Objectives: Preoperative prostate-specific antigen (PSA) velocity (PSAV), the rate of PSA rise preceding diagnosis, predicts for relapse and cancer death after prostatectomy or radiotherapy. We studied the timing, cutoff levels, and method of calculation to better define its usefulness.
Methods: The rates of biochemical relapse were examined in 471 patients who underwent radical prostatectomy (RP) with previous PSA history available. PSAV was calculated by two methods, as the difference between two PSAs divided by time, or as the slope of all available PSAs within that interval. Kaplan-Meier relapse-free survival was compared among the groups with various intervals and cutoff levels in their preoperative PSAV definition. Univariate and multivariate analysis examined all preoperative factors and PSAV for their association with relapse.
Results: The two methods of PSAV calculation yielded values within 5% of each other (R2 = 0.91). A PSA history that precedes RP by at least 12 months is necessary. A PSAV cutoff level of 2 ng/mL/yr or less versus greater than 2 ng/mL/yr appeared optimal for a PSA interval spanning 12 to 24 months before RP (P = 0.008). PSAV using a longer interval (24 to 36 months) before RP appeared more sensitive, with a cutoff of 1 ng/mL/yr or less versus greater than 1 ng/mL/yr (P = 0.029) and 2 ng/mL/yr or less versus greater than 2 ng/mL/yr (P = 0.0041) associated with relapse. A preoperative PSAV of 2 ng/mL/yr or less versus greater than 2 ng/mL/yr was an independent factor associated with the risk of relapse after RP.
Conclusions: The results of our study have shown that preoperative PSAV is independently associated with relapse after RP. However, a minimum interval of 12 months before RP is needed, and a PSAV cutoff level of 2 ng/mL/yr appears optimal. A simple two-point method of calculating PSAV is reliable.