Prostate cancer–specific mortality and competing causes of mortality among elderly men after local therapy for prostate cancer.

103 Background: The benefit of definitive local therapy among elderly patients (> 65 years) with localized prostate cancer (PC) is uncertain, particularly for those with comorbid illness. Despite this uncertainty, the majority of these men currently receive local therapy. We analyzed the risk of prostate cancer-specific mortality (PCSM) relative to competing causes of mortality (CCM), stratified by disease severity and comorbidity, among contemporary men treated at two high-volume hospitals Methods: Between 1995-2005, 4237 consecutive men aged 65 years or older were managed by radical prostatectomy (N = 1634), external-beam radiotherapy (N = 1570), or brachytherapy (N = 1033) at Cleveland Clinic or Barnes-Jewish Hospital. Clinical information was obtained from prospective data bases. Comorbidity was assessed using ACE-27 and Charlson Comorbidity indices. PC risk was classified according to D’Amico criteria. Fine and Gray competing risk analysis was used to assess PCSM and CCM at 10 years. Results: Over a median follow-up of 72 months (IQR: 46-97), 88 and 748 PCSM and CCM events were observed. Among healthy men with low risk PC, 10 year PCSM was 2% and CCM was 19%. Among healthy men with high risk PC, PCSM was 11% and CCM was 27%. In the group with moderate-to-severe comorbidities, CCM was 49, 59%, and 58% and PCSM was 1%, 3%, and 21% among those with low-, intermediate- and high-risk PC, respectively. Among these unhealthy men, 26% were treated by radical prostatectomy, of whom 45% had low-risk PC and 16% had high-risk PC. Among healthy men, 41% were treated by radical prostatectomy, of whom 54% and 9% had low- and high-risk PC, respectively. Conclusions: The risk of PCSM vs. CCM for older men is low, particularly for those with moderate-to-severe comorbidity; 49-59% had died from CCM within 10 years. Current evidence suggests that local therapy for PC is associated with a 25% reduction in PCSM, at best. Thus, with active surveillance, it is unlikely that PCSM would exceed 5-7% in those with low- and intermediate-risk PC. These results should inform elderly men and physicians about the risk of PCSM and CCM when deciding upon treatment for localized PC.