e16054 Background: Biomarkers that predict aggressiveness and tumor burden of prostate cancer are needed for patients who undergo AS. One of such candidates is [-2]proPSA. It has been reported that expression of [-2]proPSA is specific to prostate cancer and associated with grade and stage. The aim of this research is to evaluate the predictive ability of [-2]proPSA and its related parameters for the biopsy reclassification during AS. Methods: 13 institutions participated in the multi-center Japanese AS study (a prospective one-arm cohort study). 134 patients were enrolled in this study between January 2002 and December 2003. Inclusion criteria are as follows; (1) stage T1cN0M0, (2) age from 50 to 80, (3) serum PSA = 20 ng/ ml or less, (4) one or two positive cores per 6 to 12 systematic biopsy cores, (5) Gleason score = 6 or less and (6) maximum cancer involvement in positive core = 50% or less. We prospectively collected blood samples from patients who opted AS as an initial treatment. The serum samples were stocked frozen until testing. The patients who remained on AS for 1-year were recommended to undergo repeat biopsy. Reclassification was judged by a central pathologist who was also involved in the initial diagnosis at inclusion. Results: 67 patients underwent repeat biopsy at 1-year after AS. Biopsy reclassification was histopathologically found in 25 of 67 patients (37.3%). Table shows the patients’ backgrounds and the cutoff values of predictive variables at 90% and 80% sensitivity and specificity. p2PSA, %p2PSA and phi; (p2PSA/fPSA) × √(tPSA) levels at baseline and p2PSA and phi at 1-year after AS were significantly higher in the reclassification group than those in the non-reclassification group. At 90% sensitivity, the cutoff values of p2PSA, %p2PSA and phi at baseline were 12.52, 1.51 and 37.91 with a specificity of 23.81%, 26.19% and 26.19%, respectively. In both 90% and 80% sensitivity, p2PSA and its related parameters have higher specificity than PSA and its related parameters. Conclusions: [-2]proPSA and its related parameters could predict pathological reclassification in repeat biopsy during AS.