Racial and ethnic differences in risk of second primary cancers among prostate cancer survivors
- PDF / 479,822 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 59 Downloads / 189 Views
ORIGINAL PAPER
Racial and ethnic differences in risk of second primary cancers among prostate cancer survivors Diana R. Withrow1 · Sara J. Schonfeld1 · Rochelle E. Curtis1 · Lindsay M. Morton1 · Michael B. Cook1 · Eboneé N. Butler1 · Amy Berrington de González1 Received: 3 January 2020 / Accepted: 6 August 2020 © This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2020
Abstract Purpose Previous studies have shown an overall decreased risk of second cancers among prostate cancer survivors, but this has not been comprehensively examined by race/ethnicity. We conducted a retrospective cohort study of 716,319 one-year survivors of prostate cancer diagnosed at ages 35–84 during 2000–2015 as reported to 17 US Surveillance, Epidemiology and End Results (SEER) registries. Methods We estimated standardized incidence ratios (SIRs) for second primary non-prostate malignancies by race/ethnicity (non-Latino white, Black, Asian/Pacific Islander [API] and Latino), by Gleason, and by time since prostate cancer diagnosis. Poisson regression models were used to test heterogeneity between groups with the expected number as the offset. Results 60,707 second primary malignancies were observed. SIRs for all second cancers combined varied significantly by race/ethnicity: SIRwhite: 0.88 (95% confidence interval: 0.87–0.89), SIRLatino: 0.92 (0.89–0.95), S IRBlack: 0.97 (0.95–0.99), and SIRAPI: 1.05 (1.01–1.09) (p-heterogeneity 5 years) to explore the potential effects of medical surveillance bias whereby cancer patients are more likely than the general population to receive a subsequent cancer diagnosis because they are under closer medical scrutiny, assuming such a bias would have a greater impact in earlier years post diagnosis [12]. Stratification by Gleason score was used to explore effect modification by prostate cancer aggressiveness. Pathologists assign two Gleason values to each prostate cancer according to the level of differentiation of the cells. The cut-off we used (≤ 3 + 4, ≥ 4 + 3) corresponds to the American Joint Committee on Cancer Gleason Grade Groups (1–2 vs. 3 and up) and distinguishes well or mostly well-differentiated (less aggressive, better prognosis) glands from poorly or mostly poorly differentiated glands (more aggressive, poorer prognosis) [13, 14]. Analyses of Gleason score were restricted to 2004 and later when Gleason score was routinely collected. To test heterogeneity, we fitted Poisson regression models with observed number of events as the outcome and the expected number as the offset. Likelihood ratio tests were used to test for heterogeneity in SIRs. We adjusted for age at prostate cancer diagnosis in the Poisson models through stratification. We examined heterogeneity in SIRs both by race and within race. For all second cancers combined, we examined (a) heterogeneity in SIRs by race within strata defined by age at prostate cancer diagnosis (3 groups), Gleason score (3 groups including unknown), latency (2 groups), and time per
Data Loading...
