Abstract 6461: Excess risk due to smoking and modifying effect of quitting on bladder cancer incidence: the multiethnic cohort

Abstract

Abstract Background: Smoking is an established risk factor for urothelial cancer of the bladder (BC), attributed to over 50% of cases among men, however associations outside of European-descent populations are infrequently reported. Further, absolute risk of BC by smoking history has not yet been reported. Characterizing these associations will identify value of smoking cessation across non-European ancestry populations. Methods: We estimated the association between pack-years, years quit, and BC among 181,231 Multiethnic Cohort Study participants from five major racial/ethnic groups (African Americans [AA], European Americans [EA], Japanese Americans [JA], Latinos [LA], Native Hawaiians [NH]) 45-75 years old at enrollment (1993-1996). Smoking history was assessed by self-report at cohort entry and on a follow-up survey (2003-2008). Associations were estimated using an excess relative risk model to distinguish between variables that affect baseline risk and variables (e.g. years quit) that only affect excess risk due to smoking, while adjusting for sex, age, race/ethnicity, diabetes, body mass index, and cigarettes smoked per day. BC cases were identified through SEER cancer registries to 2017. We tested for heterogeneity of smoking variables by race/ethnicity and estimated incidence across age, race-ethnicity, pack-years, and years quit. Results: Over a mean 19 years of follow-up, 1,914 BC cases were identified. Age-standardized incidence rates of BC were highest among EA (51/100,000) and lowest among LA (24/100,000). Never smokers were most common among JA and LA (51%) and least common in other racial/ethnic groups (39-40%). In the full study sample, 50 pack-years smoked was associated with BC (HR=3.2 95% CI 2.7, 3.6). The association was heterogeneous by race/ethnicity (p=0.02) and statistically significant within each racial/ethnic group (p<0.05), except NH (p=0.07). Associations were strongest among EA (HR=4.0), followed by AA (HR=3.6), JA (HR=2.7), LA (HR=2.6), and NH (HR=2.4). There was no statistically significant heterogeneity of smoking cessation (p=0.08) or cigarettes smoked per day (p=0.82) by race/ethnicity. Each year quit resulted in a less than 1% decrease in the excess relative risk from smoking (p=0.12). The absolute risk models showed highest incidence at age 70 among EA current smokers (176/100,000) and lowest among AA never smokers (27/100,000). At age 70, current smokers across all ethnic groups had lower risk than EA who quit at age 45. Conclusion: Pack-years smoked is significantly associated with BC risk. EA and AA groups experience the greatest excess relative risk from smoking and NH the lowest. Reduction in excess relative risk following smoking cessation is weak, meaning the cumulative effect of pack-years smoked on BC risk cannot easily be attenuated with quitting. This contrasts with the risk reduction seen for lung and pancreatic cancer. Citation Format: David Bogumil, Victoria Cortessis, Daniel Stram, Veronica W. Setiawan, Christopher Haiman, Loic Le Marchand, Gertraud Maskarinec. Excess risk due to smoking and modifying effect of quitting on bladder cancer incidence: the multiethnic cohort. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6461.