Biochemical Recurrence (BCR) after radical prostatectomy (RP) has traditionally been quoted to occur roughly in one-third of patients at 10-years after definitive treatment based on data from cohorts treated between 1969 and 2005 [1], [2], [3], [4], [5], [6]. Over the last 20 years, there has been a marked change in prostate cancer (PC) management favoring active surveillance (AS) for lower-risk disease, with AS/conservative management emerging as the most common management for low-risk patients in the Veterans Affairs Health System (VA) in 2010 [7]. We recently showed that among men within the VA SEARCH cohort, the percent of patients undergoing RP for low-risk disease decreased from 51% to 7% between 2000 and 2017, offset by an increase in the proportion of high-risk cancers undergoing RP from 18% to 33% [8]. Additionally, over this time there have been improvements in risk stratification related to adoption of prostate MRI image-guided biopsy, which tends to increase the detection of high grade disease and the apparent risk of high-risk tumors [9], [10], [11], [12], [13]. The net result is that men undergoing RP in the post-AS era have higher risk tumors. These developments have the potential to alter risks of BCR overall and within tumor-risk subgroups. However, contemporary trends in BCR risk are unknown.

In this study, we used the VA SEARCH database to assess how BCR risk after RP has changed in the pre- and post-AS eras and to provide contemporary estimates for BCR risk, both overall and in tumor risk subgroups. The VA SEARCH database is uniquely suited to answer this question due to its detailed information on preoperative risk stratification and follow-up PSAs after RP, both of which are limited in large secondary datasets such as SEER and NCDB. Herein, we sought to determine contemporary trends of BCR after RP. We hypothesized that overall BCR risk would increase in the AS era, due to an increasing trend towards using RP for higher risk disease and improvements in risk stratification.