A 25-year Period Analysis of Other-cause Mortality in Localized Prostate Cancer

Other-cause mortality should have decreased over time in patients with prostate cancer as patient selection for diagnosis and treatment may have improved. We tested this hypothesis in 367,884 patients with prostate cancer treated with radical prostatectomy or radiotherapy within the Surveillance, Epidemiology, and End Results database and observed important other-cause mortality reduction over the last 25 years. Background: We examined the changes over time in other-cause mortality (OCM) rates in patients with clinically localized prostate cancer (PCa) as an indicator of patient selection. Patients and Methods: Within the Surveillance, Epidemiology, and End Results database (1987-2011), we identified patients with PCa treated with either radical prostatectomy (RP) (n = 230,969; 62.8%) or external beam radiation therapy (EBRT) (n = 136,915; 37.2%). Temporal trends and multivariable Cox regression analyses assessed OCM at 5 years using stratification according to year of diagnosis (1987-1991 vs. 1992-1996 vs. 1997-2001 vs. 2002-2006 vs. 2007-2011), age group, and ethnicity. Results: In patients who had undergone RP, the OCM rates at 5 years of follow-up decreased over time from 7.9% to 2.4% (slope, -0.25%/year) versus from 15.2% to 9.9% after EBRT (slope, -0.29%/year). The greatest decrease in 5-year OCM rates over time was recorded in patients >= 75 years (16.0%-12.0%; slope, -0.25%/year), followed by younger age categories (70-74 years, -0.21%/year; 65-69 years, -0.17%/year; 60-64 years, -0.10%/year; < 60 years, -0.07%/year), as well as in African-American men (11.0%-5.1%; slope, -0.32%/year), followed by Caucasian (7.6%-3.4%; slope, -0.21%/year) and Hispanic men (7.0%-3.1%; slope, -0.20%/year; all P < .001), as corroborated in multivariable Cox regression models. Conclusions: OCM rates were highest in oldest individuals and in African-American men. In both groups, an important 5-year OCM reduction over the 25-year study span was recorded. Nonetheless, these 2 patient groups may still represent the ideal target for better patient selection based on OCM considerations, because their most recent OCM rates exceeded those of, respectively, younger and Caucasian patients.