Objectives: For prostate cancer screening, the role of prostate-specific antigen (PSA) velocity (PSAV) in conjunction with total PSA is controversial. We evaluated the relationship of PSAV to histologic findings on biopsy and assessed whether PSAV provides independent predictive information.
Methods: From a community-based cohort of 25,276 men screened from 1991 to 2001, 1851 underwent a first biopsy for an elevated PSA and nonsuspicious digital rectal examination with a PSAV available from the year before biopsy. We analyzed the association between PSAV and biopsy histology.
Results: The histologic findings on biopsy were cancer in 468 (25%), prostatic inflammation in 135 (7%), and benign prostate tissue in 1248 (68%). The cancer detection rate was associated with PSAV and, depending on PSAV, ranged from 13% to 36% (P <0.001). Among men with a PSAV less than 0.5 ng/mL per year, the cancer rate ranged from 27% to 36%, at a PSAV of 0.5 to 3.0 ng/mL per year was 24% to 28%, and at a PSAV greater than 3.0 ng/mL per year was 13% to 18%. On multivariable analysis adjusting for age and PSA, PSAV was independently associated with risk of cancer on biopsy (P <0.0005). The rate of prostatic inflammation was directly associated with PSAV (PSAV of 3.0 ng/mL per year or less: 5% to 9%; PSAV greater than 3.0 ng/mL per year: 11% to 13%, P = 0.01).
Conclusions: In screened men with an elevated PSA undergoing biopsy, PSAV provides independent predictive information for estimating prostate cancer risk. Modest increases in PSA are associated with an increased risk of cancer, whereas more dramatic PSA rises are associated with a diminishing risk of cancer and higher rate of inflammation.