The prognostic significance of advanced age in patients with bladder cancer treated with radical cystectomy


Berkan Resorlu, Urology, School of Medicine, University of Ankara, Kardesler Koop., 3.cad., 30.sok. No:3/4 Asagi Ayvali Ankara 06010, Turkey.



To evaluate the association of patient age with pathological and long-term oncological outcomes after radical cystectomy (RC) for bladder carcinoma, as this disease, like many others, increases in incidence with age.


We retrospectively reviewed 241 consecutive patients with invasive bladder cancer who had RC between 1990 and 2007. The age at RC was analysed both as a continuous and categorical (≤50 years, 38 patients; 51–69, 172; or ≥70, 31) variable. Survival was also analysed.


Increasing age, analysed as a continuous and categorical variable, was associated with advanced pathological stage (P = 0.009 and 0.006, respectively). The 5-year cancer-specific survival rates for patients according to the age groups were 78.5%, 44.9% and 28.1%, respectively, and Kaplan-Meier analysis showed an increased risk of bladder cancer-specific death with advancing age (P < 0.001). Being older at RC was an important prognostic factor for disease-specific survival in a multivariate Cox regression model. Patients aged ≥70 years had a significantly higher risk of disease than patients aged ≤50 years (P = 0.002).


Higher age at RC is significantly associated with the risk of pathologically advanced disease and poorer cancer-specific survival. More prospective work is needed to examine the impact of age on tumour biology and cancer-specific survival.


radical cystectomy


cancer-specific survival.


Bladder carcinoma occurs predominantly in older patients and overall, 20–40% of patients with TCC of the bladder will present or develop muscle-invasive disease [1]. As the population ages due to improvements in health care, the incidence of bladder cancer, and accordingly the number of cases of invasive bladder cancer, is expected to rise [2]. Currently radical cystectomy (RC) is considered the treatment of choice for muscle-invasive bladder carcinoma [3,4] and therefore, in future years, urologists will have frequent opportunities to consider RC for elderly patients.

In several studies, as the overall and disease recurrence-free survival after RC was worse among older than younger patients, some physicians have advocated the use of other, less invasive and potentially less effective forms of treatment for invasive bladder carcinoma in the elderly [1,5]. However, other studies showed that RC with urinary diversion can be safe in carefully selected elderly patients [6–8]. In the present study we retrospectively assessed the influence of age on the pathological characteristics and long-term oncological outcomes after RC.


We retrospectively analysed 241 consecutive patients who had had RC, pelvic lymphadenectomy and urinary diversion for bladder cancer between 1990 and 2007 at our department. Data were collected from retrospective reviews of hospital and physicians’ office records, and by contact with the patients. Bladder cancer was histologically diagnosed by transurethral resection; all patients had a physical examination, laboratory studies and chest radiography, and CT and abdominal ultrasonography were used for clinical staging. If indicated, a bone scan was also taken. The tumour stage and grade were recorded according to the 2002 TNM system and the WHO system, respectively. The indication for RC included muscle-invasive bladder cancer, or non-muscle-invasive disease refractory to intravesical chemotherapy and/or immunotherapy. No patient had distant metastatic disease at the time of RC. The surgical procedures for RC remained unchanged during the study period. In men, the bladder, prostate and seminal vesicles were resected, while in women, the bladder, anterior vaginal wall and uterus were removed. Urethrectomy was performed in cases diagnosed as having histologically confirmed cancer of the prostatic urethra before RC or at the time of surgery, by examining frozen sections of the urethral margin. The patients were initially seen 1 month after surgery, then every 3 months for the first year, every 6 months for the second year, and annually thereafter. Follow-up visits consisted of a physical examination and serum chemistry evaluation. Diagnostic imaging was used at least annually or when clinically indicated. In all, 241 patients were categorized into three age groups aged ≤50 years at RC (38, 15.8%), 51–69 years (172, 71.4%) and ≥70 years (31, 12.8%).

The chi-squared test was used to evaluate the association between categorical variables. Differences in variables with a continuous distribution across dichotomous or ranked categories were assessed using the Mann–Whitney U-test and the Kruskal–Wallis nonparametric anova, respectively. For the multivariate survival analyses we used the Cox proportional hazard regression model. The cancer-specific survival (CSS) rates of patients were estimated by the Kaplan–Meier method, and the differences in the rates of survival determined by the log-rank test. In all tests, P < 0.05 were considered to indicate significance.


The median (range) age of the 241 patients (214 men, 88.8%; and 27 women, 11.2%) at RC was 59.8 (29–83) years. The association of age at RC with clinicopathological characteristics is shown in Table 1; the mean (range) follow-up was 34 (1–175) months for patients alive at the last follow-up. There was a significant association between age (both as a continuous and a categorical variable) and the proportion of patients with advanced stage (P = 0.009 and 0.006, respectively; Table 1). When evaluating other pathological variables there was no significant difference amongst the three age groups in the proportion of patients with high-grade disease or lymph-node positivity (Table 1). However, when age was used as a continuous variable there was a significant association between age and other pathological variables, i.e. grade and lymph-node positivity (Table 1). In univariate Cox regression analyses, age, analysed as a continuous variable, was associated with bladder cancer-specific death (hazard ratio 1.036, 95% CI 1.011–1.062, P = 0.005). The 5-year CSS rates for the patients aged ≤50, 51–69 and ≥70 years were 78.5%, 44.9% and 28.1%. Kaplan-Meier analysis showed an increased risk of bladder cancer-specific death with advancing age (log rank test P < 0.001; Fig. 1). To determine statistically important risk factors for disease-specific survival, age, sex, hydronephrosis, tumour stage, grade, nodal status and lymphovascular invasion were considered as independent variables in the multivariate model. Age, presence of hydronephrosis and nodal status were significant risk factors for CSS after using a forward-Wald elimination technique (Table 2). In the final model, CSS significantly decreased for elderly patients (P = 0.002).

Table 1.  The association of age at RC as a continuous and categorical variable with pathological characteristics in 241 patients who had RC for bladder carcinoma
 N (%)ContinuousCategorical, n (%)
Median (range)P≤5051–69≥70P
  • *

    Statistically significant at P < 0.05.

All241 (100)60 (29–83) 38 (15.8)172 (71.4)31 (12.8) 
 Male214 (88.8)62 (33–83) 32 (84.2)156 (90.7)26 (83.9) 
 Female 27 (11.2)60 (29–80)0.921 6 (15.8) 16 (9.3) 5 (16.1)0.335
pT stage       
 T0 31 (12.9)58 (33–75)  4 (10.5) 26 (15.1) 1 (3.2) 
 T1 52 (21.6)56.5 (29–73) 17 (44.7) 32 (18.6) 3 (9.7) 
 T2 66 (27.4)60 (46–83)  5 (13.2) 51 (29.7)10 (32.3) 
 T3 57 (23.7)63 (36–78)  8 (21.1) 39 (22.7)10 (32.3) 
 T4 35 (14.5)60 (37–81)0.009* 4 (10.5) 24 (14.0) 7 (22.5)0.006*
 0 (no tumour) 49 (20.3)60 (33–80)  7 (18.4) 37 (21.5) 5 (16.1) 
 Low (I, II) 78 (32.4)57.5 (37–79) 16 (42.1) 55 (32.0) 7 (22.6) 
 High (III) 114 (47.3)62 (29–83)0.044*15 (39.5) 80 (46.5)19 (61.3)0.368
Metastases to lymph nodes       
 pN021560 (29–83) 37 (97.4)153 (89.0)25 (80.6) 
 pN1  1163 (57–72)  0  9 (5.2) 2 (6.5) 
 pN2 1563 (44–81)0.019* 1 (2.6) 10 (5.8) 4 (12.9)0.219
Figure 1.

Kaplan-Meier estimates of bladder CSS probabilities according to age at RC in 241 patients treated with RC and bilateral lymphadenectomy for bladder carcinoma.

Table 2. 
Multivariate Cox regression analysis for CSS
VariableHazard ratio (95% CI)P
Age, years (reference, 50)  
 51–692.154 (1.017–4.561)0.045
 ≥704.110 (1.694–9.969)0.002
Hydronephrosis (reference, no)  
 Yes3.308 (2.113–5.181)0.001
Nodal involvement (reference, no)  
 Yes1.921 (1.095–3.368)0.023


Improvements in technology and health care have contributed to an increase in the average life span of many populations. One of the natural results of this is the disproportionate increase in the incidence of bladder cancer in the elderly. Overall, 20–40% of patients with bladder cancer will present with or develop muscle-invasive disease. RC has become the reference standard for treating muscle-invasive bladder cancer [9]. Over the past 20 years, the morbidity and mortality rates for patients who have had RC have dramatically decreased. Coupled with improvements in lower urinary tract reconstruction, including continent urinary diversions and orthotopic neobladders, the morbidity experienced by patients who have had RC has dramatically improved, with benefits on quality-of-life [10]. Several investigators examined the morbidity of RC in the elderly [7,10–16]; in general, these reports showed that RC can be safe in properly selected elderly patients.

In this context we examined the association between patient age and pathological outcome, and bladder CSS in 241 patients treated at our centre. Advanced age was significantly associated with adverse pathological features and poorer disease-specific outcomes when examined as both categorical and continuous variables. Although some studies have not shown age to be an independent predictor of survival [17,18], in 1994, Thrasher et al.[11] found age >65 years to be an independent predictor of disease-specific mortality after RC. Similarly, a report by Takashi et al.[19] found, in a multivariate analysis of 264 patients, that age was an independent predictor of survival, but that stage, size of the tumour and the presence of irritative voiding symptoms were more important. In 2005, Clark et al.[1] analysed age as a categorical variable by decade, finding an increased risk of extravesical disease, disease recurrence, and poorer overall survival among the older patients. More recently, in a large series of RC, Nielsen et al.[20] found that higher chronological age was independently associated with pathologically advanced disease and bladder cancer-specific mortality after RC. The finding of an increased risk of extravesical disease and worse CSS among these elderly patients might possibly be due to a reluctance to treat these patients with major surgery, and this might cause a delay in undergoing surgery. Although studies on bladder cancer failed to show a linear relationship between delay and prognosis, most confirmed that delays are associated with a worse outcome [21]. Therefore, a delay due to a reluctance to treat elderly patients with major surgery might be an explanation for the disparity in survival. Another explanation might be a difference in the biology of tumours stratified by age. However, this is a controversial issue, and Clark et al.[1] found no difference in a molecular analysis of p53 among their patient age groups. Further work is needed on the association between age and tumour biology.

RC with urinary diversion is a safe and effective treatment for properly selected elderly patients with invasive bladder carcinoma who are in reasonably good health. However, higher age at RC is significantly associated with the risk of pathologically advanced disease and poorer CSS. To reduce the disparities in survival we recommend that clinicians strive to schedule their elderly patients efficiently and complete surgery as soon as possible. More prospective work is needed to examine the effect of age on tumour biology. Finally, more studies examining the effect of age on disease recurrence-free and CSS are warranted.


None declared.