Presentation Authors: Jyoti Chouhan*, Ethan Matz, Marc Colaco, Amy Pearlman, James Lovato, Ryan Terlecki, Winston-Salem, NC
Introduction: The Centers for Disease Control (CDC) lists urothelial cancer as a tobacco-related malignancy (TRM). The National Health Interview Survey found declining rates of tobacco use among United States (U.S.) adults. Data from outside the U.S. suggests that, while risk of lung cancer mortality (LCM) may decrease after tobacco cessation, risk of bladder cancer mortality (BCM) appears stable. Also, racial disparities have been noted for incidence and mortality in TRM. We sought to determine the trend in BCM relative to LCM for whites, blacks and Hispanics over the past decade.
Methods: The CUDA study is a comprehensive review of all-cause mortality data from death records warehoused by the CDC from 2005-2016, and to our knowledge, the first to do so relative to urologic disease. The total number of deaths per year for this database ranged from 2.31-2.63 million. Subset analysis was performed using ICD-10 codes C67 (bladder cancer) and C33-34 (trachea/bronchus/lung cancer) and stratified by race and ethnicity. Time series regression analysis was performed for the entire interval.
Results: Over the study interval, BCM accounted for 0.61-0.69%, 0.36-0.43% and 0.38-0.43% of total mortality for whites, blacks, and Hispanics, respectively. A significant increase was seen in all groups; (p < 0.01, R2=0.87, F(1,11)=64.47, [white], P < 0.01, R2=0.62, F(1,11)=16.35 [Black], p=0.019, R2=0.44, F(1,11)=7.75 [Hispanic]. See Graphs 1a-c._x000D_
LCM accounted for 5.9-7.1%, 5.3-6.4%, and 3.3-4.0% of total mortality for whites, blacks and Hispanics, respectively. A significant decrease was seen in all groups; (p < 0.01, R2=0.89, F(1,11)=82.46, p < 0.01 [white], p < 0.01, R2=0.63, F(1,11)=17.65 [Black], p < 0.01, R2=0.82, F(1,11)=44.47, [Hispanic]. See Graphs 2a-c.
Conclusions: Over time, BCM and LCM show inverse trends. This may suggest differences in organ response to tobacco exposure, recovery following cessation, natural history over time, or attribution bias. This data is highly relevant for patient counseling. Future studies are needed, and mortality data will need to be viewed alongside incidence.