Utilization and predictors of neoadjuvant chemotherapy for muscle-invasive bladder cancer in the Veterans Health Administration.

Abstract

4594 Background: The survival benefit with neoadjuvant chemotherapy for bladder cancer was established in 2003. However, adoption of this paradigm in clinical practice has been slow. We explored the use of neoadjuvant chemotherapy and identified predictors of its use in a contemporary cohort in the Veterans Health Administration (VA). Methods: Using the national VA Clinical Cancer Registry, all patients diagnosed with clinical stage T2-4, N0 or Nx, M0 bladder cancer from 1997 to 2007 were stratified into surgically (radical cystectomy [RC], n=1,211) and nonsurgically managed groups (n=2,125). Receipt of neoadjuvant chemotherapy was defined as chemotherapy treatment up to 6 months before RC as well as initial treatment only with chemotherapy (without subsequent surgery or radiation) in the nonsurgical group. Temporal trends in neoadjuvant chemotherapy use were evaluated with a chi square test. Predictors of neoadjuvant chemotherapy were examined using a multivariable logistic regression model incorporating demographic, socioeconomic, comorbid, pathologic and hospital factors. Results: 6.3% and 8.3% of patients received neoadjuvant chemotherapy in the surgical and non-surgical group, respectively. Analysis of temporal trends in chemotherapy use demonstrated an increase in neoadjuvant chemotherapy use over time (p<0.0001); from 3% (2003 and before) to 14% annually (2007). On multivariable analysis of both groups, older age and more recent diagnosis were predictive of neoadjuvant chemotherapy use (Table). Other covariates did not predict receipt of neoadjuvant chemotherapy. Conclusions: While overall use of neoadjuvant chemotherapy in the VA population with bladder cancer remains low, use thereof is slowly increasing, with a more recent diagnosis most strongly predicting its use. [Table: see text]