Lymph node counts are valid indicators of the quality of surgical care in bladder cancer: a population-based study

To describe lymph node counts in routine clinical practice and evaluate their association with outcomes to explore its utility as a quality indicator.

Electronic records of treatment and surgical pathology reports were linked with the population-based Ontario Cancer Registry to identify all patients who underwent cystectomy between 1994 and 2008. Temporal trends were described over 3 periods: 1994 to 1998, 1999 to 2000, and 2004 to 2008. Multivariate generalized linear regression analysis was used to determine the factors associated with the use of pelvic lymph node dissection (PLND). A Cox proportional hazards regression model was used to explore the associations between PLND and survival.

The study population included 2,802 patients. Use of PLND (50%, 62%, and 85%, correspondingly), median node yield (5, 6, and 9, correspondingly), and node density (56%, 50%, and 39%, correspondingly) all improved over the study periods, 1994 to 1998, 1999 to 2000, and 2004 to 2008 (P<0.001). In multivariate analysis, factors associated with not having PLND include advanced age, female sex, lower socioeconomic status, low surgeon volume, and partial cystectomy. In adjusted analyses, patients who did not receive a PLND had inferior overall (hazard ratio = 1.26, 95% CI: 1.15-1.38) and cancer-specific (hazard ratio = 1.23, 95% CI: 1.11-1.36) survival. Node yield, as well as density, was also associated with long-term survival.

There is significant variation in use and quality of PLND at cystectomy in routine practice. Node counts are independently associated with long-term survival, and this association is persistent despite adjustment for provider-related variables. These results suggest that lymph node counts are a valid quality indicator of surgical care of muscle-invasive bladder cancer.