Prognostic impact of time to undetectable prostate-specific antigen in patients with positive surgical margins following radical prostatectomy.

Abstract

The purpose of this article was to determine the impact of time to undetectable prostate-specific antigen (PSA) for predicting biochemical recurrence (BCR) in patients with a positive surgical margin (PSM) following radical prostatectomy (RP). A PSM is an independent predictor of BCR; however, not all patients develop BCR later on. A retrospective analysis was conducted on 1,117 consecutive prostate cancer patients who underwent RP without neoadjuvant or adjuvant therapy from July 2005 to December 2009. Of these, 516 (46.2 %) patients without PSMs, and 214 (19.2 %) patients with PSMs who later achieved undetectable PSA, defined as <0.01 ng/ml, were identified. Patients with PSMs were stratified according to time to undetectable PSA dichotomized at 6 weeks and compared with patients without PSMs. Patients with PSMs who did not achieve undetectable PSA were excluded. BCR was defined as two consecutive increases of post-undetectable PSA ≥0.2 ng/ml. During the median follow-up of 58.2 months, patients with PSMs who achieved undetectable PSA in <6 weeks had comparable 5-year BCR-free survival rates to those without PSMs; however, patients with PSMs who achieved undetectable PSA in ≥6 weeks showed significantly lower rates compared with both patients without PSMs (59.2 vs 74.3 %; p < 0.001) and patients with PSMs who achieved undetectable PSA in <6 weeks (59.2 vs 78.8 %; p = 0.004). Among patients with PSMs, multivariate analysis revealed time to undetectable PSA at ≥6 weeks and seminal vesicle invasion to be independent predictors of BCR. No perioperative factors were associated with undetectable PSA at ≥6 weeks. Patients with PSMs who achieve undetectable PSA in <6 weeks show comparable risks of BCR to patients with negative surgical margins.