- Email: [email protected]
Dynamic Prognostication Using Conditional Recurrence and Progression Estimates for Patients with Nonmuscle Invasive Bladder Cancer
Author's Accepted Manuscript Dynamic prognostication using conditional recurrence and progression estimates for patients with non-muscle invasive bladder cancer Carmen V. Leitner , Ines A. Ederer , Michela de Martino , Sebastian L. Hofbauer , Ilaria Lucca , Aurélie Mbeutcha , Romain Mathieu , Andrea Haitel , Martin Susani , Shahrokh F. Shariat , Tobias Klatte PII: DOI: Reference:
S0022-5347(16)00225-1 10.1016/j.juro.2016.01.102 JURO 13295
To appear in: The Journal of Urology Accepted Date: 20 January 2016 Please cite this article as: Leitner CV, Ederer IA, de Martino M, Hofbauer SL, Lucca I, Mbeutcha A, Mathieu R, Haitel A, Susani M, Shariat SF, Klatte T, Dynamic prognostication using conditional recurrence and progression estimates for patients with non-muscle invasive bladder cancer, The Journal of Urology® (2016), doi: 10.1016/j.juro.2016.01.102. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain.
Embargo Policy All article content is under embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date. Questions regarding embargo should be directed to [email protected].
ACCEPTED MANUSCRIPT
Dynamic prognostication using conditional recurrence and progression estimates for patients with non-muscle invasive bladder cancer
RI PT
Carmen V. Leitner1, Ines A. Ederer1, Michela de Martino1, Sebastian L. Hofbauer1, Ilaria Lucca1, Aurélie Mbeutcha1, Romain Mathieu1, Andrea Haitel2, Martin Susani2, Shahrokh F. Shariat1,3,4, Tobias Klatte1
1
M AN U
SC
Department of Urology, Medical University of Vienna, Vienna General Hospital, Vienna, Austria, 2 Clinical Institute of Pathology, Medical University of Vienna, Vienna General Hospital, Vienna, Austria, 3Department of Urology, University of Texas Southwestern Medical Center, Dallas, TX, 4 Department of Urology, Weill Cornell Medical College, New York-Presbyterian Hospital, New York, NY
Word count: 2498 Tables: 5
*Corresponding author: Tobias Klatte Department of Urology
TE D
Figures: 2
EP
Medical University of Vienna Währinger Gürtel 18-20
AC C
A-1090 Vienna, Austria
Phone: +43-1-40400-26020 Fax: +43-1-40400-23320
Email: [email protected]
Keywords: recurrence; progression; BCG; mitomycin; risk group
1
ACCEPTED MANUSCRIPT
ABSTRACT Purpose: Conditional estimates provide a dynamic prediction of outcomes, but there are no data
RI PT
for non-muscle invasive bladder cancer (NMIBC). The purpose of this study was to assess the changes in conditional recurrence and progression rates after transurethral resection of the bladder (TURB) and to explore the prognostic impact of established factors and risk groups over
SC
time.
Methods: We retrospectively analyzed data from 1292 consecutive patients with a newly
recurrence and time to progression.
M AN U
diagnosed Ta/T1 BC who underwent a TURB. The endpoints of this study were time to first
Results: The 2-year recurrence rate at baseline was 36% and improved as a function of time that the patient had been free of disease recurrence. After 6-, 12-, 24-, 36-, and 48-months, the
TE D
2-year conditional recurrence rate improved to 31% (14% improvement compared with baseline), 22% (39%), 16% (56%), 13% (64%), and 11% (69%), respectively. Comparably, conditional progression rates improved with increasing follow-up, although relative differences were less distinct. The prognostic impact of established factors and the NMIBC risk group
EP
progressively decreased over time and finally disappeared. BCG, however, had a protective
AC C
effect on progression even after 3 years. We also provide tables with dynamic prognostic information at all analyzed time points. Conclusions: In patients with primary Ta/T1 BC, recurrence and progression rates improve over time. The prognostic impact of established factors and risk groups decreases and finally disappears. The effect of BCG on progression is long-lasting. Conditional outcome estimates may improve patient counseling and individualize surveillance planning.
2
ACCEPTED MANUSCRIPT
INTRODUCTION Bladder cancer (BC) is the second most common urologic malignancy with more than 400,000 new cases worldwide yearly.1 It is expected to remain a significant burden for healthcare
RI PT
systems in the next decades because of the aging and continuous growth of the world population as well as the high prevalence of smoking.2 As a measure for the needs of medical care, the 5-year global BC prevalence has been estimated to be more than 1.1 million.3 At initial
SC
diagnosis, approximately 75% of patients present with a non-muscle invasive BC (NMIBC).4 A risk-based therapeutic approach is recommended for these patients, which is usually comprised
M AN U
of transurethral resection of the tumor (TURB) with or without immediate and adjuvant intravesical instillation therapy.5 NMIBC generally carries a high risk of disease recurrence albeit a relatively low risk of disease progression and death.6 Frequent endoscopic therapeutic and surveillance procedures are necessary, making BC the most expensive cancer to treat per
TE D
patient.7
Cancer prognosis is typically assessed at the time of diagnosis and estimated for a certain time interval (e.g. 5-year survival). However, these estimates become less relevant when the followup time increases and the impact of established prognostic factors changes. In contrast to this
EP
classical concept, conditional survival (CS) accounts for survival time and thereby provides better and more dynamic estimates of outcome probabilities at each follow-up point.8 It
AC C
measures the probability that a patient will survive some additional number of years, given that the patient has already survived for a certain period of time.9 It is thought that CS may help to individualize the prediction of prognosis, patient counseling and postoperative surveillance schedules.9
The benefits of CS estimates have already been demonstrated in large population-based cohorts of patients with various types of cancer.10–12 Further, several studies have investigated the relevance and usefulness of CS in urologic malignancies, including renal cell carcinoma13–15 3
ACCEPTED MANUSCRIPT
and penile cancer.16 In BC, there are two reports addressing CS for patients treated with radical cystectomy (RC).8,17 These studies showed that the risk of mortality decreases with the increasing survivorship and that the impact of classical prognostic factors such as tumor stage
RI PT
and lymph node metastasis becomes less significant over time. Despite its high prevalence and clinical relevance, however, no data are available on conditional outcomes in patients with NMIBC.
SC
For this reason, the aim of this study was to evaluate conditional recurrence and progression in a single-center cohort of patients with primary NMIBC and to explore the prognostic impact of
PATIENTS AND METHODS
TE D
Inclusion/exclusion criteria
M AN U
established factors and risk groups over time.
We screened our prospectively maintained single-institutional database of 2133 consecutive patients who underwent a TURB of a newly diagnosed bladder tumor between 1993 and 2014
EP
for inclusion. Patients with a primary urothelial NMIBC classified as either a Ta or T1 lesion were included; therefore, we excluded patients with benign bladder tumors, other malignant tumors
AC C
which histologically were not defined as urothelial BC, primary T2 disease, a T2 tumor on restaging TURB, previous upper tract urothelial carcinoma and those who underwent a RC for primary NMIBC. A total of 1292 patients were finally analyzed.
4
ACCEPTED MANUSCRIPT
Variables A computerized database abstracting clinical and pathological data of these patients was generated. Database variables included age, gender, pathological tumor stage, grade, tumor
RI PT
size, number of tumors, immediate and adjuvant instillation therapies, postoperative follow-up interval, recurrence and progression.
TURB specimens were processed according to standard pathological procedures and analyzed
SC
by two experts in genitourinary pathology (AH, MS). Tumors were classified according to the 2009 UICC classification18 and the tumor grade was assigned according to the 1973 World
M AN U
Health Organization scheme. The tumor size was evaluated as categorical variable according to the 3 cm cut-off.6 Based on pathological tumor classification, grade, number of tumors, tumor size and concomitant carcinoma in situ (CIS), patients were further stratified into risk groups
Treatment and follow-up
TE D
according to guidelines.5
EP
All patients underwent a TURB. Over the years, there was no standard or routine repeat TURB. Among the 526 high-risk cases, 68% had a repeat TURB and 87% had detrusor muscle in the
AC C
specimen. Immediate postoperative and adjuvant instillation therapies were administered at the surgeon’s discretion and according to guideline criteria. Bacillus Calmette-Guérin (BCG) was usually given in case of high-risk NMIBC weekly for 6 weeks followed by one year of maintenance. Patients were generally followed every 3 to 6 months during the first 2 years after diagnosis, biannually up to 5 years, and annually thereafter. Follow-up consisted of a patient’s medical history, physical examination, urinary cytology, cystoscopy, and another TURB if suspicious lesions were detected.
5
ACCEPTED MANUSCRIPT
Statistical analyses The endpoints of this study were the time to first recurrence in the bladder and the time to progression. Progression was defined as disease relapse in the bladder at tumor stage T2 or
RI PT
higher or the development of metastases and calculated from the date of first TURB. In our cohort, no patient developed lymph node or distant metastases without local T2 disease. Because patients with NMIBC are at a significant risk of dying from competing causes,
SC
competing-risks analyses were used to determine the rates of recurrence and progression. Cumulative incidence functions were generated, and conditional recurrence and progression
M AN U
rates were obtained using the multiplicative law of probability. Estimates were stratified by clinical and pathological prognostic variables, including age, gender, tumor stage, grade, CIS, size, number of tumors and risk group, and compared using Gray’s tests. Multivariable estimates from Fine-Gray’s regression models were obtained at baseline as subhazard ratios (SHR) and 95% confidence intervals (95% CI), and from landmark multivariable analyses at 6, 12, 24, 36,
TE D
and 48 months. All analyses were conducted with the “survival” and the “cmprsk” package in R 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria). Statistical testing was two-sided
AC C
RESULTS
EP
and a p-value <0.05 was considered statistically significant.
The median age of the 1292 patients was 69.3 years (IQR 61.5-77.3 years). About two-thirds of all tumors were staged Ta (n=872) and the majority were graded as G2 (n=677, 52.4%). A single immediate postoperative instillation of mitomycin C and adjuvant BCG was administered in 244 (19%) and 361 cases (28%), respectively. The median follow-up was 48 months. During the observation period, 517 patients (40%) developed disease recurrence, 82 (6%) tumors
6
ACCEPTED MANUSCRIPT
progressed, and 250 patients (19%) died without having experienced disease recurrence or
RI PT
progression.
Conditional recurrence rates
The 2-year recurrence rate at baseline was 36% and improved as a function of time that the
SC
patient had been free of disease recurrence. After 6-, 12-, 24-, 36-, and 48-months, the 2-year conditional recurrence rate improved to 31% (14% improvement compared with baseline), 22%
M AN U
(39%), 16% (56%), 13% (64%), and 11% (69%), respectively (Table 1, Fig. 1). Table 1 shows the 2-year recurrence rates in relation to clinical and pathological characteristics. At baseline, the majority of variables were significantly associated with recurrence, including tumor stage (p=0.04), grade (p=0.03), number of tumors (p<0.01), CIS (p=0.04), tumor size (p=0.03) and risk group (p<0.01). There was a significant difference in 2-year recurrence rates
TE D
between low- and intermediate-risk patients at baseline (p<0.01) and after 6 months (p<0.01), while patients of the intermediate- and high-risk group had similar recurrence rates at each time point (each p>0.15). Both intermediate and high-risk patients showed the greatest improvements
EP
in conditional recurrence rates. The impact of all prognostic factors progressively decreased over time, and no variable was longer associated with recurrence after 24 months survivorship.
AC C
Tumor stage regained statistical significance after 48 months (p<0.01). Table 2 details conditional recurrence rates for certain time intervals for all patients and according to risk group. For example, the baseline 5-year recurrence rate was 47%, while only 20% of patients recurred at 5 years if they had been without recurrence for 2 years. There were only clinically relevant differences according to risk group in the first year after TURB.
7
ACCEPTED MANUSCRIPT
Conditional progression rates Figure 2 shows cumulative incidence curves for progression. In general, conditional progression rates improved as a function of time that the patient had been progression-free, although relative
RI PT
differences were less distinct (Fig. 2). Table 3 shows the changing impact of clinical and pathological variables on progression. Tumor stage, grade, CIS and risk group (low/intermediate- versus high-risk) were significantly associated with 2-year conditional progression
SC
rates at baseline and continuously predictive for the first 2 years (all p-values <0.01). Patients of the high-risk group had the greatest improvement in conditional progression rates (Table 3).
M AN U
Table 4 shows the conditional progression rates for the analyzed time points for all patients and according to the risk group. For example, the baseline 5-year progression rate was 7%, but improved to 3% if the tumor did not progress within the first 2 years. A patient who had remained
TE D
progression-free for 4 years had a 1% risk of progression in year 5.
Impact of instillation therapies and multivariable analyses
EP
A single immediate postoperative instillation of mitomycin C had no impact on recurrence and progression of the tumor at any time point on both univariable (data not shown) and multivariable
AC C
analyses (Table 5). While adjuvant BCG did not affect recurrence, it decreased the risk for progression on the univariable analysis at all analyzed time points (each p<0.05). On the multivariable analysis (Table 5) that adjusted for age, gender and risk group, BCG remained associated with a lower risk for progression even after 3 years.
8
ACCEPTED MANUSCRIPT
DISCUSSION We evaluated conditional recurrence and progression outcomes of patients with primary NMIBC. We demonstrated that the risk of recurrence and progression evolves over follow-up time. With
RI PT
increasing survivorship, recurrence and progression rates improved and the prognostic impact of established factors decreased and finally disappeared. The effect of adjuvant BCG on progression was long-lasting. We further provided tables with dynamic prognostic information
SC
that can be used during regular follow-up visits for patient counseling and surveillance planning in the near future. These estimates may be informative for both physicians and patients and may
M AN U
contribute to a more individualized approach for predicting patients’ outcomes. In NMIBC, the risk of recurrence and progression is related to clinical and pathological variables and is typically assessed at the time of TURB.6 Several prognostic factors have been combined in many prognostic models,5,6 but their accuracy is not satisfactory19 and they do not account for
TE D
the length of survivorship. Given these limitations, we studied conditional recurrence and progression estimates in patients with NMIBC. We showed that traditional recurrence and progression measures at the time of diagnosis underestimate patients’ long-term outcomes. Furthermore, we demonstrated that recurrence and progression rates improved the longer the
EP
patient had remained recurrence- and progression-free. These observations are important because many patients are followed intensively even years after TURB and demand information
AC C
regarding their long-term prognosis. Therefore, we provided tables with estimates for certain time points, which may be used for a more individualized patient counseling and surveillance planning. For example, a patient with high-risk NMIBC returning for follow-up at 36-months has a 13% risk of recurrence and a 2% risk of progression at 5 years (Tables 1-4). We also assessed how recurrence and progression evolved according to well-established clinical and pathological prognostic factors. The risk group was only associated with recurrence at 0, 6 and 12 months follow-up, and lost statistical significance at 24 months and thereafter. The 9
ACCEPTED MANUSCRIPT
analysis of progression showed comparable results. Interestingly, BCG was associated with a reduced risk of progression at all analyzed follow-up points. These data suggest that the protective effect of BCG induction and maintenance is long-lasting, which is supported by other
RI PT
authors.20,21 Due to the limited follow-up and the decreasing number of patients in each substratum, no multivariable analysis for BCG was possible beyond 36 months.
In the literature, the greatest differences in conditional outcomes rates have been demonstrated
SC
for patients with more advanced tumors compared to patients with low-risk tumors;8,13,17,22 and so did our study if patients were stratified according to risk groups. For example, patients of the
M AN U
high-risk group had a 2-year recurrence rate of 38% at the time of TURB, which improved to 13% (relative improvement 63%) after a 2-year follow-up. In contrast, the improvement was only 26% in patients with low-risk tumors. This example underscores that the conditional recurrence rates markedly improved over time in patients with higher risk disease. Similar data were seen in
priori risk of progression.
TE D
conditional progression rates, although gains were fairly lower due to the considerably lower a
Several limitations need to be mentioned. Due to the retrospective design, our study was prone to selection and collection bias. There was no slide re-review to confirm stage or grade, but only
EP
two pathologists were involved in the evaluation of BC cases over the years. Also, there was a lack of standardization in BC management, which included the indication for a restaging TURB,
AC C
instillation therapies and surveillance. Several prognostic factors such as smoking history, T1 sub-stage, lymphovascular invasion and duration of BCG therapy were not collected. Given the relatively short duration of follow-up, no conclusions can be drawn regarding long-term outcomes, especially beyond 5 years. Analyses of the small number of patients who were followed this long would have resulted in decreasing statistical power and an increased risk for type I statistical error. The waning prognostic impact of outcome predictors over time may reflect their biological irrelevance, but may also be a statistical artifact of decreasing sample size with 10
ACCEPTED MANUSCRIPT
longer follow-up. As only 82 tumors (6.4%) progressed to a muscle-invasive stage, this study had limited statistical power to analyze progression as an endpoint. Larger studies on patients at the highest risk of progression, i.e. those with T1G3 disease, may be a direction for further
RI PT
research. Our progression rate was substantially lower than in other studies,6 but the reasons for this difference remain unknown. We were not able to demonstrate an impact of a single immediate postoperative instillation of mitomycin C on recurrence. Mitomycin C reduces the risk
SC
for recurrence in general, but not in all subgroups.23 Our results may therefore be due to missing adjustment for confounding variables, but also due to missing sub-analyses of the target group
M AN U
that could benefit most. Future studies may focus on this particular subgroup. As only 19% of our patients received immediate mitomycin C, our cohort can be regarded as undertreated according to current standards. A greater use of mitomycin C has the potential to decrease the burden of NMBIC24 and may have impacted the recurrence and progression rates in our study. Finally, our study must be subjected to external validation, preferably on prospective cohorts of
EP
CONCLUSIONS
TE D
already existing randomized trials.
In patients with primary NMIBC, the risk of recurrence and progression evolves over follow-up
AC C
time. With increasing survivorship, recurrence and progression rates improve and the prognostic impact of established factors decreases. BCG appears to have a long-lasting protective effect on progression. We provide tables with dynamic prognostic information that can be used during regular follow-up visits for patient counseling and surveillance planning. These estimates may be informative for physicians and patients and may contribute to a more individualized approach for predicting patients’ outcomes. External validation is required before these data can be implemented in clinical routine.
11
ACCEPTED MANUSCRIPT
ACKNOWLEDGMENT
RI PT
This study was in part funded by the Medical-Scientific Fund of the Mayor of the City of Vienna (project number 14069/2014). The study was approved by the institutional review board of the
AC C
EP
TE D
M AN U
SC
Medical University of Vienna (protocol registration number 1573/2015).
12
ACCEPTED MANUSCRIPT
REFERENCES Ferlay J, Soerjomataram I I, Dikshit R et al: Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 2015; 136: E359.
2.
Smith ND, Prasad SM, Patel AR et al: Bladder Cancer Mortality in the United States: A Geographic and Temporal Analysis of Socioeconomic and Environmental Factors. J Urol 2015: doi: 10.1016/j.juro.2015.07.091 (in press).
3.
Bray F, Ren J-S, Masuyer E et al: Global estimates of cancer prevalence for 27 sites in the adult population in 2008. Int J Cancer 2013; 132: 1133.
4.
Ederer IA, Lucca I, Hofbauer SL et al: Histopathology and prognosis of de novo bladder tumors following solid organ transplantation. World J Urol 2015; 33: 2087.
5.
Babjuk M, Burger M, Zigeuner R et al: EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder: update 2013. Eur Urol 2013; 64: 639.
6.
Sylvester RJ, van der Meijden APM, Oosterlinck W et al: Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 2006; 49: 466.
7.
Yeung C, Dinh T and Lee J: The health economics of bladder cancer: an updated review of the published literature. Pharmacoeconomics 2014; 32: 1093.
8.
Ploussard G, Shariat SF, Dragomir A et al: Conditional Survival After Radical Cystectomy for Bladder Cancer: Evidence for a Patient Changing Risk Profile over Time. Eur Urol 2014; 66: 361.
9.
Zabor EC, Gonen M, Chapman PB et al: Dynamic prognostication using conditional survival estimates. Cancer 2013; 119: 3589.
10.
Ito Y, Nakayama T, Miyashiro I et al: Conditional survival for longer-term survivors from 2000-2004 using population-based cancer registry data in Osaka, Japan. BMC Cancer 2013; 13: 304.
11.
Yu XQ, Baade PD and O’Connell DL: Conditional survival of cancer patients: an Australian perspective. BMC Cancer 2012; 12: 460.
12.
Janssen-Heijnen MLG, Gondos A, Bray F et al: Clinical relevance of conditional survival of cancer patients in europe: age-specific analyses of 13 cancers. J Clin Oncol 2010; 28: 2520.
13.
Bianchi M, Becker A, Hansen J et al: Conditional survival after nephrectomy for renal cell carcinoma (RCC): changes in future survival probability over time. BJU Int 2013; 111: E283.
14.
Abdollah F, Suardi N, Capitanio U et al: The key role of time in predicting progression-free survival in patients with renal cell carcinoma treated with partial or radical nephrectomy: conditional survival analysis. Urol Oncol 2014; 32: 43.e9.
15.
Harshman LC, Xie W, Bjarnason GA et al: Conditional survival of patients with metastatic renal-cell carcinoma treated with VEGF-targeted therapy: a population-based study. Lancet Oncol 2012; 13: 927.
16.
Thuret R, Sun M, Abdollah F et al: Conditional survival predictions after surgery for patients with penile carcinoma. Cancer 2011; 117: 3723.
AC C
EP
TE D
M AN U
SC
RI PT
1.
13
ACCEPTED MANUSCRIPT
Sun M, Abdollah F, Bianchi M et al: Conditional survival of patients with urothelial carcinoma of the urinary bladder treated with radical cystectomy. Eur J Cancer 2012; 48: 1503.
18.
Edge SB, Byrd DR, Compton CC et al eds: Urinary bladder. In: AJCC Cancer Staging Manual. 7th ed. New York, NY: Springer 2010; pp 497–505.
19.
Xylinas E, Kent M, Kluth L et al: Accuracy of the EORTC risk tables and of the CUETO scoring model to predict outcomes in non-muscle-invasive urothelial carcinoma of the bladder. Br J Cancer 2013; 109: 1460.
20.
Lamm DL: Getting the Most Out of Bacillus Calmette-Guérin for Treatment of Bladder Cancer. J Urol 2016; 195: 7.
21.
Yokomizo A, Kanimoto Y, Okamura T et al: Randomized Controlled Study of the Efficacy, Safety and Quality of Life of Low-Dose Bacillus Calmette-Guérin Instillation Therapy for Non-Muscle-Invasive Bladder Cancer. J Urol 2016; 195: 41.
22.
Ploussard G, Xylinas E, Lotan Y et al: Conditional Survival Nephroureterectomy for Upper Tract Carcinoma. Eur Urol 2015; 67: 803.
23.
Sylvester RJ, Oosterlinck W, Holmang S et al: Systematic Review and Individual Patient Data Meta-analysis of Randomized Trials Comparing a Single Immediate Instillation of Chemotherapy After Transurethral Resection with Transurethral Resection Alone in Patients with Stage pTa-pT1 Urothelial Carcinoma. Eur Urol 2015: doi 10.1016/j.eururo.2015.05.050 (in press).
24.
Lee CT, Barocas D, Globe DR et al: Economic and humanistic consequences of preventable bladder tumor recurrences in nonmuscle invasive bladder cancer cases. J Urol 2012; 188: 2114.
After
Radical
AC C
EP
TE D
M AN U
SC
RI PT
17.
14
ACCEPTED MANUSCRIPT
FIGURE LEGENDS Figure 1
Cumulative incidence curves for recurrence (black line) overlaid by conditional
RI PT
recurrence estimates at 6 months (red line), 12 months (green line), 24 months
Figure 2
SC
(blue line), 36 months (yellow line) and 48 months (pink line).
Cumulative incidence curves for progression (black line) overlaid by conditional
M AN U
progression estimates at 6 months (red line), 12 months (green line), 24 months
AC C
EP
TE D
(blue line), 36 months (yellow line) and 48 months (pink line).
15
ACCEPTED MANUSCRIPT
2-year conditional recurrence rates in relation to clinical and pathological characteristics.
35% 37% 0.65
29% 32% 0.69
22% 22% 0.46
16% 16% 0.60
1001 (77.5) 291 (22.5)
36% 35% 0.80
31% 31% 0.98
22% 20% 0.72
16% 18% 0.92
872 (67.5) 420 (32.5)
35% 40% 0.04
30% 32% 0.14
23% 20% 0.66
275 (21.3) 677 (52.4) 340 (26.3)
30% 37% 41% 0.03
27% 32% 33% 0.33
21% 22% 24% 0.67
718 (55.6) 574 (44.4)
29% 45% <0.01
24% 40% <0.01
1135 (87.8) 157 (12.2)
35% 44% 0.04
31% 31% 0.64
1138 (88.1) 154 (11.9)
165 (12.8) 601 (46.5) 526 (40.7)
13% 14% 0.54
8% 12% 0.71
13% 14% 0.95
11% 8% 0.62
17% 14% 0.20
13% 14% 0.09
8% 17% <0.01
16% 16% 17% 0.79
13% 14% 13% 0.75
11% 11% 10% 0.72
18% 28% <0.01
13% 21% 0.14
10% 18% 0.45
10% 11% 0.75
22% 23% 0.53
16% 16% 0.59
13% 14% 0.52
11% 9% 0.39
TE D
M AN U
SC
447 (34.6) 845 (65.4)
RI PT
n (%) -
EP
Cohort, n 2-yr conditional recurrence rate Age <65 yrs ≥65 yrs p-value Gender Male Female p-value Tumor stage pTa pT1 p-value Tumor grade G1 G2 G3 p-value Number of tumors Solitary Multiple p-value CIS Absent Present p-value Tumor size <3cm ≥3cm p-value Risk group Low Intermediate High p-value
Time elapsed since transurethral resection of the bladder cancer Baseline 6 months 12 months 24 months 36 months 48 months 1292 895 706 518 409 311 36% 31% 22% 16% 13% 11%
AC C
Table 1
35% 47% 0.03
30% 42% 0.05
21% 33% 0.03
16% 15% 0.46
13% 9% 0.66
10% 15% 0.27
19% 39% 38% <0.01
18% 35% 31% <0.01
15% 26% 20% 0.06
14% 19% 14% 0.55
12% 15% 13% 0.58
9% 9% 14% 0.38
ACCEPTED MANUSCRIPT
Table 2
24 months
36 months
13% 6% 13% 16%
26% 11% 27% 29% 16% 6% 17% 18%
36% 19% 39% 38% 28% 14% 31% 29% 15% 9% 18% 14%
41% 26% 44% 42% 33% 20% 37% 33% 22% 15% 26% 20% 9% 8% 10% 7%
48 months
60 months
45% 30% 49% 45% 38% 25% 43% 38% 28% 21% 33% 25% 16% 14% 19% 14% 9% 8% 10% 8%
47% 31% 51% 47% 41% 28% 45% 40% 31% 25% 35% 29% 20% 18% 22% 18% 13% 12% 15% 13% 5% 5% 5% 6%
TE D
M AN U
SC
12 months
EP
All LR Baseline IR HR All LR 6 months IR HR All LR 12 Recurrence months IR rate to a HR certain no. of All months from LR 24 TURB months IR HR All LR 36 months IR HR All LR 48 months IR HR
6 months
AC C
Total time without recurrence from TURB
RI PT
Conditional recurrence rates for patients with NMIBC who reach a certain time point without having developed recurrence. For example, if a patient stayed recurrence-free for 48 months, his chance of developing recurrent disease at 60 months is 5%. Data are presented for the overall cohort and according to risk group (LR, low-risk; IR, intermediate-risk; HR, high-risk).
ACCEPTED MANUSCRIPT
2-year conditional progression rates in relation to clinical and pathological characteristics.
1% 5% 0.06
1% 3% 0.15
1% 3% 0.86
0.5% 3% 0.68
1001 (77.5) 291 (22.5)
5% 4% 0.95
4% 3% 0.91
3% 5% 0.75
2% 4% 0.56
2% 3% 0.56
2% 2% 0.67
872 (67.5) 420 (32.5)
1% 11% <0.01
1% 9% <0.01
1% 10% <0.01
1% 6% <0.01
2% 3% 0.06
2% 3% 0.10
275 (21.3) 677 (52.4) 340 (26.3)
1% 3% 11% <0.01
1% 2% 10% <0.01
1% 2% 11% <0.01
1% 2% 6% 0.01
1% 3% 2% 0.62
2% 2% 3% 0.52
718 (55.6) 574 (44.4)
4% 6% 0.07
3% 5% 0.05
3% 5% 0.18
2% 3% 0.34
2% 2% 0.71
2% 2% 0.63
1135 (87.8) 157 (12.2)
4% 8% <0.01
3% 6% <0.01
3% 10% <0.01
2% 8% <0.01
2% 2% 0.34
2% 2% 0.33
RI PT
2% 4% 0.13
SC
4% 5% 0.30
TE D
M AN U
447 (34.6) 845 (65.4)
EP
Cohort, n 2-year conditional progression rate Age <65 yrs ≥65 yrs p-value Gender Male Female p-value Tumor stage pTa pT1 p-value Tumor grade G1 G2 G3 p-value Number of tumors Solitary Multiple p-value CIS Absent Present p-value Tumor size <3cm ≥3cm p-value Risk group Low Intermediate High p-value
Time elapsed since transurethral resection of the bladder cancer Baseline 6 months 12 months 24 months 36 months 48 months 1292 1018 916 761 639 525 4% 4% 4% 3% 2% 2%
n (%) -
AC C
Table 3
1138 (88.1) 154 (11.9)
4% 4% 0.92
4% 3% 0.99
4% 2% 0.76
3% 1% 0.82
2% 0% 0.77
2% 2% 0.37
165 (12.8) 601 (46.5) 526 (40.7)
2% 2% 9% <0.01
2% 1% 8% <0.01
1% 1% 8% <0.01
0% 2% 5% <0.01
0% 3% 2% 0.27
3% 1% 2% 0.33
ACCEPTED MANUSCRIPT
Table 4
Progression 12 months rates to a certain no. of months from TURB 24 months
36 months
48 months
36 months
1% 0% 1% 3%
2% 1% 1% 4% 1% 1% 0.4% 2%
4% 2% 2% 9% 3% 2% 1% 6% 2% 1% 1% 5%
6% 2% 2% 11% 5% 2% 1% 10% 4% 1% 1% 8% 2% 0% 0.3% 4%
48 months
M AN U
SC
24 months
TE D
6 months
12 months
EP
Baseline
All LR IR HR All LR IR HR All LR IR HR All LR IR HR All LR IR HR All LR IR HR
6 months
AC C
Total time without progression from TURB
RI PT
Conditional progression rates for patients with NMIBC who reach a certain time point without having developed progression. For example, if a patient stayed progression-free for 24 months, his chance of developing progression at 60 months is 3%. Data are presented for the overall cohort and according to risk group (LR, low-risk; IR, intermediate-risk; HR, high-risk).
7% 2% 3% 12% 6% 2% 3% 11% 5% 1% 2% 9% 3% 0% 2% 5% 1% 0% 1% 1%
60 months 7% 2% 4% 13% 6% 2% 4% 11% 5% 1% 3% 10% 3% 0% 3% 6% 2% 0% 3% 2% 1% 0% 1% 1%
ACCEPTED MANUSCRIPT
Table 5 Multivariable Fine-Gray’s competing risks regression models predicting recurrence (A) and progression (B). A 24 months
0.96 0.80-1.15 0.63
1.04 0.85-1.29 0.69
0.92 0.71-1.19 0.53
0.93 0.66-1.32 0.69
1.00 0.81-1.23 1.00
0.98 0.77-1.27 0.90
1.05 0.76-1.44 0.79
0.99 0.64-1.51 0.95
1.00
1.00
1.00
1.77 1.32-2.38 <0.01
1.80 1.29-2.50 <0.01
1.51 1.03-2.21 0.04
1.61 1.16-2.23 <0.01
1.73 1.19-2.51 <0.01
1.30 0.83-2.05 0.26
1.01 0.84-1.23 0.87
0.98 0.78-1.23 0.87
1.16 0.90-1.50 0.25
0.91 0.66-1.26 0.58
36 months
48 months
0.91 0.60-1.37 0.64
0.96 0.57-1.62 0.87
RI PT
12 months
SC
6 months
1.22 0.59-2.51 0.60
1.00
1.00
1.00
1.24 0.77-2.00 0.38
1.22 0.58-2.18 0.51
1.14 0.54-2.41 0.74
1.27 0.72-2.24 0.41
1.38 0.70-2.73 0.35
1.89 0.84-4.29 0.13
1.10 0.83-1.45 0.50
1.15 0.79-1.66 0.47
1.17 0.75-1.83 0.48
0.98 0.52-1.75 0.93
1.02 0.67-1.54 0.94
0.88 0.50-1.53 0.66
0.90 0.48-1.72 0.76
0.67 0.31-1.41 0.29
EP
TE D
M AN U
0.98 0.59-1.65 0.95
AC C
Age >65 yr SHR 95% CI p-value Male gender SHR 95% CI p-value Risk group Low SHR Intermediate SHR 95% CI p-value High SHR 95% CI p-value Mitomycin C SHR 95% CI p-value BCG SHR 95% CI p-value
Baseline
ACCEPTED MANUSCRIPT
B 12 months
24 months
36 months
48 months
1.17 0.72-1.91 0.52
1.41 0.82-2.45 0.22
1.67 0.90-3.07 0.10
1.71 0.82-3.58 0.16
1.14 0.51-2.55 0.75
1.27 0.47-3.41 0.63
0.95 0.57-1.61 0.86
0.92 0.52-1.63 0.77
0.85 0.46-1.59 0.62
0.74 0.34-1.59 0.44
1.25 0.43-3.70 0.68
1.22 0.34-4.34 0.76
1.00
1.00
1.00
1.00
1.00
1.00
1.54 0.53-4.45 0.43
1.37 0.46-4.07 0.57
1.66 0.48-5.78 0.42
2.13 0.47-9.59 0.32
1.97 0.43-8.97 0.38
1.29 0.26-6.42 0.75
4.19 1.47-11.97 <0.01
3.78 1.30-11.01 0.02
5.19 1.52-17.67 <0.01
4.78 1.05-21.65 0.02
2.52 0.50-12.61 0.26
2.41 0.45-12.93 0.30
1.48 0.90-2.44 0.13
1.36 0.77-2.39 0.29
1.03 0.55-1.91 0.94
1.05 0.47-2.35 0.91
1.13 0.41-3.12 0.82
0.85 0.26-2.81 0.79
0.55 0.30-0.99 0.04
0.50 0.25-0.97 0.04
0.55 0.27-1.12 0.09
0.26 0.08-0.84 0.02
0.11 0.02-0.83 0.03
N.A.
SC
M AN U
TE D
EP
RI PT
6 months
AC C
Age >65 yr SHR 95% CI p-value Male gender SHR 95% CI p-value Risk group Low SHR Intermediate SHR 95% CI p-value High SHR 95% CI p-value Mitomycin C SHR 95% CI p-value BCG SHR 95% CI p-value
Baseline
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT ABBREVIATIONS BC: bladder cancer BCG: Bacillus Calmette-Guérin
NMIBC: non-muscle invasive bladder cancer
AC C
EP
TE D
M AN U
TURB: transurethral resection of the bladder
SC
CS: conditional survival
RI PT
CIS: carcinoma in situ
1
S0022-5347(16)00225-1 10.1016/j.juro.2016.01.102 JURO 13295
To appear in: The Journal of Urology Accepted Date: 20 January 2016 Please cite this article as: Leitner CV, Ederer IA, de Martino M, Hofbauer SL, Lucca I, Mbeutcha A, Mathieu R, Haitel A, Susani M, Shariat SF, Klatte T, Dynamic prognostication using conditional recurrence and progression estimates for patients with non-muscle invasive bladder cancer, The Journal of Urology® (2016), doi: 10.1016/j.juro.2016.01.102. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain.
Embargo Policy All article content is under embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date. Questions regarding embargo should be directed to [email protected].
ACCEPTED MANUSCRIPT
Dynamic prognostication using conditional recurrence and progression estimates for patients with non-muscle invasive bladder cancer
RI PT
Carmen V. Leitner1, Ines A. Ederer1, Michela de Martino1, Sebastian L. Hofbauer1, Ilaria Lucca1, Aurélie Mbeutcha1, Romain Mathieu1, Andrea Haitel2, Martin Susani2, Shahrokh F. Shariat1,3,4, Tobias Klatte1
1
M AN U
SC
Department of Urology, Medical University of Vienna, Vienna General Hospital, Vienna, Austria, 2 Clinical Institute of Pathology, Medical University of Vienna, Vienna General Hospital, Vienna, Austria, 3Department of Urology, University of Texas Southwestern Medical Center, Dallas, TX, 4 Department of Urology, Weill Cornell Medical College, New York-Presbyterian Hospital, New York, NY
Word count: 2498 Tables: 5
*Corresponding author: Tobias Klatte Department of Urology
TE D
Figures: 2
EP
Medical University of Vienna Währinger Gürtel 18-20
AC C
A-1090 Vienna, Austria
Phone: +43-1-40400-26020 Fax: +43-1-40400-23320
Email: [email protected]
Keywords: recurrence; progression; BCG; mitomycin; risk group
1
ACCEPTED MANUSCRIPT
ABSTRACT Purpose: Conditional estimates provide a dynamic prediction of outcomes, but there are no data
RI PT
for non-muscle invasive bladder cancer (NMIBC). The purpose of this study was to assess the changes in conditional recurrence and progression rates after transurethral resection of the bladder (TURB) and to explore the prognostic impact of established factors and risk groups over
SC
time.
Methods: We retrospectively analyzed data from 1292 consecutive patients with a newly
recurrence and time to progression.
M AN U
diagnosed Ta/T1 BC who underwent a TURB. The endpoints of this study were time to first
Results: The 2-year recurrence rate at baseline was 36% and improved as a function of time that the patient had been free of disease recurrence. After 6-, 12-, 24-, 36-, and 48-months, the
TE D
2-year conditional recurrence rate improved to 31% (14% improvement compared with baseline), 22% (39%), 16% (56%), 13% (64%), and 11% (69%), respectively. Comparably, conditional progression rates improved with increasing follow-up, although relative differences were less distinct. The prognostic impact of established factors and the NMIBC risk group
EP
progressively decreased over time and finally disappeared. BCG, however, had a protective
AC C
effect on progression even after 3 years. We also provide tables with dynamic prognostic information at all analyzed time points. Conclusions: In patients with primary Ta/T1 BC, recurrence and progression rates improve over time. The prognostic impact of established factors and risk groups decreases and finally disappears. The effect of BCG on progression is long-lasting. Conditional outcome estimates may improve patient counseling and individualize surveillance planning.
2
ACCEPTED MANUSCRIPT
INTRODUCTION Bladder cancer (BC) is the second most common urologic malignancy with more than 400,000 new cases worldwide yearly.1 It is expected to remain a significant burden for healthcare
RI PT
systems in the next decades because of the aging and continuous growth of the world population as well as the high prevalence of smoking.2 As a measure for the needs of medical care, the 5-year global BC prevalence has been estimated to be more than 1.1 million.3 At initial
SC
diagnosis, approximately 75% of patients present with a non-muscle invasive BC (NMIBC).4 A risk-based therapeutic approach is recommended for these patients, which is usually comprised
M AN U
of transurethral resection of the tumor (TURB) with or without immediate and adjuvant intravesical instillation therapy.5 NMIBC generally carries a high risk of disease recurrence albeit a relatively low risk of disease progression and death.6 Frequent endoscopic therapeutic and surveillance procedures are necessary, making BC the most expensive cancer to treat per
TE D
patient.7
Cancer prognosis is typically assessed at the time of diagnosis and estimated for a certain time interval (e.g. 5-year survival). However, these estimates become less relevant when the followup time increases and the impact of established prognostic factors changes. In contrast to this
EP
classical concept, conditional survival (CS) accounts for survival time and thereby provides better and more dynamic estimates of outcome probabilities at each follow-up point.8 It
AC C
measures the probability that a patient will survive some additional number of years, given that the patient has already survived for a certain period of time.9 It is thought that CS may help to individualize the prediction of prognosis, patient counseling and postoperative surveillance schedules.9
The benefits of CS estimates have already been demonstrated in large population-based cohorts of patients with various types of cancer.10–12 Further, several studies have investigated the relevance and usefulness of CS in urologic malignancies, including renal cell carcinoma13–15 3
ACCEPTED MANUSCRIPT
and penile cancer.16 In BC, there are two reports addressing CS for patients treated with radical cystectomy (RC).8,17 These studies showed that the risk of mortality decreases with the increasing survivorship and that the impact of classical prognostic factors such as tumor stage
RI PT
and lymph node metastasis becomes less significant over time. Despite its high prevalence and clinical relevance, however, no data are available on conditional outcomes in patients with NMIBC.
SC
For this reason, the aim of this study was to evaluate conditional recurrence and progression in a single-center cohort of patients with primary NMIBC and to explore the prognostic impact of
PATIENTS AND METHODS
TE D
Inclusion/exclusion criteria
M AN U
established factors and risk groups over time.
We screened our prospectively maintained single-institutional database of 2133 consecutive patients who underwent a TURB of a newly diagnosed bladder tumor between 1993 and 2014
EP
for inclusion. Patients with a primary urothelial NMIBC classified as either a Ta or T1 lesion were included; therefore, we excluded patients with benign bladder tumors, other malignant tumors
AC C
which histologically were not defined as urothelial BC, primary T2 disease, a T2 tumor on restaging TURB, previous upper tract urothelial carcinoma and those who underwent a RC for primary NMIBC. A total of 1292 patients were finally analyzed.
4
ACCEPTED MANUSCRIPT
Variables A computerized database abstracting clinical and pathological data of these patients was generated. Database variables included age, gender, pathological tumor stage, grade, tumor
RI PT
size, number of tumors, immediate and adjuvant instillation therapies, postoperative follow-up interval, recurrence and progression.
TURB specimens were processed according to standard pathological procedures and analyzed
SC
by two experts in genitourinary pathology (AH, MS). Tumors were classified according to the 2009 UICC classification18 and the tumor grade was assigned according to the 1973 World
M AN U
Health Organization scheme. The tumor size was evaluated as categorical variable according to the 3 cm cut-off.6 Based on pathological tumor classification, grade, number of tumors, tumor size and concomitant carcinoma in situ (CIS), patients were further stratified into risk groups
Treatment and follow-up
TE D
according to guidelines.5
EP
All patients underwent a TURB. Over the years, there was no standard or routine repeat TURB. Among the 526 high-risk cases, 68% had a repeat TURB and 87% had detrusor muscle in the
AC C
specimen. Immediate postoperative and adjuvant instillation therapies were administered at the surgeon’s discretion and according to guideline criteria. Bacillus Calmette-Guérin (BCG) was usually given in case of high-risk NMIBC weekly for 6 weeks followed by one year of maintenance. Patients were generally followed every 3 to 6 months during the first 2 years after diagnosis, biannually up to 5 years, and annually thereafter. Follow-up consisted of a patient’s medical history, physical examination, urinary cytology, cystoscopy, and another TURB if suspicious lesions were detected.
5
ACCEPTED MANUSCRIPT
Statistical analyses The endpoints of this study were the time to first recurrence in the bladder and the time to progression. Progression was defined as disease relapse in the bladder at tumor stage T2 or
RI PT
higher or the development of metastases and calculated from the date of first TURB. In our cohort, no patient developed lymph node or distant metastases without local T2 disease. Because patients with NMIBC are at a significant risk of dying from competing causes,
SC
competing-risks analyses were used to determine the rates of recurrence and progression. Cumulative incidence functions were generated, and conditional recurrence and progression
M AN U
rates were obtained using the multiplicative law of probability. Estimates were stratified by clinical and pathological prognostic variables, including age, gender, tumor stage, grade, CIS, size, number of tumors and risk group, and compared using Gray’s tests. Multivariable estimates from Fine-Gray’s regression models were obtained at baseline as subhazard ratios (SHR) and 95% confidence intervals (95% CI), and from landmark multivariable analyses at 6, 12, 24, 36,
TE D
and 48 months. All analyses were conducted with the “survival” and the “cmprsk” package in R 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria). Statistical testing was two-sided
AC C
RESULTS
EP
and a p-value <0.05 was considered statistically significant.
The median age of the 1292 patients was 69.3 years (IQR 61.5-77.3 years). About two-thirds of all tumors were staged Ta (n=872) and the majority were graded as G2 (n=677, 52.4%). A single immediate postoperative instillation of mitomycin C and adjuvant BCG was administered in 244 (19%) and 361 cases (28%), respectively. The median follow-up was 48 months. During the observation period, 517 patients (40%) developed disease recurrence, 82 (6%) tumors
6
ACCEPTED MANUSCRIPT
progressed, and 250 patients (19%) died without having experienced disease recurrence or
RI PT
progression.
Conditional recurrence rates
The 2-year recurrence rate at baseline was 36% and improved as a function of time that the
SC
patient had been free of disease recurrence. After 6-, 12-, 24-, 36-, and 48-months, the 2-year conditional recurrence rate improved to 31% (14% improvement compared with baseline), 22%
M AN U
(39%), 16% (56%), 13% (64%), and 11% (69%), respectively (Table 1, Fig. 1). Table 1 shows the 2-year recurrence rates in relation to clinical and pathological characteristics. At baseline, the majority of variables were significantly associated with recurrence, including tumor stage (p=0.04), grade (p=0.03), number of tumors (p<0.01), CIS (p=0.04), tumor size (p=0.03) and risk group (p<0.01). There was a significant difference in 2-year recurrence rates
TE D
between low- and intermediate-risk patients at baseline (p<0.01) and after 6 months (p<0.01), while patients of the intermediate- and high-risk group had similar recurrence rates at each time point (each p>0.15). Both intermediate and high-risk patients showed the greatest improvements
EP
in conditional recurrence rates. The impact of all prognostic factors progressively decreased over time, and no variable was longer associated with recurrence after 24 months survivorship.
AC C
Tumor stage regained statistical significance after 48 months (p<0.01). Table 2 details conditional recurrence rates for certain time intervals for all patients and according to risk group. For example, the baseline 5-year recurrence rate was 47%, while only 20% of patients recurred at 5 years if they had been without recurrence for 2 years. There were only clinically relevant differences according to risk group in the first year after TURB.
7
ACCEPTED MANUSCRIPT
Conditional progression rates Figure 2 shows cumulative incidence curves for progression. In general, conditional progression rates improved as a function of time that the patient had been progression-free, although relative
RI PT
differences were less distinct (Fig. 2). Table 3 shows the changing impact of clinical and pathological variables on progression. Tumor stage, grade, CIS and risk group (low/intermediate- versus high-risk) were significantly associated with 2-year conditional progression
SC
rates at baseline and continuously predictive for the first 2 years (all p-values <0.01). Patients of the high-risk group had the greatest improvement in conditional progression rates (Table 3).
M AN U
Table 4 shows the conditional progression rates for the analyzed time points for all patients and according to the risk group. For example, the baseline 5-year progression rate was 7%, but improved to 3% if the tumor did not progress within the first 2 years. A patient who had remained
TE D
progression-free for 4 years had a 1% risk of progression in year 5.
Impact of instillation therapies and multivariable analyses
EP
A single immediate postoperative instillation of mitomycin C had no impact on recurrence and progression of the tumor at any time point on both univariable (data not shown) and multivariable
AC C
analyses (Table 5). While adjuvant BCG did not affect recurrence, it decreased the risk for progression on the univariable analysis at all analyzed time points (each p<0.05). On the multivariable analysis (Table 5) that adjusted for age, gender and risk group, BCG remained associated with a lower risk for progression even after 3 years.
8
ACCEPTED MANUSCRIPT
DISCUSSION We evaluated conditional recurrence and progression outcomes of patients with primary NMIBC. We demonstrated that the risk of recurrence and progression evolves over follow-up time. With
RI PT
increasing survivorship, recurrence and progression rates improved and the prognostic impact of established factors decreased and finally disappeared. The effect of adjuvant BCG on progression was long-lasting. We further provided tables with dynamic prognostic information
SC
that can be used during regular follow-up visits for patient counseling and surveillance planning in the near future. These estimates may be informative for both physicians and patients and may
M AN U
contribute to a more individualized approach for predicting patients’ outcomes. In NMIBC, the risk of recurrence and progression is related to clinical and pathological variables and is typically assessed at the time of TURB.6 Several prognostic factors have been combined in many prognostic models,5,6 but their accuracy is not satisfactory19 and they do not account for
TE D
the length of survivorship. Given these limitations, we studied conditional recurrence and progression estimates in patients with NMIBC. We showed that traditional recurrence and progression measures at the time of diagnosis underestimate patients’ long-term outcomes. Furthermore, we demonstrated that recurrence and progression rates improved the longer the
EP
patient had remained recurrence- and progression-free. These observations are important because many patients are followed intensively even years after TURB and demand information
AC C
regarding their long-term prognosis. Therefore, we provided tables with estimates for certain time points, which may be used for a more individualized patient counseling and surveillance planning. For example, a patient with high-risk NMIBC returning for follow-up at 36-months has a 13% risk of recurrence and a 2% risk of progression at 5 years (Tables 1-4). We also assessed how recurrence and progression evolved according to well-established clinical and pathological prognostic factors. The risk group was only associated with recurrence at 0, 6 and 12 months follow-up, and lost statistical significance at 24 months and thereafter. The 9
ACCEPTED MANUSCRIPT
analysis of progression showed comparable results. Interestingly, BCG was associated with a reduced risk of progression at all analyzed follow-up points. These data suggest that the protective effect of BCG induction and maintenance is long-lasting, which is supported by other
RI PT
authors.20,21 Due to the limited follow-up and the decreasing number of patients in each substratum, no multivariable analysis for BCG was possible beyond 36 months.
In the literature, the greatest differences in conditional outcomes rates have been demonstrated
SC
for patients with more advanced tumors compared to patients with low-risk tumors;8,13,17,22 and so did our study if patients were stratified according to risk groups. For example, patients of the
M AN U
high-risk group had a 2-year recurrence rate of 38% at the time of TURB, which improved to 13% (relative improvement 63%) after a 2-year follow-up. In contrast, the improvement was only 26% in patients with low-risk tumors. This example underscores that the conditional recurrence rates markedly improved over time in patients with higher risk disease. Similar data were seen in
priori risk of progression.
TE D
conditional progression rates, although gains were fairly lower due to the considerably lower a
Several limitations need to be mentioned. Due to the retrospective design, our study was prone to selection and collection bias. There was no slide re-review to confirm stage or grade, but only
EP
two pathologists were involved in the evaluation of BC cases over the years. Also, there was a lack of standardization in BC management, which included the indication for a restaging TURB,
AC C
instillation therapies and surveillance. Several prognostic factors such as smoking history, T1 sub-stage, lymphovascular invasion and duration of BCG therapy were not collected. Given the relatively short duration of follow-up, no conclusions can be drawn regarding long-term outcomes, especially beyond 5 years. Analyses of the small number of patients who were followed this long would have resulted in decreasing statistical power and an increased risk for type I statistical error. The waning prognostic impact of outcome predictors over time may reflect their biological irrelevance, but may also be a statistical artifact of decreasing sample size with 10
ACCEPTED MANUSCRIPT
longer follow-up. As only 82 tumors (6.4%) progressed to a muscle-invasive stage, this study had limited statistical power to analyze progression as an endpoint. Larger studies on patients at the highest risk of progression, i.e. those with T1G3 disease, may be a direction for further
RI PT
research. Our progression rate was substantially lower than in other studies,6 but the reasons for this difference remain unknown. We were not able to demonstrate an impact of a single immediate postoperative instillation of mitomycin C on recurrence. Mitomycin C reduces the risk
SC
for recurrence in general, but not in all subgroups.23 Our results may therefore be due to missing adjustment for confounding variables, but also due to missing sub-analyses of the target group
M AN U
that could benefit most. Future studies may focus on this particular subgroup. As only 19% of our patients received immediate mitomycin C, our cohort can be regarded as undertreated according to current standards. A greater use of mitomycin C has the potential to decrease the burden of NMBIC24 and may have impacted the recurrence and progression rates in our study. Finally, our study must be subjected to external validation, preferably on prospective cohorts of
EP
CONCLUSIONS
TE D
already existing randomized trials.
In patients with primary NMIBC, the risk of recurrence and progression evolves over follow-up
AC C
time. With increasing survivorship, recurrence and progression rates improve and the prognostic impact of established factors decreases. BCG appears to have a long-lasting protective effect on progression. We provide tables with dynamic prognostic information that can be used during regular follow-up visits for patient counseling and surveillance planning. These estimates may be informative for physicians and patients and may contribute to a more individualized approach for predicting patients’ outcomes. External validation is required before these data can be implemented in clinical routine.
11
ACCEPTED MANUSCRIPT
ACKNOWLEDGMENT
RI PT
This study was in part funded by the Medical-Scientific Fund of the Mayor of the City of Vienna (project number 14069/2014). The study was approved by the institutional review board of the
AC C
EP
TE D
M AN U
SC
Medical University of Vienna (protocol registration number 1573/2015).
12
ACCEPTED MANUSCRIPT
REFERENCES Ferlay J, Soerjomataram I I, Dikshit R et al: Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 2015; 136: E359.
2.
Smith ND, Prasad SM, Patel AR et al: Bladder Cancer Mortality in the United States: A Geographic and Temporal Analysis of Socioeconomic and Environmental Factors. J Urol 2015: doi: 10.1016/j.juro.2015.07.091 (in press).
3.
Bray F, Ren J-S, Masuyer E et al: Global estimates of cancer prevalence for 27 sites in the adult population in 2008. Int J Cancer 2013; 132: 1133.
4.
Ederer IA, Lucca I, Hofbauer SL et al: Histopathology and prognosis of de novo bladder tumors following solid organ transplantation. World J Urol 2015; 33: 2087.
5.
Babjuk M, Burger M, Zigeuner R et al: EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder: update 2013. Eur Urol 2013; 64: 639.
6.
Sylvester RJ, van der Meijden APM, Oosterlinck W et al: Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 2006; 49: 466.
7.
Yeung C, Dinh T and Lee J: The health economics of bladder cancer: an updated review of the published literature. Pharmacoeconomics 2014; 32: 1093.
8.
Ploussard G, Shariat SF, Dragomir A et al: Conditional Survival After Radical Cystectomy for Bladder Cancer: Evidence for a Patient Changing Risk Profile over Time. Eur Urol 2014; 66: 361.
9.
Zabor EC, Gonen M, Chapman PB et al: Dynamic prognostication using conditional survival estimates. Cancer 2013; 119: 3589.
10.
Ito Y, Nakayama T, Miyashiro I et al: Conditional survival for longer-term survivors from 2000-2004 using population-based cancer registry data in Osaka, Japan. BMC Cancer 2013; 13: 304.
11.
Yu XQ, Baade PD and O’Connell DL: Conditional survival of cancer patients: an Australian perspective. BMC Cancer 2012; 12: 460.
12.
Janssen-Heijnen MLG, Gondos A, Bray F et al: Clinical relevance of conditional survival of cancer patients in europe: age-specific analyses of 13 cancers. J Clin Oncol 2010; 28: 2520.
13.
Bianchi M, Becker A, Hansen J et al: Conditional survival after nephrectomy for renal cell carcinoma (RCC): changes in future survival probability over time. BJU Int 2013; 111: E283.
14.
Abdollah F, Suardi N, Capitanio U et al: The key role of time in predicting progression-free survival in patients with renal cell carcinoma treated with partial or radical nephrectomy: conditional survival analysis. Urol Oncol 2014; 32: 43.e9.
15.
Harshman LC, Xie W, Bjarnason GA et al: Conditional survival of patients with metastatic renal-cell carcinoma treated with VEGF-targeted therapy: a population-based study. Lancet Oncol 2012; 13: 927.
16.
Thuret R, Sun M, Abdollah F et al: Conditional survival predictions after surgery for patients with penile carcinoma. Cancer 2011; 117: 3723.
AC C
EP
TE D
M AN U
SC
RI PT
1.
13
ACCEPTED MANUSCRIPT
Sun M, Abdollah F, Bianchi M et al: Conditional survival of patients with urothelial carcinoma of the urinary bladder treated with radical cystectomy. Eur J Cancer 2012; 48: 1503.
18.
Edge SB, Byrd DR, Compton CC et al eds: Urinary bladder. In: AJCC Cancer Staging Manual. 7th ed. New York, NY: Springer 2010; pp 497–505.
19.
Xylinas E, Kent M, Kluth L et al: Accuracy of the EORTC risk tables and of the CUETO scoring model to predict outcomes in non-muscle-invasive urothelial carcinoma of the bladder. Br J Cancer 2013; 109: 1460.
20.
Lamm DL: Getting the Most Out of Bacillus Calmette-Guérin for Treatment of Bladder Cancer. J Urol 2016; 195: 7.
21.
Yokomizo A, Kanimoto Y, Okamura T et al: Randomized Controlled Study of the Efficacy, Safety and Quality of Life of Low-Dose Bacillus Calmette-Guérin Instillation Therapy for Non-Muscle-Invasive Bladder Cancer. J Urol 2016; 195: 41.
22.
Ploussard G, Xylinas E, Lotan Y et al: Conditional Survival Nephroureterectomy for Upper Tract Carcinoma. Eur Urol 2015; 67: 803.
23.
Sylvester RJ, Oosterlinck W, Holmang S et al: Systematic Review and Individual Patient Data Meta-analysis of Randomized Trials Comparing a Single Immediate Instillation of Chemotherapy After Transurethral Resection with Transurethral Resection Alone in Patients with Stage pTa-pT1 Urothelial Carcinoma. Eur Urol 2015: doi 10.1016/j.eururo.2015.05.050 (in press).
24.
Lee CT, Barocas D, Globe DR et al: Economic and humanistic consequences of preventable bladder tumor recurrences in nonmuscle invasive bladder cancer cases. J Urol 2012; 188: 2114.
After
Radical
AC C
EP
TE D
M AN U
SC
RI PT
17.
14
ACCEPTED MANUSCRIPT
FIGURE LEGENDS Figure 1
Cumulative incidence curves for recurrence (black line) overlaid by conditional
RI PT
recurrence estimates at 6 months (red line), 12 months (green line), 24 months
Figure 2
SC
(blue line), 36 months (yellow line) and 48 months (pink line).
Cumulative incidence curves for progression (black line) overlaid by conditional
M AN U
progression estimates at 6 months (red line), 12 months (green line), 24 months
AC C
EP
TE D
(blue line), 36 months (yellow line) and 48 months (pink line).
15
ACCEPTED MANUSCRIPT
2-year conditional recurrence rates in relation to clinical and pathological characteristics.
35% 37% 0.65
29% 32% 0.69
22% 22% 0.46
16% 16% 0.60
1001 (77.5) 291 (22.5)
36% 35% 0.80
31% 31% 0.98
22% 20% 0.72
16% 18% 0.92
872 (67.5) 420 (32.5)
35% 40% 0.04
30% 32% 0.14
23% 20% 0.66
275 (21.3) 677 (52.4) 340 (26.3)
30% 37% 41% 0.03
27% 32% 33% 0.33
21% 22% 24% 0.67
718 (55.6) 574 (44.4)
29% 45% <0.01
24% 40% <0.01
1135 (87.8) 157 (12.2)
35% 44% 0.04
31% 31% 0.64
1138 (88.1) 154 (11.9)
165 (12.8) 601 (46.5) 526 (40.7)
13% 14% 0.54
8% 12% 0.71
13% 14% 0.95
11% 8% 0.62
17% 14% 0.20
13% 14% 0.09
8% 17% <0.01
16% 16% 17% 0.79
13% 14% 13% 0.75
11% 11% 10% 0.72
18% 28% <0.01
13% 21% 0.14
10% 18% 0.45
10% 11% 0.75
22% 23% 0.53
16% 16% 0.59
13% 14% 0.52
11% 9% 0.39
TE D
M AN U
SC
447 (34.6) 845 (65.4)
RI PT
n (%) -
EP
Cohort, n 2-yr conditional recurrence rate Age <65 yrs ≥65 yrs p-value Gender Male Female p-value Tumor stage pTa pT1 p-value Tumor grade G1 G2 G3 p-value Number of tumors Solitary Multiple p-value CIS Absent Present p-value Tumor size <3cm ≥3cm p-value Risk group Low Intermediate High p-value
Time elapsed since transurethral resection of the bladder cancer Baseline 6 months 12 months 24 months 36 months 48 months 1292 895 706 518 409 311 36% 31% 22% 16% 13% 11%
AC C
Table 1
35% 47% 0.03
30% 42% 0.05
21% 33% 0.03
16% 15% 0.46
13% 9% 0.66
10% 15% 0.27
19% 39% 38% <0.01
18% 35% 31% <0.01
15% 26% 20% 0.06
14% 19% 14% 0.55
12% 15% 13% 0.58
9% 9% 14% 0.38
ACCEPTED MANUSCRIPT
Table 2
24 months
36 months
13% 6% 13% 16%
26% 11% 27% 29% 16% 6% 17% 18%
36% 19% 39% 38% 28% 14% 31% 29% 15% 9% 18% 14%
41% 26% 44% 42% 33% 20% 37% 33% 22% 15% 26% 20% 9% 8% 10% 7%
48 months
60 months
45% 30% 49% 45% 38% 25% 43% 38% 28% 21% 33% 25% 16% 14% 19% 14% 9% 8% 10% 8%
47% 31% 51% 47% 41% 28% 45% 40% 31% 25% 35% 29% 20% 18% 22% 18% 13% 12% 15% 13% 5% 5% 5% 6%
TE D
M AN U
SC
12 months
EP
All LR Baseline IR HR All LR 6 months IR HR All LR 12 Recurrence months IR rate to a HR certain no. of All months from LR 24 TURB months IR HR All LR 36 months IR HR All LR 48 months IR HR
6 months
AC C
Total time without recurrence from TURB
RI PT
Conditional recurrence rates for patients with NMIBC who reach a certain time point without having developed recurrence. For example, if a patient stayed recurrence-free for 48 months, his chance of developing recurrent disease at 60 months is 5%. Data are presented for the overall cohort and according to risk group (LR, low-risk; IR, intermediate-risk; HR, high-risk).
ACCEPTED MANUSCRIPT
2-year conditional progression rates in relation to clinical and pathological characteristics.
1% 5% 0.06
1% 3% 0.15
1% 3% 0.86
0.5% 3% 0.68
1001 (77.5) 291 (22.5)
5% 4% 0.95
4% 3% 0.91
3% 5% 0.75
2% 4% 0.56
2% 3% 0.56
2% 2% 0.67
872 (67.5) 420 (32.5)
1% 11% <0.01
1% 9% <0.01
1% 10% <0.01
1% 6% <0.01
2% 3% 0.06
2% 3% 0.10
275 (21.3) 677 (52.4) 340 (26.3)
1% 3% 11% <0.01
1% 2% 10% <0.01
1% 2% 11% <0.01
1% 2% 6% 0.01
1% 3% 2% 0.62
2% 2% 3% 0.52
718 (55.6) 574 (44.4)
4% 6% 0.07
3% 5% 0.05
3% 5% 0.18
2% 3% 0.34
2% 2% 0.71
2% 2% 0.63
1135 (87.8) 157 (12.2)
4% 8% <0.01
3% 6% <0.01
3% 10% <0.01
2% 8% <0.01
2% 2% 0.34
2% 2% 0.33
RI PT
2% 4% 0.13
SC
4% 5% 0.30
TE D
M AN U
447 (34.6) 845 (65.4)
EP
Cohort, n 2-year conditional progression rate Age <65 yrs ≥65 yrs p-value Gender Male Female p-value Tumor stage pTa pT1 p-value Tumor grade G1 G2 G3 p-value Number of tumors Solitary Multiple p-value CIS Absent Present p-value Tumor size <3cm ≥3cm p-value Risk group Low Intermediate High p-value
Time elapsed since transurethral resection of the bladder cancer Baseline 6 months 12 months 24 months 36 months 48 months 1292 1018 916 761 639 525 4% 4% 4% 3% 2% 2%
n (%) -
AC C
Table 3
1138 (88.1) 154 (11.9)
4% 4% 0.92
4% 3% 0.99
4% 2% 0.76
3% 1% 0.82
2% 0% 0.77
2% 2% 0.37
165 (12.8) 601 (46.5) 526 (40.7)
2% 2% 9% <0.01
2% 1% 8% <0.01
1% 1% 8% <0.01
0% 2% 5% <0.01
0% 3% 2% 0.27
3% 1% 2% 0.33
ACCEPTED MANUSCRIPT
Table 4
Progression 12 months rates to a certain no. of months from TURB 24 months
36 months
48 months
36 months
1% 0% 1% 3%
2% 1% 1% 4% 1% 1% 0.4% 2%
4% 2% 2% 9% 3% 2% 1% 6% 2% 1% 1% 5%
6% 2% 2% 11% 5% 2% 1% 10% 4% 1% 1% 8% 2% 0% 0.3% 4%
48 months
M AN U
SC
24 months
TE D
6 months
12 months
EP
Baseline
All LR IR HR All LR IR HR All LR IR HR All LR IR HR All LR IR HR All LR IR HR
6 months
AC C
Total time without progression from TURB
RI PT
Conditional progression rates for patients with NMIBC who reach a certain time point without having developed progression. For example, if a patient stayed progression-free for 24 months, his chance of developing progression at 60 months is 3%. Data are presented for the overall cohort and according to risk group (LR, low-risk; IR, intermediate-risk; HR, high-risk).
7% 2% 3% 12% 6% 2% 3% 11% 5% 1% 2% 9% 3% 0% 2% 5% 1% 0% 1% 1%
60 months 7% 2% 4% 13% 6% 2% 4% 11% 5% 1% 3% 10% 3% 0% 3% 6% 2% 0% 3% 2% 1% 0% 1% 1%
ACCEPTED MANUSCRIPT
Table 5 Multivariable Fine-Gray’s competing risks regression models predicting recurrence (A) and progression (B). A 24 months
0.96 0.80-1.15 0.63
1.04 0.85-1.29 0.69
0.92 0.71-1.19 0.53
0.93 0.66-1.32 0.69
1.00 0.81-1.23 1.00
0.98 0.77-1.27 0.90
1.05 0.76-1.44 0.79
0.99 0.64-1.51 0.95
1.00
1.00
1.00
1.77 1.32-2.38 <0.01
1.80 1.29-2.50 <0.01
1.51 1.03-2.21 0.04
1.61 1.16-2.23 <0.01
1.73 1.19-2.51 <0.01
1.30 0.83-2.05 0.26
1.01 0.84-1.23 0.87
0.98 0.78-1.23 0.87
1.16 0.90-1.50 0.25
0.91 0.66-1.26 0.58
36 months
48 months
0.91 0.60-1.37 0.64
0.96 0.57-1.62 0.87
RI PT
12 months
SC
6 months
1.22 0.59-2.51 0.60
1.00
1.00
1.00
1.24 0.77-2.00 0.38
1.22 0.58-2.18 0.51
1.14 0.54-2.41 0.74
1.27 0.72-2.24 0.41
1.38 0.70-2.73 0.35
1.89 0.84-4.29 0.13
1.10 0.83-1.45 0.50
1.15 0.79-1.66 0.47
1.17 0.75-1.83 0.48
0.98 0.52-1.75 0.93
1.02 0.67-1.54 0.94
0.88 0.50-1.53 0.66
0.90 0.48-1.72 0.76
0.67 0.31-1.41 0.29
EP
TE D
M AN U
0.98 0.59-1.65 0.95
AC C
Age >65 yr SHR 95% CI p-value Male gender SHR 95% CI p-value Risk group Low SHR Intermediate SHR 95% CI p-value High SHR 95% CI p-value Mitomycin C SHR 95% CI p-value BCG SHR 95% CI p-value
Baseline
ACCEPTED MANUSCRIPT
B 12 months
24 months
36 months
48 months
1.17 0.72-1.91 0.52
1.41 0.82-2.45 0.22
1.67 0.90-3.07 0.10
1.71 0.82-3.58 0.16
1.14 0.51-2.55 0.75
1.27 0.47-3.41 0.63
0.95 0.57-1.61 0.86
0.92 0.52-1.63 0.77
0.85 0.46-1.59 0.62
0.74 0.34-1.59 0.44
1.25 0.43-3.70 0.68
1.22 0.34-4.34 0.76
1.00
1.00
1.00
1.00
1.00
1.00
1.54 0.53-4.45 0.43
1.37 0.46-4.07 0.57
1.66 0.48-5.78 0.42
2.13 0.47-9.59 0.32
1.97 0.43-8.97 0.38
1.29 0.26-6.42 0.75
4.19 1.47-11.97 <0.01
3.78 1.30-11.01 0.02
5.19 1.52-17.67 <0.01
4.78 1.05-21.65 0.02
2.52 0.50-12.61 0.26
2.41 0.45-12.93 0.30
1.48 0.90-2.44 0.13
1.36 0.77-2.39 0.29
1.03 0.55-1.91 0.94
1.05 0.47-2.35 0.91
1.13 0.41-3.12 0.82
0.85 0.26-2.81 0.79
0.55 0.30-0.99 0.04
0.50 0.25-0.97 0.04
0.55 0.27-1.12 0.09
0.26 0.08-0.84 0.02
0.11 0.02-0.83 0.03
N.A.
SC
M AN U
TE D
EP
RI PT
6 months
AC C
Age >65 yr SHR 95% CI p-value Male gender SHR 95% CI p-value Risk group Low SHR Intermediate SHR 95% CI p-value High SHR 95% CI p-value Mitomycin C SHR 95% CI p-value BCG SHR 95% CI p-value
Baseline
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT ABBREVIATIONS BC: bladder cancer BCG: Bacillus Calmette-Guérin
NMIBC: non-muscle invasive bladder cancer
AC C
EP
TE D
M AN U
TURB: transurethral resection of the bladder
SC
CS: conditional survival
RI PT
CIS: carcinoma in situ
1