China, the incidence of obesity was only 11.9% and the average BMI was 23.6 kg/m2 among adults in 2010–2012 (19). Therefore, the Chinese criteria, which define overweight as BMI 24–28 kg/m2 and obesity as BMI ≥28 kg/m2 (19), were considered to be more suitable for this study. The participants in our study were first divided into four groups: underweight, normal, overweight, and obese. We found that having a BMI at either end of the scale was associated with a worse survival outcome. Next, we divided the study participants into three groups: two groups of patients with extreme BMI (underweight and obese) and one group of patients with average BMI (normal/overweight). The differences in survival outcome between the patients with extreme BMI and those with average BMI were statistically significant.

For advanced BCa, RC and urinary diversion remain the gold standard (20). However, until now, it remained uncertain whether having a high BMI could result in a worse prognosis after cystectomy. A large sample study of 4,118 patients by Chromecki et al. showed that obesity was linked to poor cancer-specific outcomes (13). However, a recent prospective multicenter study of 671 patients showed inconsistent results, even finding that patients with obesity had better OS than patients of a normal weight (14). At present, there are few relevant studies conducted in the Chinese population. Because the BMI classification standard of Chinese people is significantly different from that of the western population, research in Chinese population may be helpful to further evaluate the relationship between BMI and RC prognosis. The results of the present study in a Chinese population support the view that obesity can lead to a worse prognosis after RC.

The reasons why increased BMI is related to worse oncological outcomes may be complex. One common explanation is that obese patients have an unclarified anatomy due to excessive fat tissue and are more difficult to operate on, which could result in an increased risk of surgical complications, longer operation times, and more blood loss (21). A study comprising 2,240 patients found that BMI as a continuous variable was an independent predictor of 90-day major complications (22). Another prospective study involving 548 patients who underwent laparoscopic RC also found that high BMI was independently associated with a higher overall risk of postoperative complications (23). However, the occurrence of short-term complications does not necessarily translate to a poor survival outcome. In our study, fatal complications, such as repeated infections or renal dysfunction, were not common in any of the three BMI groups. A large decline in survival in the two extreme BMI groups occurred within the first 12–16 months after surgery, which was in accordance with the observations of a previous report, in which most local recurrence occurred within 6–18 years after cystectomy and 90% of distant recurrence appeared within the first 2 years (24). In our study, the OS and CSS rates were similar, which means that BCa progression was responsible for most of the deaths recorded.

Another explanation for obesity’s association with poor oncological outcomes may lie in its close relationship with inflammation. The biological mechanism by which obesity affects cancer development involves several pathways, including the insulin/insulin-like growth factor pathway, sex steroids, adipokines, and a state of chronic inflammation (25). Considering that sex differences contribute to a poor prognosis after RC, obesity might lead to hormonal changes, as excessive lipids cells can produce estrogen (26). The increased expression of estrogen receptor beta has been implicated in the development of urothelial cancer (27). There is potential for differential expression of this receptor in patients with obesity; however, this has not specifically been studied. Moreover, obesity can increase the levels of insulin, insulin-like growth factor-1, and inflammatory cytokines, which exert mitogenic, anti-apoptotic proangiogenic functions and facilitate tumor growth (28).

One important finding of our study is that being underweight can also lead to a poor prognosis. The reason might be that underweight patients were older and tended to undergo ureterostomy. Multiple studies have evidenced that having underweight status predicts a poor prognosis of other cancers, including gastric and colorectal cancer (29,30). Theoretically, low body mass is usually an indication to low serum albumin and poor nutritional status. Moreover, excessive weight loss might be associated with cancer cachexia, which is a complex metabolic condition characterized by loss of skeletal muscle and body weight that develops in progressive disease (31).

Our study has some limitations. First, it is a retrospective design and is hard to avoid selection bias. Second, BMI alone cannot fully reflect a patient’s adiposity status. Waist circumference and the waist-to-hip ratio are commonly used in conjunction with BMI to approximate central adiposity and intra-abdominal visceral fat, which are now considered to be equally crucial indicators of health status. Third, the study did not include serum markers to reflect nutrition status. Nevertheless, our study adds more evidence of the importance of BMI to patient outcomes after RC, which has been rarely studied in Chinese population. According to the results of this research, for patients who are diagnosed with bladder cancer and potentially need to undergo RC, more attention should be paid to maintaining a normal BMI level.

Conclusions

Our results suggest that Chinese patients who are underweight or obese have a worse prognosis after undergoing RC for BCa than patients who are normal weight or overweight. Therefore, in addition to conventional risk factors, BMI could serve as a predictor of BCa prognosis. Multicenter prospective research is warranted to confirm these conclusions.

The article’s supplementary files as