Using Competing Risk of Mortality to Inform the Transition from Prostate Cancer Active Surveillance to Watchful Waiting

Abstract

BackgroundFor men on active surveillance (AS) for prostate cancer (PCa), disease progression and age-related changes in health may influence decisions about pursuing curative treatment. ObjectiveTo evaluate the predicted PCa and non-PCa mortality at the time of reclassification among men on AS, to identify clinical criteria for considering a transition from AS to watchful waiting (WW). Design, setting, and participantsPatients enrolled in a large AS program who experienced biopsy grade reclassification (Gleason grade increase) were retrospectively examined. All patients who had complete documentation of medical comorbidities at reclassification were included. Outcome measurements and statistical analysisA validated model was used to assess 10- and 15-yr untreated PCa and non-PCa mortalities based on patient comorbidities and PCa clinical characteristics. We compared the ratio of predicted PCa mortality with predicted non-PCa mortality (“predicted mortality ratio”) and divided patients into four risk tiers based on this ratio: (1) tier 1 (ratio: >0.33), (2) tier 2 (ratio 0.33–0.20), (3) tier 3 (ratio 0.20–0.10), and (4) tier 4 (ratio <0.10). Results and limitationsOf the 344 men who were reclassified, 98 (28%) were in risk tier 1, 85 (25%) in tier 2, 93 (27%) in tier 3, and 68 (20%) in tier 4 for 10-yr mortality. Fifteen-year risk tiers were distributed similarly. The 23 (6.7%) men who met the “transition triad” (age >75 yr, Charlson Comorbidity Index >3, and grade group ≤2) had a 14-fold higher non-PCa mortality risk and a lower predicted mortality ratio than those who did not (0.07 vs 0.23, p < 0.001). The primary limitations of our study included its retrospective nature and the use of predicted mortalities. ConclusionsAt reclassification, nearly half of patients had a more than five-fold and one in five patients had a more than ten-fold higher risk of non-PCa death than patients having a risk of untreated PCa death. Despite a more significant cancer diagnosis, a transition to WW for older men with multiple comorbidities and grade group <3 PCa should be considered. Patient summaryMen with favorable-risk prostate cancer and life expectancy of >10 yr are often enrolled in active surveillance, which entails delay of curative treatment until there is evidence of more aggressive disease. We examined a group of men on active surveillance who developed more aggressive disease, and found, nevertheless, that the majority of these men continued to have a dramatically higher risk of death from non–prostate cancer causes than from prostate cancer based on a risk prediction tool. For men older than 75 yr, who have multiple medical conditions and who do not have higher-grade cancer, it may be reasonable to reconsider the need for curative treatment given the low risk of death from prostate cancer compared with the risk of death from other causes.