Prostate Cancer Death After Radiotherapy or Radical Prostatectomy: A Nationwide Population-based Observational Study.

ABSTRACT

BACKGROUND: There are no conclusive results from randomized trials on radiotherapy (RT) versus radical prostatectomy (RP) for prostate cancer. Numerous observational studies have suggested that RP is associated with a lower risk of prostate cancer death, but whether results have been biased due to limited adjustments for confounding factors is unknown. OBJECTIVE: To compare the risk of prostate cancer death after RT versus RP. DESIGN, SETTING, AND PARTICIPANTS: Nationwide population-based observational study of men in the Prostate Cancer data Base Sweden 3.0 who had undergone RT or RP between 1998 and 2012. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Prostate cancer deaths were compared. Hazard ratios (HRs) were calculated in Cox regression models, including clinical T stage, M stage, Gleason grade group, serum levels of prostate-specific antigen, proportion of biopsy cores with cancer, mode of detection, comorbidity, age, educational level, and civil status. Period analysis with left truncation was performed. RESULTS AND LIMITATIONS: Primary treatment was RT or RP for 41 503 men. Treatment effect was associated with disease severity. In univariate analysis of RT versus RP, risk of prostate cancer death was higher after RT-low- and intermediate-risk cancer, HR 1.82 (95% confidence interval [CI] 1.53-2.16), and high-risk cancer, HR 1.57 (95% CI 1.33-1.85). After full adjustment in period analysis, this difference between the treatments was attenuated-low- and intermediate-risk cancer, HR 1.24 (95% CI 0.97-1.58), and high-risk cancer, HR 1.03 (95% CI 0.81-1.31). Confounding remained due to nonrandom allocation to treatment. CONCLUSIONS: In comparison with previous studies, the difference in prostate cancer mortality after RT and RP was much smaller. PATIENT SUMMARY: The difference in prostate cancer mortality after contemporary radiotherapy and radical prostatectomy was small in contrast to previous studies, indicating that potential side effects should be more emphasized when selecting treatment.