Clinical Stage Migration and Survival for Renal Cell Carcinoma in the United States

Transcription

1 ava ilable at journa l homepage: euoncology.europeanurology.com Clinical Stage Migration and Survival for Renal Cell Carcinoma in the United States Hiten D. Patel *, Mohit Gupta, Gregory A. Joice, Arnav Srivastava, Ridwan Alam, Mohamad E. Allaf, Phillip M. Pierorazio James Buchanan Brady Urological Institute and Department of Urology, Johns Hopkins University School of Medicine, Baltimore, MD, USA Article info Article history: Accepted August 24, 2018 Associate Editor: Gianluca Giannarini Keywords: Kidney cancer Stage migration Renal cell carcinoma Small renal mass Abstract Background: The rising incidence of renal cell carcinoma (RCC) since the 1980s has been accompanied by stage migration toward early-stage (stage I) cancers. Stage migration drove an apparent increase in survival for RCC since the 1980s, but it is unclear whether it remains a contributor more recently. Objective: To determine whether clinical stage migration has persisted and the relative impact of stage migration versus improvements in treatment on survival for RCC. Design, setting, and participants: An epidemiologic assessment of stage migration and survival for patients at diagnosis of RCC ( ) across >1500 facilities in the National Cancer Database. Outcome measurements and statistical analysis: Proportion of patients over time was assessed by clinical stage at diagnosis via Cochran-Armitage chi-square tests and linear regression. Mortality data were assessed with the Kaplan-Meier method for 5-yr overall survival, Cox proportional hazards regression, and propensity score matching to differentiate the impact of treatment including systemic therapy from stage migration. Results and limitations: Greater diagnosis of clinical stage I disease (70%; p < 0.001) was observed, with decreased diagnosis of stage III (8%; p < 0.001) and stage IV (11%; p < 0.001) up to 2007 followed by stabilization through Tumor size continues to decrease for localized tumors (mean 0.22 cm stage I and 1.24 cm stage II, ). Histology demonstrated significant associations with stage. Five-year overall survival improved (67.9% [2004] to 72.3% [2010]) with gains in advanced RCC but not localized tumors. Models confirmed improved survival in recent years for stage IV patients. Systemic therapy was associated with improved survival (hazard ratio [ ], p < 0.001). National Cancer Database limitations apply. Conclusions: The proportion of patients presenting with stage I RCC has stabilized (70%), suggesting that stage migration may have ended. Localized tumors are detected with decreasing size, while advanced cancers have remained stable. Only 11% of patients now present with distant metastatic disease, but 5-yr overall survival is improving in recent years due to improved treatments rather than stage migration. Patient summary: In this study, we found that stage migration toward early-stage cancers has ended for renal cell carcinoma (RCC). However, improved treatment for advanced RCC appears to be responsible for improved survival in recent years European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. James Buchanan Brady Urological Institute, The Johns Hopkins University School of Medicine, 600 N. Wolfe Street/Marburg 134, Baltimore, MD 21287, USA. Tel.: ; Fax: address: hitenpatel@jhmi.edu (H.D. Patel) / 2018 European Association of Urology. Published by Elsevier B.V. All rights reserved.

2 2 1. Introduction The incidence of renal cell carcinoma (RCC) has been increasing worldwide for almost 30 yr due to a combination of environmental risk factors and incidental diagnoses attributed to widespread use of cross-sectional imaging [1]. Over patients are diagnosed with kidney cancer each year in the USA, and the increased incidence has been accompanied by clinical stage migration toward earlierstage tumors [2]. Data from the National Cancer Database (NCDB) indicated an increase in the proportion of patients presenting with ct1 RCC from 40% before 1993 to 60% through 2004 [2]. However, it is unknown if clinical stage migration has continued into recent years, which has implications for patient outcomes. Additionally, clinicians continue to quote historical rates of 30 33% for the proportion of patients who present with metastatic RCC, but the accuracy of these estimates is questionable in recent years with one cohort noting a potential dip to <20% in Sweden [3 6]. Estimates of stage and survival for RCC are often based on rates from the Surveillance, Epidemiology, and End Results program, which reports data on kidney and renal pelvis cancers. Therefore, publicly reported data combine upper tract transitional cell carcinomas with RCC, but the best current estimates suggest that 5-yr overall survival for kidney cancer has increased from 57% in the late 1980s to >70% in recent years; it remains uncertain how much of the increase is due simply to stage migration rather than improvements in providing effective treatment [5]. Management approaches for localized disease are highly effective, so despite implementation of minimally invasive technologies and active surveillance, patient prognosis and cancer control may be unchanged [7,8]. More promisingly, evidence suggests that patients with metastatic RCC in the targeted therapy era may have improved prognosis [9]. The objective of the present study was to determine whether stage migration has persisted since 2004 and whether survival has improved for patients diagnosed with RCC across >1500 Commission on Cancer facilities. Specifically, we assessed clinical stage at presentation, histologic subtypes, and relative impact on survival that may be attributed to improvements in RCC treatment rather than stage migration. 2. Patients and methods 2.1. Patient cohort The NCDB, jointly sponsored by the American Cancer Society and American College of Surgeons, captures information on the initial diagnosis of malignancy and first course of therapy for patients across >1500 Commission on Cancer facilities in the USA. Patients 18 yr of age diagnosed with a malignancy of the kidney or renal pelvis (no exclusion for history of prior malignancies) with data on tumor staging from 2004 to 2015 were evaluated. Of patients whose information had been captured, (4.61%) were excluded due to a primary transitional cell carcinoma diagnosis and (3.62%) due to other non-rcc histologies. Of the remaining patients with RCC histologies (clear cell, not otherwise specified, papillary, chromophobe, sarcomatoid features, or cyst-associated RCC), 4732 (1.64%) were unable to be subtyped further as ct1a/b (1734) or ct2a/b (2998), (5.19%) had missing data on clinical N stage, and 5398 (1.88%) had missing data on clinical M stage, resulting in a final included sample of patients Procedures and variables Clinical and pathologic staging for all patients was based on the AJCC seventh edition manual with stratification by standard stage groupings (I IV), and clinical relevant subgroups (ct1an0m0, ct1bn0m0, cn1, and cm1) were evaluated [10]. Additional variables included were age, sex, race, Charlson Comorbidity Index (CCI), tumor size, and treatments received. Stage migration was assessed over time based on clinical staging as well as tumor size migration within each clinical stage stratum. RCC histology was evaluated, and 5-yr overall survival was quantified for patients diagnosed in 2004 and The relative impact of time on survival was analyzed by stratifying patients according to clinical stage to determine the relative impact of treatment versus time within each stratum. Systemic therapy for stage IV patients was captured and classified by whether patients received systemic therapy, had contraindications, died too early to plan for treatment, or were recommended for therapy but did not receive it Statistical analysis Baseline demographics and clinical characteristics were tabulated by clinical stage. Stage migration trends were tested using the Cochran- Armitage chi-square test for trend and departure from a linear model as well as segmented linear regression. The Kaplan-Meier method estimated 5-yr overall survival, while Cox proportional hazards regression models were constructed to evaluate the impact of time on survival by clinical stage. Multivariable models were evaluated adjusting for age, sex, race, CCI, histology, and receipt of surgical therapy. To reduce bias due to changes in treatment patterns and patient characteristics over time, propensity score matching was employed by clinical stage to balance propensity scores between earlier ( ) and more recent ( ) years. One-to-one nearest-neighbor matching was performed without replacement and a caliper set to followed by Cox proportional hazards models for matched patients. Sensitivity analyses were performed by time period and in the subset receiving treatment to exclude patients dying within 30 d. Patients with stage IV RCC were further evaluated to adjust for receipt of systemic therapy. Statistical analyses were performed using STATA v.15.0 (STATA Corp., College Station, TX, USA; 2017). 3. Results 3.1. Stage migration Of patients included from 2004 to 2015, (47.5%) had stage I ct1a, (22.8%) stage I ct1b, (10.5%) stage II, (8.3%) stage III, and (11.0%) stage IV. Detailed demographics and clinical data are reported in Supplementary Table 1. Stage migration over the 12-yr period is displayed in Figure 1A, with statistically significant trends toward greater diagnosis of clinical stage I RCC (p < 0.001) and decreased diagnosis of clinical stage III (p < 0.001) and stage IV (p < 0.001) disease up to There was no significant change for stage II RCC (p = 0.441) or in the stage I T1b subset (p = 0.509), with stage

3 3 Fig. 1 Clinical stage at presentation for renal cell carcinoma showing (A) stage migration from 2004 to 2015 across Commission on Cancer accredited facilities in the USA and (B) distribution by histologic subtypes (National Cancer Database, ). I T1a alone accounting for the increase (p < 0.001). Linear regression models confirmed similar trends, with no significant change in the diagnosis of stage I (p = 0.650) or stage IV (p = 0.124) RCC after For patients receiving surgery, pathologic stage over time is displayed in Supplementary Figure 1. Tumor size across clinical stages has notably decreased from 2004 to 2015, with statistically significant trends for clinically localized RCC (stage I II) but not for more advanced tumors with a median of 7.50 cm for stage III and 8.20 cm for stage IV RCC (Table 1). The mean change in tumor size was 0.22 cm for stage I tumors and 1.24 cm for stage II tumors. Clinical stage distribution by histology demonstrated expected patterns for clear cell, papillary, and chromophobe RCC with early stage at diagnosis for patients with chromophobe RCC (Fig.1B). Over 95% of patients with cyst-associated RCC presented with localized disease compared with >50% with sarcomatoid features having stage IV disease Survival Median follow-up for survival was 42.7 mo (interquartile range [IQR] ) for the overall cohort, 83.8 mo (IQR ) for patients diagnosed in 2004, and 62.8 (IQR ) for patients diagnosed in Figure 2A displays overall survival through 10 yr after diagnosis. Between 2004 and 2010, 5-yr overall survival improved from 67.9% to 72.3% (Fig. 2B). Confidence intervals in Figure 2B suggest that a notable gain in survival was observed for cm1 patients. Multivariable Cox proportional hazards models showed associations between age, sex, CCI, and clinical stage with survival (Supplementary Table 2). Papillary RCC had worse survival compared with clear cell RCC after adjusting for differences in stage distribution and other variables, while chromophobe RCC demonstrated the best prognosis. While cyst-associated RCC presented at an earlier stage, the prognosis was equivalent to that of clear cell RCC (hazard ratio [HR] [ ], p = 0.829). Most notably, diagnosis in recent years remained a statistically significant predictor of improved survival despite adjustment for stage distribution. On stage-stratified analyses, improvements in prognosis over time were limited to the stage IV setting, with no improvement in prognosis for stage I III patients based on standard multivariable modeling (Table 2). Stage I T1a patients had slightly worse prognosis over time in the multivariable model, but no difference was noted after propensity score matching was applied to account for changes in characteristics and treatment patterns before and after A sensitivity analysis limited to patients diagnosed in 2010 and earlier allowing for longer follow-up also demonstrated no difference in survival over time (data not shown). Propensity score matching noted improved survival for both stage III and stage IV RCC patients in recent years. Further evaluation of stage IV RCC patients indicated that patients receiving systemic therapies experienced improved survival (HR [ ], p < 0.001; Table 3). Patients with contraindications to therapy or those who died before therapy could be planned or initiated had worse survival. A subset of patients who were eligible Table 1 Change in tumor size at presentation for renal cell carcinoma from 2004 to 2015 (National Cancer Database, ) Clinical stage Median (mean), cm IQR, cm Median (mean), cm IQR, cm Coefficient, mm/yr a p value Stage I (T1a) 2.70 (2.75) (2.67) <0.001 Stage I (T1b) 5.10 (5.52) (5.18) <0.001 Stage I (All) 3.50 (3.70) (3.48) <0.001 Stage II 9.00 (10.91) (9.67) <0.001 Stage III 7.50 (8.30) (8.22) Stage IV 8.00 (9.05) (9.10) IQR = interquartile range. a Simple linear regression models.

4 4 Fig. 2 Survival implications for renal cell carcinoma shown for (A) the overall cohort via Kaplan-Meier survival curves stratified by clinical stage and (B) 5-yr survival probabilities comparing patients diagnosed in 2004 to patients diagnosed in 2010 with 95% confidence intervals (National Cancer Database, ). for and recommended to receive systemic therapy but did not receive it for various reasons, including patient or family refusal, experienced increased mortality. Notably, the year of diagnosis remained a statistically significant predictor of survival after adjustment for systemic therapy. 4. Discussion After over 25 yr, RCC stage migration has stabilized and ended according to Commission on Cancer facilities captured by the NCDB in the USA. Early evidence suggested that stage migration began before 1983 when T2 tumors (>7 cm) were more common than either T1a or T1b tumors [11]. Owing to environmental risk factors and use of crosssectional imaging, the increased incidence of RCC was accompanied by drastic stage migration [1,2,12]. The results of the present study demonstrate that stage migration leveled off after 2007, with a largely stable distribution in recent years. While the absolute increase in ct1 tumors was fairly small (4.2%) relative to the prior decade, a slow reduction in localized tumor size has continued and may be relevant among ct1a patients despite stability in AJCC stage. In 2015, 71% of patients are diagnosed with clinically localized T1 RCC carrying a favorable 5-yr overall survival prognosis of 82.4%. Only 11% of patients now present with distant metastasis, and <2% have clinical nodal metastases without evidence of distant disease dramatically lower than previously established rates of 30% [3 5]. The role of imaging in this phenomenon is undebatable, as localized tumors continue to present at smaller size and advanced RCC continues to present as relatively large tumors. Therefore, the next advancement in RCC diagnostics and early detection may lie in biomolecular testing rather than imaging. While survival has improved over time when considering all RCC patients, the primary benefit was observed in advanced RCC (stage III IV), with 5-yr survival increasing from 9.8% in 2004 to 13.2% in 2010 for patients with distant metastatic disease. The results indicate that stage migration no longer contributes to improvements in survival for RCC, and additional gains reflect improvements in advanced treatment options. There are several well-established diagnostic and management dilemmas for the incidentally diagnosed small renal mass (stage I T1a RCC). Unfortunately, no preoperative demographic or imaging characteristics can reliably distinguish between malignant and benign masses [13]. Management options for localized disease are highly effective and have essentially remained unchanged in Table 2 Cox proportional hazards regression models assessing change in survival prognosis over time by clinical stage at presentation for renal cell carcinoma (National Cancer Database, ). Univariable Multivariable a Propensity score matching b HR 95% CI p value HR 95% CI p value HR 95% CI p value Low High Low High Low High Survival over time Survival over time Survival over time Stage I (T1a; per year) < Stage I (T1b; per year) Stage II (per year) Stage III (per year) < Stage IV (per year) < < <0.001 CI = confidence interval; HR = hazard ratio. a Multivariable models adjusted for age, sex, race, Charlson Comorbidity Index, histology, and receipt of surgical therapy. b Propensity score matching performed with one-to-one matching without replacement (caliper 0.001) in a stage-by-stage fashion to balance propensity scores for patients diagnosed in earlier ( ) and later ( ) years.

5 5 Table 3 C in prognosis for clinical stage IV renal cell carcinoma over time accounting for treatment with systemic therapy in a multivariable Cox proportional hazards regression model (National Cancer Database, ). Multivariable a HR 95% CI p-value Low High Age (yr) <45 REF Sex Male REF Female CCI 0 REF < < <0.001 Histology ccrcc REF RCC NOS <0.001 prcc <0.001 chrcc Sarcomatoid <0.001 Cyst associated Systemic therapy None REF Yes <0.001 Contraindicated <0.001 Died too early <0.001 Recommended but not received Unknown <0.001 Year <0.001 CCI = Charlson Comorbidity Index; ccrcc = clear cell RCC; chrcc = chromophobe RCC; CI = confidence interval; HR = hazard ratio; NOS = not otherwise specified; prcc = papillary RCC; RCC = renal cell carcinoma. a Multivariable models were also controlled for race and receipt of surgical therapy. terms of cancer control for several decades, leading to a greater emphasis on renal functional outcomes and quality of life [14,15]. Minimally invasive techniques have been employed to reduce morbidity, but they have largely preserved rather than improved upon the disease-modifying impact of open surgery. Active surveillance is also an increasingly utilized management option with thresholds for intervention to reduce morbidity while optimizing outcomes [16,17]. Therefore, the lack of improvement we observed in survival over time is not surprising for stage I and II patients. In addition, surgery remains the mainstay of treatment for stage II III patients, given the relatively high adverse event rate for adjuvant therapies with limited to no benefit in recently reported trials [18 20]. The improvement in survival that we demonstrated for cm1 patients over time is likely due to recent advances in systemic therapy starting with the approval of sunitinib in 2006, which was quickly followed by sorafenib and temsirolimus thereafter [4,21,22]. Given rapid progress in the area since then, and approval of nivolumab plus ipilimumab in April 2018 as frontline agents for intermediate- and poor-risk cm1 RCC, it is likely that the trend will continue in the near future. Our analysis also confirmed the independent prognostic value of histology validating a prior study that suggested that papillary RCC is associated with worse survival than clear cell RCC in the metastatic setting but with better survival for localized disease [23]. Importantly, we were able to adjust for receipt of systemic therapy, which was not considered in the German cohort by Steffens et al. [23], but it is also thought that papillary RCC is more resistant to systemic or immunotherapy agents than clear cell RCC. The findings emphasize a need for more effective systemic agents to treat advanced papillary RCC, which has often been limited or completely excluded from trials reported in the past 12 yr [4,21,22,24]. Additionally, we confirm sarcomatoid features as an independent marker of adverse outcomes and that cyst-associated RCC carries a similar prognosis to clear cell RCC on a stage-by-stage basis. Our study should be considered in the context of a few potential limitations. While NCDB captures the majority of cancer diagnoses in the USA, it is not a population-based sample to derive incidence and prevalence of RCC. Information on patients with benign pathology is also not generally captured when tissue is available, and symptomatology is not recorded. However, the study focuses on stage migration and survival estimates, which are robustly captured by NCDB and have not been evaluated since 2004, prior to the targeted therapy era and widespread use of robotic surgery [25]. The estimates are also conservative as it is possible that NCDB could underestimate ct1a disease. Propensity score analyses were employed to aid in balancing differences in patient characteristics in earlier and more recent years, but matching based on propensity scores is not without limitations, including the possibility of residual confounding. Additional NCDB sites have been included over time, and patients may obtain care at more than one site. Some data elements may also be limited or variable across different sites, including comorbidities captured and advanced treatments received for RCC outside of the first course of care. Cancer outcome of recurrence and cancerspecific survival are not available in NCDB. Lastly, the analysis depends on the accuracy of clinical and pathologic staging across all Commission on Cancer sites. Collaborative Stage data elements and physician-recorded AJCC TNM elements have been captured since 2004, which have likely improved reliability and consistency compared with data prior to 2004 [26]. Despite potential limitations, our findings constitute the largest assessment of RCC stage migration to date and quantify improvements in survival in the USA. 5. Conclusions The proportion of patients presenting with early-stage (stage I) RCC has stabilized in the USA at about 70% since 2007, suggesting that stage migration may have ended. Localized tumors continue to be detected at a smaller initial size, while tumor size for stage III and IV patients has remained stable, suggesting a need for improved detection of advanced RCC. Only 11% of patients now present with

6 6 distant metastatic disease, but 5-yr overall survival is improving in recent years likely due to improved treatments rather than stage migration. Author contributions: Hiten D. Patel had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Patel, Pierorazio. Acquisition of data: Patel, Gupta, Allaf, Pierorazio. Analysis and interpretation of data: Patel, Gupta, Joice, Srivastava, Alam, Allaf, Pierorazio. Drafting of the manuscript: Patel, Gupta, Joice, Srivastava, Alam, Allaf, Pierorazio. Critical revision of the manuscript for important intellectual content: Patel, Gupta, Joice, Srivastava, Alam, Allaf, Pierorazio. Statistical analysis: Patel. Obtaining funding: None. Administrative, technical, or material support: None. Supervision: Allaf, Pierorazio. Other: None. Financial disclosures: Hiten D. Patel certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, 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. Acknowledgments: The NCDB is a joint project of the Commission on Cancer (CoC) of the American College of Surgeons and the American Cancer Society. The CoC s NCDB and the hospitals participating in the CoC NCDB are the sources of the deidentified data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi: /j.euo References [1] Chow WH, Dong LM, Devesa SS. Epidemiology and risk factors for kidney cancer. Nat Rev Urol 2010;7: [2] Kane CJ, Mallin K, Ritchey J, Cooperberg MR, Carroll PR. Renal cell cancer stage migration: analysis of the National Cancer Data Base. Cancer 2008;113: [3] Ball MW. Surgical management of metastatic renal cell carcinoma. Discov Med 2017;23: [4] Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa in metastatic renal-cell carcinoma. N Engl J Med 2007;356: [5] Choueiri TK, Motzer RJ. Systemic therapy for metastatic renal-cell carcinoma. N Engl J Med 2017;376: [6] Thorstenson A, Bergman M, Scherman-Plogell AH, et al. Tumour characteristics and surgical treatment of renal cell carcinoma in Sweden : a population-based study from the National Swedish Kidney Cancer Register. Scand J Urol 2014;48: [7] Patel HD, Kates M, Pierorazio PM, et al. Survival after diagnosis of localized T1a kidney cancer: current population-based practice of surgery and nonsurgical management. Urology 2014;83: [8] Pierorazio PM, Johnson MH, Patel HD, et al. Management of renal masses and localized renal cancer: systematic review and metaanalysis. J Urol 2016;196: [9] Patel HD, Gorin MA, Gupta N, et al. Mortality trends and the impact of lymphadenectomy on survival for renal cell carcinoma patients with distant metastasis. Can Urol Assoc J 2016;10: [10] Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, editors. American Joint Committee on Cancer Staging manual. ed. 7. New York, NY: Springer; [11] Hollingsworth JM, Miller DC, Daignault S, Hollenbeck BK. Rising incidence of small renal masses: a need to reassess treatment effect. J Natl Cancer Inst 2006;98: [12] Chow WH, Devesa SS, Warren JL, Fraumeni Jr JF. Rising incidence of renal cell cancer in the United States. JAMA 1999;281: [13] Pierorazio PM, Patel HD, Johnson MH, et al. Distinguishing malignant and benign renal masses with composite models and nomograms: a systematic review and meta-analysis of clinically localized renal masses suspicious for malignancy. Cancer 2016;122: [14] Patel HD, Pierorazio PM, Johnson MH, et al. Renal functional outcomes after surgery, ablation, and active surveillance of localized renal tumors: a systematic review and meta-analysis. Clin J Am Soc Nephrol 2017;12: [15] Patel HD, Riffon MF, JoiceGA, et al. A prospective, comparative studyof quality of life among patients with small renal masses choosing active surveillance and primary intervention. J Urol 2016;196: [16] Uzosike AC, Patel HD, Alam R, et al. Growth kinetics of small renal masses on active surveillance: variability and results from the DISSRM Registry. J Urol 2018;199: [17] Jang A, Patel HD, Riffon M, et al. Multiple growth periods predict unfavourable pathology in patients with small renal masses. BJU Int 2018;121: [18] Haas NB, Manola J, Uzzo RB, et al. Adjuvant sunitinib or sorafenib for high-risk, non-metastatic renal-cell carcinoma (ECOG-ACRIN E2805): a double-blind, placebo-controlled, randomized, phase III trial. Lancet 2016;387: [19] Ravaud A, Motzer RJ, Pandha HS, et al. Adjuvant sunitinib in highrisk renal-cell carcinoma after nephrectomy. N Engl J Med 2016;375: [20] Motzer RJ, Haas NB, Donskov F, et al. Randomized phase III trial of adjuvant pazopanib versus placebo after nephrectomy in patients with localized or locally advanced renal cell carcinoma. J Clin Oncol 2017;35: [21] Escudier B, Eisen T, Stadler WM, et al. Sorafenib in advanced clearcell renal-cell carcinoma. N Engl J Med 2007;356: [22] Hudes G, Carducci M, Tomczak P, et al. Temsirolimus, interferon alfa, or both for advanced renal-cell carcinoma. N Engl J Med 2007;356: [23] Steffens S, Janssen M, Roos FC, et al. Incidence and long-term prognosis of papillary compared to clear cell renal cell carcinoma a multicentre study. Eur J Cancer 2012;48: [24] Escudier B, Tannir NM, McDermott DF, et al. CheckMate 214: efficacy and safety of nivolumab plus ipilimumab vs sunitinib for treatment-naïve advanced or metastatic renal cell carcinoma, including IMDC risk and PD-L1 expression subgroups. Ann Oncol 2017;28(Suppl. 5):v [25] Patel HD, Mullins JK, Pierorazio PM, et al. Trends in renal surgery: robotic technology is associated with increased use of partial nephrectomy. J Urol 2013;189: [26] Bilimoria KY, Stewart AK, Winchester DP, Ko CY. The National Cancer Data Base: a powerful initiative to improve cancer care in the United States. Ann Surg Oncol 2008;15: