Purpose: To determine variables related to the finding of prostate cancer (PC) in patients who underwent surgery following at least one negative prostate biopsy (PB).
Materials and methods: A retrospective study of 170 patients who underwent transurethral resection of the prostate (TURP) or open prostatectomy between 1999 and 2007, following one or more negative PB sets. A multivariate logistic regression analysis was carried out in order to determine variables related to the finding of PC. The predictive capacity of PSA, PSA-density and PSA-velocity was assessed by means of ROC curves and the area under the curve (AUC). Sensitivity, specificity and predictive values were determined for several PSA-density and PSA-velocity cut-off points.
Results: Open prostatectomy was carried out on 104 patients (61.18%) and TURP on 66 (38.82%). PC was detected in the surgical specimen of 16 patients (9.41%). Variables associated with the finding of PC in the surgical specimen were PSA-density (OR: 1.47; 95% CI: 1.22-6.64; p: 0.007) and PSA-velocity (OR: 2.87; 95% CI: 1.60-5.12: p < 0.001). The AUCs were 0.746, 0.793 and 0.832, for PSA, PSA-density and PSA-velocity, respectively. The most sensitive PSA-density and PSA-velocity cut-off points in detecting PC were 0.15 and 1 ng/ml/year, respectively. Patients without PC showed a median PSA reduction of 9.35 ng/ml (-2.40 - 35.40), following surgery.
Conclusions: PSA-density and PSA-velocity in particular, allow for the prediction of the presence of PC in the TURP or open prostatectomy specimen of patients with previously negative PBs. Diagnostic TURP could prove useful in patients with clinical suspicion of PC, susceptible to curative treatment, with PSA-velocity > 1l ng/ml/year and one or more negative saturation biopsies.