Abstract
BACKGROUND: The association between curative treatment (CurTrt) and mortality in senior adults (≥70 years) with high-risk prostate cancer (PCa) is poorly documented. In a population-based cohort we report temporal trends in treatment and PCa-specific mortality (PCSM), investigating the association between CurTrt and mortality in senior adults with high-risk PCa, compared to findings in younger men (<70 years). METHODS: Observational study from the Cancer Registry of Norway. Patients with high-risk PCa were stratified for three diagnostic periods (2005-08, 2009-12 and 2013-16), age (<70, vs ≥70) and primary treatment (CurTrt: Radical prostatectomy (RP), Radiotherapy (RAD) vs no curative treatment (NoCurTrt)). Competing risk and Kaplan-Meier methods estimated PCSM and overall mortality (OM), respectively. Multivariable logistic regression models estimated odds for CurTrt, and multivariable Fine Gray and Cox regression models evaluated the hazard ratios for PCSM and OM. RESULTS: Of 19 763 evaluable patients, 54% were aged ≥70 years. Senior adults had more unfavorable PCa characteristics than younger men. Across diagnostic periods, use of CurTrt increased from 15% to 51% in men aged ≥70 and 65% to 81% in men aged < 70 years. With median five years follow-up, PCSM decreased in all patients (P < .05), in the third period restricted to senior adults. In all patients NoCurTrt was associated with three-fold higher 5-year PCSM and two-fold higher OM compared to CurTrt. CONCLUSIONS: In high-risk PCa patients, increased use of CurTrt, greatest in senior men, was observed along with decreased PCSM and OM in both senior and younger adults. CurTrt should increasingly be considered in men ≥70 years.