Objective: To assess the extent of early mortality and its temporal course after prostatectomy and radiotherapy in the general population. Methods: Data from the Surveillance, Epidemiology, and End Results (SEER) database and East German epidemiologic cancer registries were used for the years 2005-2013. Metastasized cases were excluded. Analyzing overall mortality, year-specific Cox regression models were used after adjusting for age (including age squared), risk stage, and grading. To estimate temporal hazards, we computed year-specific conditional hazards for surgery and radiotherapy after propensity-score matching and applied piecewise proportional hazard models. Results: In German and US populations, we observed higher initial 3-month mortality odds for prostatectomy (USA: 9.4, 95% CI: 7.8-11.2; Germany: 9.1, 95% CI: 5.1-16.2) approaching the null effect value not before 24-months (estimated annual mean 36-months in US data) after diagnosis. During the observational period, we observed a constant hazard ratio for the 24-month mortality in the US population (2005: 1.7, 95% CI: 1.5-1.9; 2013: 1.9, 95% CI: 1.6-2.2) comparing surgery and radiotherapy. The same was true in the German cohort (2005: 1.4, 95% CI: 0.9-2.1; 2013: 3.3, 95% CI: 2.2-5.1). Considering low-risk cases, the adverse surgery effect appeared stronger. Conclusion: There is strong evidence from two independent populations of a considerably higher early to midterm mortality after prostatectomy compared to radiotherapy extending the time of early mortality considered by previous studies up to 36-months.
Frontiers in oncology. 2020 Jan 17*** epublish ***
Daniel Medenwald, Dirk Vordermark, Christian T Dietzel
Department of Radiation Oncology, University Hospital Halle (Saale), Halle (Saale), Germany.