Professional Documents
Culture Documents
Grreen Et Al. - Pre Cystectomy Prediction of NOC-UC
Grreen Et Al. - Pre Cystectomy Prediction of NOC-UC
Grreen Et Al. - Pre Cystectomy Prediction of NOC-UC
Study Type – Prognosis (case series) What’s known on the subject? and What does the study add?
Level of Evidence 4 Upstaging to non-organ-confined (NOC) disease is frequent at the time of radical
cystectomy for urothelial carcinoma of the bladder (UCB). Pre-surgical models that can
accurately predict which patients are likely to have more extensive disease are sparse.
OBJECTIVE
The present study developed an accurate nomogram for the prediction of NOC-UCB
based on a cohort of patients with clinically organ-confined disease. Adoption of such
• To create an accurate pre-cystectomy
a tool into daily clinical decision-making may lead to more appropriate integration of
decision-making tool that allows for the
perioperative chemotherapy, thereby potentially improving survival in patients with
accurate identification of patients with
UCB.
clinically organ-confined urothelial
carcinoma of the bladder (UCB) who have
non-organ-confined UCB (NOC-UCB) at
RESULTS highly accurate (area under the curve
cystectomy, as identification of patients
0.828) and well calibrated, deviating <8%
with UCB most likely to benefit from
• Clinical stage distribution was 19 from ideal prediction. Decision curve
neoadjuvant chemotherapy (NACTx) is
patients with Ta, 15 with Tis, 67 with T1, analysis showed net benefit across all
hampered by inaccurate clinical staging.
and 100 with T2. threshold probabilities.
• At the time of cystectomy, NOC-UCB and
PATIENTS AND METHODS LN-positive disease were found in 71 (35%) CONCLUSIONS
and 38 (19%) of patients, respectively;
• A prospectively maintained single- 81 (40%) of patients had NOC-UCB (≥pT3/ • NOC-UCB can be predicted with high
institution database containing 201 Nany or pTany/N+). accuracy by integrating standard
patients who underwent cystectomy and • Tumour stage (P [trend] <0.001), clinicopathological factors with imaging
pelvic lymph node (LN) dissection without presence of LVI (odds ratio [OR] 5.2; information.
NACTx for UCB was analysed. P = 0.02), and radiographic evidence • This model may help to identify patients
• Predictive variables for NOC-UCB of NOC-UCB or hydronephrosis (OR 3.2; with NOC-UCB who may benefit from
included, among others, age, gender, P = 0.01) were independently associated NACTx.
transurethral resection of bladder tumour with ≥pT3 Nany UCB.
(TURBT) findings (stage, grade, histology, • Tumour stage (P [trend] < 0.001) and KEYWORDS
size, presence of carcinoma in situ, presence of LVI (OR 6.64; P = 0.01) were
lymphovascular invasion [LVI], independently associated with (≥ pT3/Nany bladder cancer, urothelial carcinoma of
multifocality), history of intravesical or pTany/N+) UCB. bladder, radical cystectomy, non-organ-
therapy, time from TURBT to cystectomy, • A nomogram to predict (≥ pT3/Nany or confined (extravesical), lymph node,
and cross-sectional imaging findings. pTany/N+) based on all three variables was predictive model
© 2 0 1 2 B J U I N T E R N A T I O N A L | doi:10.1111/j.1464-410X.2012.11370.x 1
GREEN ET AL.
INTRODUCTION resulting in wider adoption of this therapy coefficients were used to develop a
by the urology community, potentially nomogram.
Radical cystectomy with bilateral pelvic improving survival. Therefore, our primary
lymph-node dissection (PLND) is the objective was to develop a preoperative Decision curve analysis [16] was used to
mainstay therapy for non-metastatic multivariable model, integrating clinical, explore the clinical value of the model
high-risk non-muscle-invasive and pathological, and imaging variables to predicting (≥ pT3/Nany or pTany/N+)
muscle-invasive urothelial carcinoma of accurately identify patients with clinically NOC-UCB. Because the value of a true
the bladder (UCB), resulting in durable organ-confined UCB who are at increased positive (i.e. detection of NOC-UCB and
oncological control and long-term survival risk for pathologically NOC-UCB. We subsequent treatment with NACTx) may
for most patients with organ-confined UCB. hypothesised that NOC-UCB could be differ from the disadvantages resulting from
However, 42–44% of patients without lymph predicted with reasonable accuracy. a false positive (i.e. potential over-treatment
node (LN) metastases but with disease with systemic chemotherapy), the net
extending to the perivesical fat and 65–67% benefit differentially weights true and false
of those with regional LN metastases have PATIENTS AND METHODS positives by using the threshold probability
disease recurrence ≤5 years after cystectomy at which one would opt for NACTx. For
[1,2]. Disease recurrence often manifests as After Institutional Review Board approval, example, if a provider (or patient) would opt
distant metastases [2], suggesting a high we retrospectively reviewed all prospectively for NACTx with a 30% risk of NOC-UCB but
prevalence of early systemic dissemination. collected data for 296 consecutive patients would forgo NACTx with only a 29% risk,
with bladder cancer who were treated with then the threshold probability is 30%.
Level I evidence supports the use of radical cystectomy or partial cystectomy and Clinically valuable models show a net benefit
neoadjuvant chemotherapy (NACTx) for bilateral PLND. We excluded patients who to alternative treatment strategies
clinical (c) T2–T4aN0M0 UCB [3]. However, had inadequate TUR of bladder tumour throughout the clinically applicable range of
to date, NACTx has been seldom used in the (TURBT) data (11 patients), patients with threshold probabilities. Alternative strategies
general community and may indeed be less non-urothelial pathology (16), those who include treating all patients with NACTx or
commonly used than adjuvant underwent NACTx (58), patients with clinical treating no one with NACTx. For the
chemotherapy. A recent multi-institutional T3–T4 stage disease (seven) based on TURBT purposes of this analysis, we assume that
study reported that only 12% of patients and bimanual examination, and those with the preoperative identification of NOC-UCB
with cT2–T4aN0M0 received NACTx, whereas a concomitant diagnosis of high-grade would lead to treatment with NACTx. All
22% received adjuvant systemic upper tract urothelial cancer (three). The statistical tests were two-sided with
chemotherapy [4]. The exact reasons for remaining 201 patients were the subject of significance set at P < 0.05. Statistical tests
these practice patterns remain unclear, but the present analysis; all had clinically were performed with SPSS® 18 (SPSS Inc.,
under-usage of NACTx may be due to the localised disease. IBM Corp., Somers, NY, USA) and R-statistics
belief that cystectomy alone will cure a (the R foundation for Statistical Computing,
significant proportion of patients who All surgical specimens were processed version 2.1.13).
would incur unnecessary and avoidable according to standard pathological
side-effects of overtreatment with NACTx. procedures at our institution and were
histologically confirmed to be UCB by a RESULTS
The mainstay of clinical staging for UCB, an dedicated genitourinary pathologist. If the
integration of pathological examination of TURBT procedure was performed at another The descriptive variables of the 201
the transurethral resection (TUR) specimen, institution, representative slides were evaluable patients are shown in Table 1.
bimanual physical examination, and obtained and reviewed before proceeding to The median (range) age at the time of
cross-sectional imaging studies, is highly cystectomy. Tumours were staged according cystectomy was 72.9 (41–92) years and 165
inaccurate [5]. It is not infrequent to find to the 2002 American Joint Committee on (82.1%) patients were male. Overall, 71
that the clinical stage is discordant with the Cancer/Union Internationale Contre le (35.4%) patients had (≥pT3/Nany) NOC-UCB
pathological stage at cystectomy, with Cancer TNM classification [13]. Tumour and 38 (18.9%) had LN metastases (N+). In
pathological upstaging to ≥ pT3 and/or N+ grading was assessed according to the 1998 all, 81 (40.3%) patients had (≥pT3/Nany or
disease occurring in up to 60% of patients WHO/International Society of Urological pTany/N+) NOC-UCB.
[6–10]. Current predictive tools, e.g. Pathology consensus classification [14].
nomograms, which are commonly used for lymphovascular invasion (LVI) was defined Tables 2 and 3 show the univariable and
outcome prediction in prostate cancer, do as the unequivocal presence of tumour cells multivariable regression models, respectively,
not substantially improve prediction of within an endothelium-lined space, with no for the prediction of NOC-UCB. In
non-organ-confined (NOC)-UCB in patients underlying muscular walls [15]. univariable analyses, TURBT tumour size,
with clinically organ-confined disease TURBT stage, TURBT LVI, intravesical therapy
[11,12]. This is in part due to the fact that To identify predictors of pathological stage history, and an abnormal preoperative
they rely only on clinical stage and grade. at the time of cystectomy, we created imaging study were associated with
multivariable logistic regression models to NOC-UCB in both models. In multivariable
More accurate clinical staging would allow predict (≥ pT3/Nany) and (≥ pT3/Nany or analyses TURBT stage, TURBT LVI, and
for the selection of patients most likely to pTany/N+). Discrimination was measured by abnormal preoperative imaging were
benefit from NACTx, thereby possibly area under the curve (AUC). Regression independently associated with (≥pT3/Nany)
TABLE 2 Univariable preoperative logistic regression analyses predicting NOC-UCB in patients treated with cystectomy and PLND
*P value for trend. †Abnormal imaging defined by presence of hydronephrosis and/or suggestion of NOC-UCB. IVT, intravesical therapy; CIS, carcinoma in situ.
FIG. 1. Pre-cystectomy nomogram predicting NOC-UCB stage (≥pT3/Nany or pTany/N+) at cystectomy and FIG. 2. Decision curve analysis of the effect of the
its calibration plot. Abnormal imaging defined by presence of hydronephrosis and/or suggestion of preoperative prediction model for detection of
NOC-UCB. (≥pT3/Nany or pTany/N+) in 151 patients who
underwent cystectomy. Assumption is made that
Prediction of (≥pT3 / Nany) or (pTany / N+) the identification of NOC-UCB would lead to
0 10 20 30 40 50 60 70 80 90100 treatment with NACTx. Net benefit is plotted
Points 1.0 AUC = 0.828 against threshold probabilities compared with
‘NACTx for all’ strategy and ‘NACTx for none’.
TURBT T1
0.8
Observed Probability
Stage Ta-Tis 1.0
T2 NACTx None
TURBT Yes NACTx All
0.6
LVI 0.8 NACTx per
No Nomogram
Abnormal Yes 0.4
Net benefit
Imaging 0.6
No ldeal
0.2 Longistic calibration
Total Points 0.4
Nonparametric
0
20
40
60
80
120
140
160
180
200
220
0
10
0.0
Probability of 0.0 0.2 0.4 0.6 0.8 1.0 0.2
≥pT3 / Nany or
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.86
0.9
0.93
0.95
0.97
TABLE 4 Number of patients with and without pathologically confirmed NOC-UCB (pT3–4/pNany or pTany/pN+) who would receive NACTx at various
threshold probabilities based on nomogram prediction; all patients with clinically organ-confined UCB
*Number over-treated = (number of patients predicted to have NOC-UCB) − (number of patients with pathologically confirmed NOC-UCB). % over-treated =
number over-treated/ number of patients predicted to have NOC-UCB. For a risk threshold of 30%, the number over-treated is 90 − 56 = 34, and %
over-treated is 34/90 × 100 = 38%. Assumption is that prediction of NOC-UCB drives the decision to use NACTx. †Column describes the ‘undertreated’ group;
those patients with pathologically NOC-UCB who were not treated with NACTx. The percentage is based upon a total of 65 patients with pathologically
confirmed NOC-UCB.
To examine the potential clinical impact that NACTx, or alternatively, treating no one suspicion that NOC-UCB is present before
the present predictive model provides, we (Fig. 2). This was true across almost the treating a patient with NACTx, showing a
performed a decision curve analysis, a entire range of threshold probabilities. very low risk-tolerance for what they may
technique that evaluates the clinical perceive as overtreatment of organ-confined
consequences of using predictive models A recent study reported that only 12% of UCB. Using the present predictive model
[16,26]. The use of the present model to patients with cT2–T4aN0M0 received NACTx (Fig. 1) in our cohort, a highly risk-intolerant
predict (≥ pT3/Nany or pTany/N+) NOC-UCB [4] showing that the threshold probability clinician (threshold probability 90%) would
for the purpose of guiding the use of for urologists to use NACTx may be very administer NACTx to 12 patients, all of
NACTx, provided a net benefit relative to the high. In other words, many urologists whom would in fact have pathologically
two strategies of treating all patients with probably require a very high degree of confirmed NOC-UCB (Table 4).
Lastly, the present nomogram predicts that perioperative chemotherapy, thereby international radical cystectomy cohort.
for patients with clinically ≤ T1 UCB and the potentially improving survival in patients BJU Int 2011; 107: 898–904
presence of TURBT LVI and/or abnormal with UCB. The clinical value of this model 8 Bianco FJ, Justa D, Grignon DJ, Sakr
imaging, a significant number of patients needs to be further assessed in external WA, Pontes JE, Wood DP. Management
will have NOC-UCB at cystectomy. Current multi-institutional validation cohorts. of clinical T1 bladder transitional cell
European Association of Urology guidelines carcinoma by radical cystectomy. Urol
recommend NACTx for all patients with Oncol 2004; 22: 290–4
cT2–T4N0M0 UCB [5]. However, there is no SOURCE OF FUNDING 9 Dutta SC, Smith JA, Shappell SB,
such recommendation for patients with Coffey CS, Chang SS, Cookson MS.
clinically non-muscle-invasive UCB. While Supported by Frederick J. and Theresa Dow Clinical under staging of high risk
treating all cT2–T4N0M0 patients with Wallace fund of the New York Community nonmuscle invasive urothelial carcinoma
NACTx may over-treat a significant number Trust. treated with radical cystectomy. J Urol
of patients, excluding all patients with < cT2 2001; 166: 490–3
UCB from NACTx without consideration 10 Cheng L, Neumann RM, Weaver AL
given to features such as LVI on TURBT or CONFLICT OF INTEREST et al. Grading and staging of bladder
abnormal preoperative imaging, may deny carcinoma in transurethral resection
some of these patients optimal therapy. None declared. specimens. Correlation with 105
matched cystectomy specimens. Am J
The present study is not without its Clin Pathol 2000; 113: 275–9
limitations including, but not limited to the REFERENCES 11 Karakiewicz PI, Shariat SF, Palapattu
retrospective analysis and the small cohort. GS et al. Precystectomy nomogram for
For example, there is inherent difficulty in 1 Stein JP, Lieskovsky G, Cote R et al. prediction of advanced bladder cancer
determining the presence of LVI at the Radical cystectomy in the treatment of stage. Eur Urol 2006; 50: 1254–62
morphological level with significant invasive bladder cancer: long-term 12 Shariat SF, Margulis V, Lotan Y,
differences between local pathologists and results in 1,054 patients. J Clin Oncol Montorsi F, Karakiewicz PI.
central pathology review [15]. Additionally, 2001; 19: 666–75 Nomograms for bladder cancer. Eur Urol
repeat TUR was not done for every patient, 2 Madersbacher S, Hochreiter W, 2008; 54: 41–53
but rather was performed at the discretion Burkhard F et al. Radical cystectomy 13 Edge SB, Byrd DR, Compton CC, Fritz
of the surgeon, which could have for bladder cancer today – a AG, Greene FL, Trotti A eds. American
contributed to under-staging. For example, homogeneous series without Joint Committee on Cancer (AJCC)
13 of 67 patients with T1 UCB at TURBT had neoadjuvant therapy. J Clin Oncol 2003; Staging Manual, Seventh Edn. New York:
high-grade recurrent and/or BCG refractory 21: 690–6 Springer, 2010
UCB without muscularis propria in the 3 Grossman HB, Natale RB, Tangen CM 14 Eble JN, Sauter G, Epstein JI,
specimen and were directed to cystectomy. et al. Neoadjuvant chemotherapy plus Sesterhenn IA. World Health
Moreover, some data points that have been cystectomy compared with cystectomy Organization Classification of Tumours.
analysed by other authors were not included alone for locally advanced bladder Pathology and Genetics of Tumours of
such as the frequency of variant UCB cancer. N Engl J Med 2003; 349: 859–66 the Urinary System and Male Genital
histology [27]. Lastly, imaging protocols 4 Feifer A, Taylor J, Shouery M et al. Organs. Lyon: IARCC Press, 2004
were not standardised, nor was the use of B.C.A.N. Muscle-Invasive Bladder Cancer 15 Shariat SF, Svatek RS, Tilki D et al.
i.v. contrast for CT or MRI. Quality of Care Consortium. Multi- International validation of the
institutional quality-of-care initiative for prognostic value of lymphovascular
In conclusion, the current conventional non-metastatic, muscle-invasive, invasion in patients treated with radical
clinical staging for patients with UCB before transitional cell carcinoma of the cystectomy. BJU Int 2010; 105: 1402–12
cystectomy results in a high rate of staging bladder. J Clin Oncol 2011; 29 (Suppl. 7): 16 Vickers AJ, Elkin EB. Decision curve
inaccuracy; therefore more accurate tools 240 [abstract] analysis: a novel method for evaluating
are needed, particularly for the purpose of 5 Stenzl A, Cowan NC, De Santis M prediction models. Med Decis Making
efficaciously directing patients towards et al. Treatment of muscle-invasive and 2006; 26: 565–74
NACTx. The present study provides such a metastatic bladder cancer: update of the 17 Sonpavde G, Shariat SF. Preoperative
tool and reiterates that LVI found at TURBT EAU guidelines. Eur Urol 2011; 59: chemotherapy for bladder cancer: a
and abnormal preoperative imaging findings 1009–18 standard waits to be optimally deployed.
can predict NOC-UCB at the time of 6 Shariat SF, Palapattu GS, Karakiewicz Cancer 2012; 118: 8–11
cystectomy in patients with presumed PI et al. Discrepancy between clinical 18 Stimson CJ, Cookson MS, Barocas DA
clinically localised UCB. Decision curve and pathologic stage: impact on et al. Preoperative hydronephrosis
analysis of the nomogram shows clinical prognosis after radical cystectomy. Eur predicts extravesical and node positive
benefit both for clinicians hesitant to use Urol 2007; 51: 137–51 disease in patients undergoing
NACTx and for those who liberally 7 Svatek RS, Shariat SF, Novara G et al. cystectomy for bladder cancer. J Urol
recommend for its use. Adoption of such a Discrepancy between clinical and 2010; 183: 1732–7
tool into daily clinical decision-making may pathological stage: external validation of 19 Favaretto RL, Shariat SF, Savage C
lead to more appropriate integration of the impact on prognosis in an et al. Combining imaging and
ureteroscopy variables in a preoperative cystectomy. J Clin Oncol 2005; 23: do more good than harm? J Clin Oncol
multivariable model for prediction of 6533–9 2011; 29: 2951–2
muscle-invasive and non-organ confined 23 Hong SK, Kwak C, Jeon HG, Lee E, Lee 27 Black PC, Brown GA, Dinney CP. The
disease in patients with upper tract SE. Do vascular, lymphatic, and impact of variant histology on the
urothelial carcinoma. BJU Int 2012; 109: perineural invasion have prognostic outcome of bladder cancer treated with
77–82 implications for bladder cancer after curative intent. Urol Oncol 2009; 27:
20 Quek ML, Stein JP, Nichols PW et al. radical cystectomy? Urology 2005; 65: 3–7
Prognostic significance of 697–702
lymphovascular invasion of bladder 24 Leissner J, Koeppen C, Wolf HK. Correspondence: Shahrokh F. Shariat, Brady
cancer treated with radical cystectomy. Prognostic significance of vascular and Urologic Health Center, Weill Cornell Medical
J Urol 2005; 174: 103–6 perineural invasion in urothelial bladder College, New York Presbyterian Hospital, 525
21 Canter D, Guzzo T, Resnick M et al. cancer treated with radical cystectomy. East 68th Street, Box 94, Starr 900, New
The presence of lymphovascular invasion J Urol 2003; 169: 955–60 York, NY 10065, USA.
in radical cystectomy specimens from 25 Resnick MJ, Bergey M, Magerfleisch L, e-mail: sfshariat@gmail.com
patients with urothelial carcinoma Tomaszewski JE, Malkowicz SB, Guzzo
portends a poor clinical prognosis. BJU TJ. Longitudinal evaluation of the Abbreviations: AUC, area under the curve;
Int 2008; 102: 952–7 concordance and prognostic value LN, lymph node; LVI, lymphovascular
22 Lotan Y, Gupta A, Shariat SF et al. of lymphovascular invasion in invasion; NACTx, neoadjuvant
Lymphovascular invasion is transurethral resection and radical chemotherapy; NOC, non-organ-confined;
independently associated with overall cystectomy specimens. BJU Int 2011; PLND, pelvic lymph-node dissection;
survival, cause-specific survival, and 107: 46–52 TUR(BT), transurethral resection (of bladder
local and distant recurrence in patients 26 Vickers AJ. Prediction models: tumour); UCB, urothelial carcinoma of the
with negative lymph nodes at radical revolutionary in principle, but do they bladder.