e546 Background: The goal of this investigation is to use longitudinal Active Surveillance (AS) data to identify whether biopsies could be eliminated from a 10-year AS plan calling for annual biopsy without substantially prolonging the time to detection of prostate cancer (CaP) grade progression. Methods: With data from 1,500 men with very-low or low-risk CaP enrolled in AS at Johns Hopkins, we developed a hidden Markov model to estimate probabilities of progression to Gleason 7 or higher. We then simulated all potential AS biopsy strategies where it was assumed biopsy would be performed no more often than annually and either would or would not be performed each year after diagnosis. For every strategy the model was used to predict the average time from occurrence to the detection of grade progression. Potential alternatives to annual biopsy were identified. Results: The estimated 10 year probability of grade progression was 40.0% and the annual probability was 4.0%. Simulation of an annual biopsy strategy estimated that for men who experience grade progression the mean time from occurrence to detection would be 14.1 months. Alternative strategies that reduced the number of biopsies received over 10 years increased the time to detecting grade progression; however several strategies eliminated biopsies with only small increases in the time to detecting grade progression (Table). For instance, a strategy calling for 6 biopsies over 10 years (biopsies in years 1-3,5,7, and 8) would only increase the average time to detecting grade progression by 4.6 months while eliminating 4 biopsies. Conclusions: While annual biopsy was associated with the shortest interval to detecting grade progression, several alternative strategies may allow for less frequent biopsy without sizable increases in time to detecting grade progression. [Table: see text]