SEARCH

SEARCH BY CITATION

Keywords:

  • comorbidity;
  • cystectomy;
  • ACE-27;
  • ASA;
  • urothelial carcinoma;
  • perioperative mortality

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Study Type – Prognosis (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

The degree of comorbidity significantly affects the course of patients with bladder cancer undergoing radical cystectomy (RC).

To our knowledge this is the first study comparing four different comorbidity indices in patients undergoing RC for urothelial carcinoma to assess the best clinical predictors for 90-day perioperative mortality. We concluded that the ASA score should be the method of choice, as it showed a predictive ability superior to that of ECOG and CCI, and is much easier to generate than the ACE-27.

OBJECTIVE

  • • 
    To evaluate which of the following among the Adult Comorbidity Evaluation-27 (ACE-27), the Charlson Comorbidity Index (CCI), the Eastern Cooperative Oncology Group performance status (ECOG) and the American Society of Anesthesiologists (ASA) comorbidity scores correlate best with perioperative mortality after radical cystectomy (RC) for urothelial carcinoma (UC) of the bladder.

PATIENTS AND METHODS

  • • 
    A study was carried out on 555 unselected consecutive patients without neoadjuvant chemotherapy who underwent RC for UC of the bladder from 2000 to 2010 at one of two institutions.
  • • 
    Patients' medical records were reviewed retrospectively.
  • • 
    We established a defined binary linear progression model based on clinical variables to predict perioperative mortality <90 days after RC (90PM). To this model we added, individually, the comorbidity indices ACE-27, CCI, ECOG, and ASA to assess their predictive capacity regarding 90PM.

RESULTS

  • • 
    The overall 90PM was 7.9%.
  • • 
    Age (P= 0.01) and clinical distant metastatic tumour stage (P= 0.002) were independent predictors for 90PM in the multivariate analysis.
  • • 
    Each of the four investigated comorbidity indices was able to significantly increase the predictive capacity of the basic model: ECOG +13.5%, (odds ratio [OR]: 1.61, P= 0.036; area under the curve [AUC] 74.7), ASA Score +28.3% (OR: 2.19, P= 0.004; AUC 76.1), Charlson Index +12.3% (OR: 1.31, P= 0.047; AUC 73.8) and ACE-27 + 29.8% (OR: 1.72, P= 0.004; AUC 76.1).

CONCLUSIONS

  • • 
    ASA and ACE-27 show a nearly identical clinical predictive value for perioperative mortality. Both scores could be considered for clinical practice.
  • • 
    With regard to ease of generation and availability, the ASA score can be regarded as the best instrument.

Abbreviations
ACE-27

Adult comorbidity Evaluation 27

CCI

Charlson Comorbidity Index

ECOG

Eastern Cooperative Oncology Group

ASA

American Society of Anesthesiologists

RC

radical cystectomy

UC

urothelial carcinoma

90PM

perioperative mortality <90 days after RC

OR

odds ratio

AUC

area under the curve

BMI

body mass index

BM

basic model

ROC

receiver-operating characteristic

IQR

interquartile range

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Radical cystectomy (RC) is the ‘gold standard’ therapy for urothelial carcinoma (UC) invading the bladder muscle and for high-risk non-muscle invasive UC of the bladder [1–3]. Despite medical improvements, RC is still associated with significant perioperative mortality <90 days after RC (90PM), which ranges from 2.6% to 7.3% in the current literature [4–11]. Disease stage, but also age, gender, body mass index (BMI) and hospital surgery volume have been reported as important determinants of mortality after RC [12]. In particular, the age of patients undergoing RC has increased, resulting in an increased risk of perioperative complications and mortality. Different studies have reported the impact of comorbidities on prognosis and patient mortality after surgery and particularly after RC [13–15]. The most commonly used comorbidity indices in the literature are the Adult Comorbidity Evaluation-27 (ACE-27), the Charlson Comorbidity Index (CCI) and the American society of Anaesthesiologists (ASA) score. ACE-27 and CCI scores were significantly associated with overall survival in previous studies on mortality after RC [14–16]. Both scores were significantly associated with 90-day readmission and mortality [8,17,18]. No report has yet established whether the ASA score and the Eastern Cooperative Oncology Group (ECOG) performance status could also be useful in predicting the 90PM. In the present study, we analysed the ability of the aforementioned comorbidity indices in predicting 90PM in patients treated with RC.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

STUDY DESIGN AND POPULATION

This was an institutional review board-approved study, retrospectively analysing patients' medical records. The study population comprised 555 patients diagnosed with UC of the bladder who underwent RC between January 2000 and April 2010 at the General Hospital of Bolzano in Italy and the University of Regensburg in Germany. The indication for RC was a high-risk non-muscle-invasive or muscle-invasive tumour stage, based on histopathological specimens obtained by transurethral resection. The patients were staged using CT/MRI before surgery. RC was performed by several different surgeons. Standard lymphadenectomy comprised dissection of all lymphatic tissue up to and including the iliac bifurcation. Patients who had undergone neoadjuvant chemotherapy (n= 21) or showing histology other than UC were excluded from the study.

PATHOLOGICAL EVALUATION

All tumours were histologically confirmed UC, including histological variants of UC. Tumour staging was performed according to the American Joint Committee on Cancer-Union Internationale Contre le Cancer TNM classification [19].

COMORBIDITY INDICES

Four comorbidity indices were assessed: ASA, ACE-27, CCI and ECOG. The ASA score was obtained prospectively because it was the standard assessment criterion of comorbidity in both clinics involved in the present study. ACE-27, CCI and ECOG were obtained retrospectively based on the information from patients' documents.

The ACE-27 is a 27-item comorbidity instrument validated on adult oncology patients including 513 patients with bladder cancer [20]. Specific diseases and conditions are classified into three grades according to their severity. Once all organ systems have been classified, an overall comorbidity score is assigned based on the highest ranked single condition. If two or more moderate conditions occur in different organ systems or disease groupings, the overall score is classified as severe. The ASA score was described in 1940 and is the oldest evaluation score assessing perioperative risk [21]. The CCI was initially described in 1987 by Charlson et al. [22] to predict 1-year mortality in medical patients. The age-adjusted CCI, which is calculated with additional points added for age, was not addressed in the present study because the variable ‘age’ was already part of the basic model (BM), resulting in collinearity (phi-value > 0.55). The ECOG score was published in 1982 [23]. It is a classification of the wellbeing of cancer patients and runs from 0 to 5, with ECOG 0 describing perfect health and ECOG 5 equating to death.

STATISTICAL ANALYSES

Patients' clinical characteristics were assessed. Clinical variables were tested for collinearity. Absence of factor collinearity was defined as a phi-value below 0.55 (P < 0.001). No variable had to be excluded from further analyses. A binary linear regression model (odds ratio [OR], 95% CI) addressing 90PM was created only including these preoperatively available clinical variables (the BM [Table 1]). In addition, the comorbidity indices were tested for collinearity. Based on a Spearman-correlation with phi-values >0.5, no combined inclusion of the indices into the prediction models was performed. Every comorbidity score was, therefore, separately added to the BM and different binary linear regression models were calculated. The statistical quality of the different multivariate models was assessed using the Omnibus test comparing the model coefficients. Briefly, compared with the BM, the changing deviance of step 1 of the backward elimination (−2 multiplied log-likelihood) was analysed by means of the development of likelihood ratio chi-square (percentage change of model quality including the affiliated P values in comparison with the other multivariate Cox models). Predictive ability was assessed using the area under the receiver-operating characteristic (ROC) curve. All reported P values were two-sided and a P value of ≤0.05 was considered to indicate statistical significance. All statistical tests were performed using SPSS version 17.0 (IBM Corp., Somers, NY, USA).

Table 1.  Association of clinical features with 90PM: defined binary logistic regression model (BM)
 OR95% CI P
Institution0.890.43–1.830.752
Sex0.740.30–1.760.493
BMI1.020.95–1.090.626
cTN Stage1.940.89–4.200.094
cM + Stage3.871.39–10.790.010
Age1.061.02–1.100.002
Constant0.001<0.001

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

STUDY POPULATION

The characteristics of the 555 patients are listed in Table 2. Of these patients, 450 (81%) were male and 105 (19%) female, and all of them were Caucasian. The median (interquartile range [IQR]) age was 70 (63–76) years. A median (IQR) of 13 (9–19) lymph nodes were removed during RC, and 155 (27.9%) patients had positive lymph nodes. The comorbidity distribution of the patient cohort can be seen in Table 2. The overall 90PM was 7.9%. Adjuvant chemotherapy regimes were administered to 108 (19.5%) patients.

Table 2.  Characteristics of 555 patients treated with RC for UC of the bladder stratified by institution
 Total cohortRegensburgBolzano P
No. of patients555354201 
Median (IQR) age70 (63–76)69 (64–77)72 (63–78)0.000
 <50 years, n (%)19 (3.4)17 (4.8)2 (1.0) 
 50–59 years, n (%)77 (13.9)54 (15.3)23 (11.4) 
 60–69 years, n (%)160 (28.8)118 (33.3)42 (20.9) 
 70–79 years, n (%)222 (40.0)127 (35.9)95 (47.3) 
 >79 years, n (%)77 (13.9)38 (10.7)39 (19.4) 
Gender, n (%)
 Male450 (81.1)289 (81.6)161 (80.1)0.656
 Female105 (18.9)65 (18.4)40 (19.9) 
BMI classification, n (%)   <0.000
 <18.5 (underweight)14 (2.6)8 (2.3)6 (3.0) 
 18.5–25 (normal weight)213 (39.2)115 (33.5)98 (48.8) 
 25–30 (obese I)227 (41.7)146 (42.6)81 (40.3) 
 30–35 (obese II)67 (12.3)52 (15.2)15 (7.5) 
 >35 (obese III)23 (4.2)22 (6.4)1 (0.5) 
ASA score, n (%)   0.016
 189 (16.4)50 (14.7)39 (19.4) 
 2265 (48.9)156 (45.7)109 (54.2) 
 3186 (34.3)134 (39.3)52 (25.9) 
 42 (0.4)1 (0.3)1 (0.5) 
ACE-27 classification, n (%)   <0.000
 None112 (20.7)71 (20.8)41 (20.4) 
 Mild119 (22.0)55 (16.1)64 (31.8) 
 Moderate191 (35.2)127 (37.2)64 (31.8) 
 Severe120 (22.1)88 (25.8)32 (15.9) 
CCI score, n (%)   <0.000
 0–1152 (27.9)75 (21.6)77 (38.9) 
 2–3320 (58.7)215 (62.0)105 (53.0) 
 4–567 (12.3)52 (15.0)15 (7.6) 
 6–76 (1.1)5 (1.4)1 (0.5) 
ECOG score, n (%)   <0.000
 0123 (22.7)75 (21.9)58 (28.8) 
 1290 (53.4)149 (43.6)90 (44.8) 
 2112 (20.6)108 (31.6)43 (21.4) 
 318 (3.3)10 (2.9)10 (5.0) 
 40 (0.0)0 (0)0 (0) 
Tumour stage cM1, n (%)32 (5.8)24 (6.8)8 (4.0)0.174
90PM, n (%)44 (7.9)29 (8.2)15 (7.5)0.760

Comparing both participating institutes, the Regensburg University cohort included significantly younger patients (P < 0.000), with a higher BMI (P < 0.000) and higher comorbidity (P= 0.016 for ASA and P < 0.000 for ACE27, CCI and ECOG, respectively). 90PM and cM-stage did not differ significantly (Table 2).

DEFINED BM

Age, gender, BMI, institution and clinical TNM were tested for correlations and no variable had to be excluded. A binary logistic regression model addressing 90PM was calculated from these variables (Table 1). Age (OR 1.06, 95% CI 1.02–1.10, P= 0.01) and clinical distant metastatic tumour stage (cM+) (OR 3.87, 95% CI 1.39–10.79, P= 0.002) were independent predictors for 90PM. BMI, gender, institution and clinical lymph node status were insignificant. For the BM a chi-squared index (likelihood ratio chi-squared) of 30.69 and an AUC of 71.6 (95% CI 0.631–0.800) were calculated (Table 3, Fig. 1).

Table 3.  Comparison of the different models consisting of the BM plus one comorbidity index by OR of the comorbidity index within the model, including its P value, chi-squared index of the models and the AUC from the ROC curve
ModelOR of comorbidity score (95% CI) P Chi-squared index of model, ratio %AUC of model (95% CI) P value of model
BM30.6971.5 (63–80)
BM + ECOG1.61 (1.03–2.52)0.03634.85 (+13.6)74.7 (67–82)<0.001
BM + ASA2.19 (1.28–3.76)0.00439.38 (+28.3)76.1 (70–83)<0.001
BM + CCI1.30 (1.00–1.70)0.04734.45 (+12.2)73.8 (66–82)<0.001
BM + ACE-271.71 (1.19–2.46)0.00439.86 (+29.8)76.1 (70–83)<0.001
image

Figure 1. Diagnostic performance analysis (ROC curve) of the BM and the BM + comorbidity indices for the prediction of 90PM.

Download figure to PowerPoint

PREDICTIVE ABILITY OF THE COMORBIDITY INDICES

Each comorbidity score was added to the BM separately. The odds ratios of the comorbidity indices and chi-squared index as well as the AUC from the ROC-curve were calculated and compared. Age and clinical status of distant metastatic disease remained significant variables in any regression model.

ASA

The ASA score turned out to be a significant variable (P= 0.004, OR 2.19, 95% CI 1.28–3.76) when added to the BM (Table 3). The predictive capacity of the model was significantly improved, with a chi-squared index of 39.38 (+28.3%) and an AUC of 76.1 (95% CI 70–83 [Table 3, Fig. 1). In detail, patients aged ≥ 80 years, with an ASA score ≥ 3, show an 8-times higher risk of dying within 90 days after RC than patients <80 years with ASA ≤ 2 (P < 0.00).

ECOG

Within the regression model, ECOG score was shown to be a highly significant variable (P= 0.036, OR 1.61, 95% CI 1.03–2.52) as a predictor for 90PM (Table 3). The predictive capacity of the model was significantly improved. A chi-squared index of 34.85 (+13.5%) and an AUC of 74.7 (95% CI 67–82) were calculated (Table 3, Fig. 1).

CCI

The CCI was also a significant variable when added to the BM (P= 0.047, OR 1.30; 95% CI 1.00–1.70 [Table 3]). The predictive capacity was 34.45 (+12.2%) and the AUC was 73.8 (95% CI 66–82 [Table 3, Fig. 1]).

ACE-27

The ACE-27, a comorbidity score that was created for patients with cancer, was also a significant predictor for 90PM (P= 0.004, OR 1.71, 95% CI 1.19–2.47 [Table 3]). The chi-squared index was increased to 39.86 (+28.9%) and the AUC was 76.1 (95% CI: 70–83 [Table 3, Fig. 1]).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

We assessed four validated comorbidity scores to abstract and classify comorbidity data. To our knowledge, the present study is the first to compare four different comorbidity indices in patients undergoing RC for UC, assessing which of these indices is the best clinical predictor of 90PM. Restricting the BM to clinical variables allowed valid counselling of the patients before RC.

Three main findings emerged from the present study. The first finding was that age and clinical metastatic tumour stage are independently associated with 90PM after RC. This finding is consistent with the current literature. A recent retrospective study conducted by May et al. [24], involving 2483 patients, showed that 90PM is significantly higher in ≥75-year-old patients undergoing RC than in the younger group (6.2% vs. 3.7%, P= 0.026). Liberman et al. [25] showed that in septuagenarian and octogenarian patients, the overall 90PM rate was 5.4% and 9.2%, respectively, compared with 2.0% for their younger counterparts (OR 2.9 and 5.0, respectively; P > 0.001). In a nomogram to predict 90PM in patients with UC after RC, it was shown that age was the strongest indicator of 90PM [5]. In addition, in the same nomogram, advanced tumour stage was associated with increased 90PM, which is consistent with our findings where age and clinical metastatic tumour stage were significantly associated with 90PM. These known facts allow us to underline the high risk of palliative RC in old and comorbid patients with a clinically distant metastatic tumour stage.

The second finding was that comorbidity, assessed by validated comorbidity indices (ACE-27, ECOG, CCI and ASA), is an independent predictor of 90PM. ACE-27 was shown to improve the basic regression model by +29.8% (OR 1.72). These results are consistent with those of a study published by Fairey et al. [8], who reported on 314 patients with severe comorbidity conditions undergoing RC, assessed by ACE-27, and found that these were associated with an increased risk of 90PM (OR 6.4, 95% CI 1.1–66.4, P= 0.03). The ASA score is the second most appropriate tool, improving the predictive capacity of the BM by ≈+28%. These findings are consistent with the data reported by Boström et al. [9], who reported on a cohort of 258 patients where an ASA score ≥ 3 was independently associated with 90PM (OR 3.25, P= 0.036). In addition, Malavaud et al. [26] showed that an ASA score ≥ 3 was associated with major complications and most specifically those related to the type of urinary diversion. The CCI improves the prediction of 90PM of our BM by 12.3%. In a study population of 1121 patients, Koppie et al. showed that overall survival, but not progression-free survival, is significantly associated with the age-adjusted CCI [16]. In a study by Lund et al. [27] involving 3997 patients with invasive UC categorized using the CCI, 5-year mortality rates were 45%, 30% and 20% with a CCI score = 0, CCI score = 1–2 and CCI score > 3, respectively. There have been no studies on the predictive ability of the CCI for 90PM after RC. The ECOG score, which is based on the physical status of the patient, improved the BM by +13.5%, an improvement that was highly significant. To our knowledge the ECOG has not previously been assessed with regard to patients undergoing RC for UC.

Based on their huge predictive capacity, comorbidity scores are essential in the preoperative risk assessment of patients undergoing RC. A recently published nomogram predicting 90PM in patients with UC after RC does not consider this fact [5]. The ACE-27 and ASA scores can be recommended for preoperative assessment of 90PM, because they are superior to the CCI and the ECOG scores; however, our third finding was that, from a practical standpoint, the ASA score should be highlighted as the preferred instrument, because it is much easier to generate than the ACE-27. The ASA score stratifies the patients into low risk or high risk of 90PM groups, because most patients fall into ASA categories 2 or 3.

Whilst we are aware that currently there is no equal alternative treatment to RC for muscle-invasive tumour stages, an improved assessment of the risk of 90PM could help clinicians hold better-informed discussions with their patients about the risks and benefits of RC.

The present study has certain limitations that must be considered when interpreting the findings. The study is retrospective in nature and is therefore limited in terms of data acquisition. Perioperative standards were not equally defined for the two institutions within the study period. Equally, it should be noted that the rate of 90PM (7.9%) was in the upper range of the reported percentage in the current literature. In several contemporary RC series, 90PM rates ranged between 2.6% and 7.3% [6–9,11,17,25]. By contrast, the 30-day mortality is 2.8%, which can be considered to be within the established range. An explanation is the older age (54% of patients > 70 years) and higher comorbidity rate of the present study population (57% of the patients had an ACE-27 ≥ 2) compared with those in the current literature. Both involved centres are reference centres, resulting in negatively selected patients regarding tumour stage, age and comorbidity. Despite the conspicuous differences of the populations of the included centres (age, BMI and comorbidities), the 90PM did not differ significantly, suggesting that younger age balances the higher comorbidity of the patients. Furthermore, in the present study the cause of death remains unspecified. Thus, postoperative clinical pathways are not influenced by our results. And finally, it should be noted that all existing comorbidity scores only reflect a certain proportion of the spectrum of comorbidities that could possibly result in mortality, meaning that a substantial number of key drivers of death are unknown or not incorporated.

In conclusion, in the present series of 555 patients undergoing RC for UC, age and clinical metastatic tumour staging are independent predictors of 90PM. All four evaluated comorbidity indices (ACE-27, CCI, ASA and ECOG) are independent predictors for 90PM. ASA and ACE-27 are superior to CCI and ECOG, and therefore both could be considered for clinical practice; however, the ASA score is most widely available and simple to generate.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

This article is dedicated to Karl Fritsche. We thank Jenny Wagner and Patrick Goldsworthy for their excellent assistance.

Hans Martin Fritsche and Matthias May had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

CONFLICT OF INTEREST

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Financial disclosures: the authors certify that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (e.g. employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: none.

Funding/Support and role of the sponsor: none.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES
  • 1
    Stein JP, Lieskovsky G, Cote R et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol 2001; 19: 66675
  • 2
    Herr HW, Dotan Z, Donat SM, Bajorin DF. Defining optimal therapy for muscle invasive bladder cancer. J Urol 2007; 177: 43743
  • 3
    Stenzl A, Cowan NC, De Santis M et al. Treatment of muscle-invasive and metastatic bladder cancer: update of the EAU guidelines. Eur Urol 2011; 59: 100918
  • 4
    Ramani VA, Bromage SJ, Clarke NW. A contemporary standard for morbidity and outcome after radical cystectomy. BJU Int 2009; 104: 62832
  • 5
    Isbarn H, Jeldres C, Zini L et al. A population based assessment of perioperative mortality after cystectomy for bladder cancer. J Urol 2009; 182: 707
  • 6
    Donat SM, Shabsigh A, Savage C et al. Potential impact of postoperative early complications on the timing of adjuvant chemotherapy in patients undergoing radical cystectomy: a high-volume tertiary cancer center experience. Eur Urol 2009; 55: 17785
  • 7
    Hollenbeck BK, Miller DC, Taub DA et al. The effects of adjusting for case mix on mortality and length of stay following radical cystectomy. J Urol 2006; 176: 13638
  • 8
    Fairey A, Chetner M, Metcalfe J et al. Associations among age, comorbidity and clinical outcomes after radical cystectomy: results from the Alberta Urology Institute radical cystectomy database. J Urol 2008; 180: 12834; discussion 34
  • 9
    Boström PJ, Kossi J, Laato M, Nurmi M. Risk factors for mortality and morbidity related to radical cystectomy. BJU Int 2009; 103: 1916
  • 10
    Porter MP, Gore JL, Wright JL. Hospital volume and 90-day mortality risk after radical cystectomy: a population-based cohort study. World J Urol 2011; 29: 737
  • 11
    Gregg JR, Cookson MS, Phillips S et al. Effect of preoperative nutritional deficiency on mortality after radical cystectomy for bladder cancer. J Urol 2011; 185: 906
  • 12
    Goossens-Laan CA, Gooiker GA, van Gijn W et al. A systematic review and meta-analysis of the relationship between hospital/surgeon volume and outcome for radical cystectomy: an update for the ongoing debate. Eur Urol 2011; 59: 77583
  • 13
    Hines RB, Chatla C, Bumpers HL et al. Predictive capacity of three comorbidity indices in estimating mortality after surgery for colon cancer. J Clin Oncol 2009; 27: 433945
  • 14
    Megwalu II, Vlahiotis A, Radwan M, Piccirillo JF, Kibel AS. Prognostic impact of comorbidity in patients with bladder cancer. Eur Urol 2008; 53: 5819
  • 15
    Fairey AS, Jacobsen NE, Chetner MP et al. Associations between comorbidity, and overall survival and bladder cancer specific survival after radical cystectomy: results from the Alberta Urology Institute Radical Cystectomy database. J Urol 2009; 182: 8592; discussion 3
  • 16
    Koppie TM, Serio AM, Vickers AJ et al. Age-adjusted Charlson comorbidity score is associated with treatment decisions and clinical outcomes for patients undergoing radical cystectomy for bladder cancer. Cancer 2008; 112: 238492
  • 17
    Stimson CJ, Chang SS, Barocas DA et al. Early and late perioperative outcomes following radical cystectomy: 90-day readmissions, morbidity and mortality in a contemporary series. J Urol 2010; 184: 1296300
  • 18
    Taylor JM, Feifer A, Savage CJ et al. Evaluating the utility of a preoperative nomogram for predicting 90-day mortality following radical cystectomy for bladder cancer. BJU Int 2011; doi:10.1111/j.1464-410X.2011.10391.x. [Epub ahead of print]
  • 19
    Sobin LH, Gospodarowicz MK, Wittekind C, International Union against Cancer. TNM Classification of Malignant Tumours. West Sussex: UK Wiley-Blackwell, 2010
  • 20
    Piccirillo JF, Tierney RM, Costas I, Grove L, Spitznagel EL Jr. Prognostic importance of comorbidity in a hospital-based cancer registry. JAMA 2004; 291: 24417
  • 21
    Saklad M. Grading of Patients for Surgical Procedures. Anesthesiology 1941; 2: 2814
  • 22
    Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 37383
  • 23
    Oken MM, Creech RH, Tormey DC et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol 1982; 5: 64955
  • 24
    May M, Fritsche HM, Gilfrich C et al. [Influence of older age on survival after radical cystectomy due to urothelial carcinoma of the bladder: survival analysis of a German multi-centre study after curative treatment of urothelial carcinoma of the bladder.]. Urologe A 2011; 50: 8219
  • 25
    Liberman D, Lughezzani G, Sun M et al. Perioperative mortality is significantly greater in septuagenarian and octogenarian patients treated with radical cystectomy for urothelial carcinoma of the bladder. Urology 2011; 77: 6606
  • 26
    Malavaud B, Vaessen C, Mouzin M, Rischmann P, Sarramon J, Schulman C. Complications for radical cystectomy. Impact of the American Society of Anesthesiologists score. Eur Urol 2001; 39: 7984
  • 27
    Lund L, Jacobsen J, Clark P, Borre M, Norgaard M. Impact of comorbidity on survival of invasive bladder cancer patients, 1996–2007: a Danish population-based cohort study. Urology 2010; 75: 3938