LAUSR

INTRODUCTION Clinical outcomes for urothelial cancer treatments are equivalent whether patients are treated with surgery or trimodality treatment, with survival outcomes reported to be between 50% and 60% regardless of treatment. Although there are subtle differences between international guidelines, the European Society for Medical Oncology, European Association of Urology, American Society of Clinical Oncology, and American Urological Association all state that bladder preservation should be considered an alternative to radical cystectomy for suitable patients. National Institute for Health and Care Excellence guidelines from the United Kingdom to our knowledge are the first to state that patients should be counseled by both a surgeon specializing in radical cystectomy and an oncologist specializing in bladder preservation so that the individual can make an informed decision. In the absence of grade A evidence based on a randomized clinical trial comparing radical cystectomy with bladder preservation, the best evidence to date comes from meta-analysis, which has found that bladder preservation with trimodality treatment is superior to that of radical cystectomy with or without chemotherapy, with a hazard ratio in favor of trimodality treatment of 1.22. The challenge with all of these nonrandomized comparative studies is the heterogeneity of the populations being compared in terms of age, comorbidities, and performance status. The recent US National Cancer Data Base (NCDB) update from 2004 to 2013 has resulted in an increase in studies analyzing the database, 2 of which recently were published in Cancer. The study by Cahn et al compared overall survival (OS) in patients undergoing radical cystectomy (22,680 patients) with that of patients undergoing bladder preservation therapy (9620 patients). Patients treated with radiotherapy (RT) were stratified by those receiving radical RT alone (dose> 50 grays; 2540 patients) or chemoradiotherapy (1489 patients). Improved OS was observed for those patients receiving surgery in all age groups. Propensity score matching among the populations did not eliminate the difference. However, with more vigorous modeling, the magnitude of the difference in OS became smaller. In the study by Korpics et al, elderly patients (defined as those aged>80 years) were analyzed for OS within a comparison of RT versus chemoradiotherapy. In this study, 739 patients were found to have received RT alone and 630 patients received chemoradiotherapy. The median follow-up was 21 months and the 2-year OS rate was 42% versus 56% (P< .0001) when comparing RT with chemoradiotherapy. Multivariate analysis demonstrated that chemoradiotherapy and a higher RT dose were associated with improved OS. After propensity score matching, a significant benefit of chemoradiotherapy remained. There are inherent difficulties in using the NCDB due to its retrospective nature and the lack of pertinent information that can be associated with outcomes. In particular, the lack of data regarding performance status, specific chemotherapy drugs used, and sequencing of chemotherapy and RT (sequential vs concurrent) is significant. Given the known discrepancies between surgical and nonsurgical candidates in the United States, with patients treated with radical cystectomy acknowledged as being younger, fitter, and with fewer medical comorbidities, OS may not be the optimal outcome when comparing treatments. Many groups advocate the reporting of disease-specific survival as a more objective outcome in cases in which the disease occurs predominantly in the elderly. However, these data are unavailable in the NCDB. The