Usefulness of the (F/T)/PSA Density Ratio to Detect Prostate Cancer

Abstract

Prostate-specific antigen (PSA) is the most important marker in the detection of prostate cancer, but its reliability is limited by a low specificity. To improve the specificity of PSA, a new parameter, based on the combination of free/total PSA (F/T) with PSA density, is proposed. From April 1999 to June 2003, a series of 647 men with high serum PSA levels were submitted to ultrasound-guided multiple prostate biopsies. Total PSA and free PSA were measured by immunoassay (DASIT, Italy). Receiver operating characteristic (ROC) curves of PSA density, F/T and (F/T)/PSA density (F/T divided by PSA/prostate volume) were constructed, and relative specificity and sensitivity were calculated. Prostate cancer was present in 284 cases (137 of them with PSA values between 4 and 10 ng/ml), 31 patientshad intraepithelial neoplasia and the remaining cases showed benign prostatic hyperplasia. Analysis of ROC curves in all examined patients showed that the area under the curve (AUC) for (F/T)/PSA density (0.849) was significantly higher than the AUC for F/T (0.681) (p = 0.012). Similar results were observed in the cases with PSA levels between 4 and 10 ng/ml, where the AUC for (F/T)/PSA density (0.768) was significantly higher than the AUC for F/T (0.610) (p = 0.041). In our study, the prostate cancer risk became very high when (F/T)/PSA density values decreased to <1.5. The use of the (F/T)/PSA density ratio may be useful in the detection of prostate cancer, reducing unnecessary prostate biopsies. A cutoff of 1.5 for the (F/T)/PSA density value is recommended for patients with PSA levels between 4 and 10 ng/ml.