A SelectMDx/magnetic resonance imaging-based nomogram to diagnose prostate cancer

Abstract

Background: 70%-80% of prostate cancer (PCa) biopsies performed in the US annually may be unnecessary. Specific antigen testing (PSA) and tans rectal ultrasound (TRUS) are imprecise predictive methods for risk of PCa. Novel strategies are critical to guide biopsy decision-making.

Aim: We assessed the utility and accuracy of combining Select MDx and multiparametric magnetic resonance imaging (mpMRI) scores for predicting risk of PCa.

Methods and results: Our study was conducted at Mount Sinai hospital at Urology department in New York City from January 2020 to April 2021. Total 129 men performed select MDx test. Indications for prostate biopsy were high-risk Select MDx score, suspicious DRE, PI-RADS scores 3/4/5 on mpMRI, or any combination of these. Fifty-one percentage of 129 patients underwent systemic or combined systemic and MRI/US (ultrasound) fusion biopsy; All men underwent 3 T MRI of Prostate w/wo contrast using standard protocols prior to biopsy. A single surgeon performed prostate biopsies. Gleason score ≥3 + 3 on biopsy is defined as outcome. Descriptive statistics were calculated as cross tables. Binary logistic regression model is used to determine the outcome. The nomogram was based on the coefficients of the logit function. ROCs were plotted and decision curve analysis was performed. Using both high-risk Select MDx and PI-RADS scores of 4/5, 87% of biopsies could have been avoided, while detecting 64% of PCa and missing 36%. If biopsies were performed on men with positive Select MDx or PI-RADS 4/5 results, 16% of biopsies could have been avoided while detecting all PCa. Combining these scores improved specificity and accuracy for the detection of PCa over either used alone. Study limitations include limited sample size, sole institution study, and risk or overfitting for the proposed model which may limit generalizability.

Conclusion: Combining SelectMDx and mpMRI PI-PADS scores of 4/5 may be useful for PCa biopsy decision-making.

Keywords: PCa biomarkers; SelectMDx; mpMRI PI RADS score; prostate biopsy.

FIGURE 1

Nomogram prediction model for predicting prostate cancer after prostate biopsy. The prediction of adverse pathology probability from the nomogram includes the following steps: (1) Locate the patient's variable MRI PI‐RADS score on the corresponding axis. (2) Draw a line straight down to the score axis to determine how many points toward the probability of cancer the patient receives for his PIRADS score. (3) Repeat the process for the Select MDx score. (4) Total the points for each of the predictors. (5) Locate the final sum on the total score axis. (6) Draw a line straight up to find the patient's probability [Prob] of having Prostate cancer. MRI PIRADS, magnetic resonance imaging Prostate Imaging and Reporting Data System v.2 score; PI‐RADS 0, No PI‐RADS score or PI‐RADS score 1 or 2.

FIGURE 2

Area under curve characteristics of the prediction model for prostate cancer. PIRADS, magnetic resonance imaging Prostate Imaging and Reporting Data System v.2 score.

FIGURE 3

Decision curve analyses showing the net benefit associated with the use of nomogram‐derived probability for prediction of prostate cancer rather than relying on Select MDx score or PI‐RADS score information only. PIRADS, magnetic resonance imaging Prostate Imaging and Reporting Data System v.2 score.