Giovannone R1,De Berardinis E1, Sciarra A1, Panebianco V2, Shahabadi H1, Gentile V1, Salciccia

Prostate cancer gene 3 (PCA3) and multiparametric magnetic resonance (mMRI) can reduce unnecessary biopsies: decision curve analysis to evaluate predictive models.

Giovannone R1,De Berardinis E1, Sciarra A1, Panebianco V2, Shahabadi H1, Gentile V1, Salciccia S1andBusetto GM1.

1 – Department of Urology, Sapienza University of Rome – Rome – Italy

2 – Department of Radiology,Sapienza University of Rome – Rome – Italy

Introduction

Several new biomarkers have been proposed to overcome the current limits of PSA. Since its introduction in clinical practice, urinary prostate cancer gene 3 (PCA3) assay has shown promising results for prostate cancer (PC) detection, staging and prognosis information. Furthermore multiparametric magnetic resonance imaging (mMRI) has the ability to better describe several aspects of the natural history of PC.

Materials and methods

To assess the role of mMRI and PCA3 tests in identifying PC patients previously negative to a prostate biopsy, we conducted a prospective study on 171 patients with clinically suspected prostate cancer, previously resulting negative to the TRUS-guided prostate biopsy but with persistent high PSA serum levels (between 4 and 10 ng/ml). All patients were submitted to PCA3 test and mMRI prior to second TRUS-guided prostate biopsy.

Accuracy and reliability, most used diagnosis tests, have been evaluated. Four multivariate logistic regression models have been analyzed, in terms of discrimination and cost benefit, to assess the clinical role of PCA3 and mMRI in predicting biopsy outcome. Furthermore a decision curve analysis has been plotted.

Results

The repeated TRUS biopsy identified 68 (41.7%) new PC patients; 95 patients (58.3%) were not evaluated according to the Gleason score as the biopsy was negative. The sensitivity and specificity of PCA3 test (cut-off 35) and mMRI were respectively; 68% and 49%, 74% and 90%. Decision curve analysis showed a good performance in predictin PC cancer with the model including mMRI and PCA adjusted by age, PSA and DRE (full model).

Limitations: not randomized study, small sample size.

Discussion

The hypothesis in our study is that the potential value of PCA3 as a biomarker for PC diagnosis could be improved by using the mMRI for directing prostate biopsy in order to overcome the current limits of random prostate biopsy. In our trial for a PCA3 score 35 alone as predictor of PC, we reported a sensibility and specificity of 68% and 49% respectively and 0.59 of AUC which appears to be similar to other experiences in the same population of patients(persistently elevated PSA serum levels and first negative biopsy), and therefore not very useful as a biomarker to detect PC patients.

In clinical practice, however, the application of the base clinical model + PCA3 resulted in no net benefit gain with a cut-off of 20%, while using the clinical base model + MRSI and even more full clinical model (same cut-off), could avoid a large number of unnecessary biopsies at the cost of only a small number of patients with PC being advised not to perform biopsy. Starting from a cut-off of 26% the net benefit gain of using the base clinical model + PCA3 without mMRI starts to be remarkable.

Conclusions

According to our experience the use of mMRI to guide prostate biopsy can increase both the accuracy of this procedure and the sensitivity of the PCA3 test.