INTRODUCTION AND OBJECTIVE:
To evaluate if new parameters based on MRI and MRI-targeted biopsy add accuracy in the EAU risk groups to predict who will develop metastatic prostate cancer after radical prostatectomy (RP).
To estimate the risk of metastatic and biochemical (BCR) recurrence incorporating volume of poorly differentiated cancer based on MRI and MRI-targeted biopsy before RP.
Retrospective cohort of 713 consecutive men who underwent prostate MRI-targeted biopsies and RP between 2009 and 2018 in our center. Preoperative parameters included PSA, cT-stage, MRI tumor volume (TV) based on lesion largest diameter, percent of Gleason pattern 4/5 (%GP4/5), PSA density (PSAD), volume of GP4/5 (VolGP4/5). Their predictive values for metastases and BCR were assessed using univariable and multivariable Fine-and Gray models; models’ performances were compared using the Akaike criteria information (AIC) and Harrell’s C-index.
During a median follow-up of 57 months, 132/713 had BCR, 48/713 had metastatic recurrence, and the 5-year probability of developing metastases was 5.6% (95%CI, 3.9-7.7). In univariable model, PSAD, %GP4/5, TV, VolGP4/5 were significantly associated with metastatic recurrence and BCR. Risk of metastases increased with %GP4/5, with a subHazard Ratio (SHR) of 13.36 (95%CI, 5.59-31.92) between 40 and 75%, and 19.57 (95%CI, 7.21-53.09) for 100% using %GP4/5 <40% as reference. VolGP4/5, treated as categorical variable (<0.5; 0.5-1.0; 1.01-3.2; >3.2 ml), was the parameter with lowest AIC and the highest C-index for metastatic recurrence C-index=0.82 (95%CI, 0.76-0.88) and for BCR C-index=0.73 (95%CI, 0.68-0.78). EAU risk groups classification showed C-index=0.74 (95%CI, 0.67-0.80) for metastatic recurrence and C-index=0.67 (95%CI, 0.63-0.72) for BCR. Performance of multivariable model including %GP4/5 and TV for predicting risk of metastatic recurrences showed C-index=0.86 (95%CI, 0.79-0.91) (Figure 1) and C-index=0.77 (95%CI, 0.72-0.82) for BCR.
Preoperative estimation of volume of poorly differentiated cancer based on MRI and targeted biopsies showed high performance for predicting metastatic recurrence after radical prostatectomy. These results may assist for better personalized patient management.
Source of Funding: