Surgery versus radiation for clinically positive nodal prostate cancer in an other cause mortality risk weighted cohort.

Abstract

This study examined cancer control metrics between surgery and radiation for clinically positive nodal prostate cancer in an other-cause mortality weighted cohort, to circumvent limitations in previous studies. The Surveillance, Epidemiology, and End Results Research Plus database was queried to identify men with clinically positive nodal prostate cancer at diagnosis between 2004 and 2017 who were treated with surgery or radiation. A competing-risks regression model was used to calculate the 10-year other-cause mortality risk using available covariates, including treatment type. Inverse probability of treatment weighting was then used to balance covariates, including other-cause mortality risk. Then, competing-risks cumulative incidence curves and multivariable models, which were weighted on the calculated other-cause mortality risk, were used to examine the impact of treatment type on cancer-specific mortality, after accounting for covariates. 4739 patients underwent surgery whereas 1039 underwent radiation. The median follow-up was 4.7years (2.6-8.2). Other-cause mortality was statistically different between treatment arms in the unweighted cohort (Gray's p = 0.005), but that difference disappeared in the weighted cohort (Gray's p = 0.2). At 10years, the cancer-specific mortality rate was 27.6% (22.2-33.9) for radiation versus 18.1% (16.2-20.3) for surgery (p < 0.001). On competing-risks multivariable analysis, radiation had 1.86-fold (95% CI 1.69-2.12) higher hazard likelihood from one year to the next compared to surgery (p < 0.001). Clinically positive nodal patients treated with radiation fare worst cancer-specific mortality than those that underwent surgery, using calculated other-cause mortality risk.