Abstract

AIMS AND BACKGROUND:

The aim of this study was to assess the validity of predictors of prostate biopsy outcome in order to improve their positive predictive value.

MATERIAL AND METHODS:

The study material consisted of a consecutive series of 410 prostate biopsies performed during 2003. The variables tested as possible predictors were age, findings at palpation (DRE) and ultrasonography (TRUS), total prostate-specific antigen (PSA), and free-to-total prostate-specific antigen (F/T) ratio. The association with biopsy outcome (cancer vs non-cancer) was investigated by univariate and multivariate analysis.

RESULTS:

All tested variables showed a statistically significant and independent association with biopsy outcome both in univariate and multivariate analysis. Nevertheless, no variable had good performance as a biopsy indicator: depending on the considered variable, three to nine cancer biopsies would be delayed in order to avoid ten benign biopsies. Using 0.12, 0.15 and 0.20 as the cutoff for F/T would avoid 77.3%, 64.4% and 43.1% of benign biopsies but would delay 54.0%, 35.6% and 21.0% of cancer biopsies, respectively.

CONCLUSION:

Although it may contribute to diagnostic suspicion, F/T should never exclude a biopsy indicated because of suspicion arising from other diagnostic tests.