Objective: To determine if adjustment of prostate-specific antigen velocity (PSAV) for variation improves prediction of cancer in men with previous negative prostate biopsy.
Patients and methods: Records of men undergoing prostate biopsy between 1999 and 2004 by a single urologist were reviewed to identify men with at least three follow up PSA measurements. Patients with atypia, high grade prostatic intraepithelial neoplasia or cancer on baseline biopsy were excluded. Men were rebiopsied if perceived to have rising PSA. Men with cancer, no cancer, or no repeat biopsy were compared for PSAV and a new parameter, PSAV%/Variation. PSAV was calculated by linear regression, and adjusted to percent change (PSAV%). Diagnostic accuracy was assessed by receiver operating characteristic curve.
Results: Of 118 men who met inclusion criteria, 32 had repeat biopsies. Nine biopsies were positive (group 1) and 22 were negative (group 2). The PSAV%, PSAV, and PSAV%/Variation for groups 1 versus 2 was 22.9% and 1.7% (p = 0.004), 1.12 versus 0.4 ng/ml/year (p = 0.007), and 1.07 verus 0.03 (p < 0.001), respectively. PSAV%/Variation had the largest area under the curve (0.881), compared with PSAV (0.744) and PSAV% (0.784). At cut off of 0.77, specificity was 86.4% and sensitivity was 87.5% for PSAV%/Variation. At the same sensitivity level, the specificities of PSAV% and PSAV were 77.3% and 63.6%, respectively.
Conclusion: Correction for variation could potentially make PSAV a more reliable parameter in patients with prior negative biopsy. The results of our preliminary study warrant further analysis in a larger prospective cohort.