Post-prostatectomy PSA slope (ProsVue) in Gleason score ≥7 disease

Abstract

e15113 Background: A case-cohort designed multicenter clinical study addressed the prognostic value of the linear slope of PSA in 304 men after radical prostatectomy (RP) – 64 with clinical recurrence and 240 with stable disease. PSA was measured with Nucleic Acid Detection Immunoassay (NADiA) technology having a limit of quantification of 0.00065 ng (0.65 pg) per mL. Least-squares linear PSA slope (ProsVue) was calculated using 3 serum samples drawn 1.5-20 months (mo) post-RP. Multivariate Cox proportional hazards regression analysis showed that a ProsVue cutpoint of ≤2.0 pg/mL/mo was an independent prognostic factor for identifying men at reduced risk for recurrence along with clinical evaluation (HR=9.824). Of traditional risk factors – pre-RP PSA, Gleason score (GS) and pathologic stage, only GS was an independent prognostic factor (HR=5.3). This study tested the prognostic performance of ProsVue in a 164-man subset with pathologic GS ≥7. Methods: Recurrence was documented by positive imaging, positive biopsy or prostate cancer death. The same ProsVue cutpoint dichotomized men “at reduced” or “not at reduced” recurrence risk compared to clinically recurrent or stable disease status. Uni- and multivariate Cox proportional hazards regression were employed. Results: Median age, pre-RP PSA and percent tumor volume in the subset were 62.6, 7.2 ng/mL, and 20%. Distributions of GS 7, 8, 9 were 76.8%, 14.6% and 8.5%, respectively. Pathologic stages were T0-T2, T3a, T3b and T4 in 41%, 57%, 29% and 7% of men, respectively. Median follow-up in stable and recurrent groups was 10.8 and 4.8 yr. The table compares results in the full cohort and the GS ≥7 subgroup. Conclusions: ProsVue prognostic performance is maintained with GS ≥7. Performance parameters, especially the multivariate hazard ratio (HR) are minimally attenuated vs. the full case-cohort. [Table: see text]