Douglas A. Mata, Susan Groshen, Friedrich-Carl von Rundstedt, Donald G. Skinner, Walter M. Stadler, Richard J. Cote, John P. Stein, Seth P. Lerner

Previous studies have shown that variability in surgical technique can affect the outcomes of cooperative group trials. We analyzed measures of surgical quality and clinical outcomes in patients enrolled in the p53-MVAC trial. We performed a post-hoc analysis of patients with pT1-T2N0M0 urothelial carcinoma of the bladder following radical cystectomy (RC) and bilateral pelvic and iliac lymphadenectomy (LND). Measures of surgical quality were examined for associations with time to recurrence (TTR) and overall survival (OS). We reviewed operative and/or pathology reports for 440 patients from 35 sites. We found that only 31% of patients met all suggested trial eligibility criteria of having ≥15 lymph nodes identified in the pathologic specimen (LN#) and having undergone both extended and presacral LND. There was no association between extent of LND, LN#, or presacral LND and TTR or OS after adjustment for confounders and multiple testing. We demonstrated that there was substantial variability in surgical technique within a cooperative group trial. Despite explicit entry criteria, many patients did not undergo per-protocol LNDs. While outcomes were not apparently affected, it is nonetheless evident that careful attention to study design and quality monitoring will be critical to successful future trials. J. Surg. Oncol. © 2015 Wiley Periodicals, Inc.