Presentation Authors: Armando Stabile*, Paolo Dell'Oglio, Milan, Italy, Matteo Soligo, Rochester , MN, Francesco De Cobelli, Giorgio Gandaglia, Nicola Fossati, Milan, Italy, Luca Boeri, Rochester, MN, Simone Scuderi, Emanuele Zaffuto, Federico Dehò, Antonio Esposito, Giuseppe Fallara, Alessandro Del Maschio, Milan, Italy, R. Jeffrey Karnes, Rochester, MN, Francesco Montorsi, Alberto Briganti, Milan, Italy
Introduction: There is a lack of evidence on the ability of MRI to detect clinically significant prostate cancer (csPCa) in young patient. This is key due to an increasing number of men referred for early diagnosis. We hypothesized that the diagnostic performances of MRI for csPCa varies according to patient age.
Methods: We identified 993 patients with a positive MRI of the prostate (PI-RADSâ‰¥3) who underwent subsequent fusion MRI-targeted biopsy (TBx) and concomitant systematic biopsy (SBx) at a two referral Centres between 2013 and 2018. All patients were either biopsy naive or had a previous biopsy and all had a negative DRE. All MRIs were classified according to PI-RADS system by expert radiologists. The study outcomes were csPCa (defined as Gleason at biopsyâ‰¥3+4) at SBx and TBx. Two multivariable logistic regression models (MVA) were developed per each outcome. Covariates included age at biopsy, PSA density, PI-RADS (3 vs 4 vs 5), previous biopsy status (naive vs non naive) and the number of either SBx cores (Model 1) or TBx cores (Model 2). The hypothesis that MRI accuracy in detecting csPCa differed by age was tested using an interaction term between TBx vs SBx.
Results: The overall rate of csPCa was 54% (n=534). Overall, 344 (34.6%) and 484 (48.7%) patients had csPCa at SBx and TBx, respectively. The rate of csPCa missed by SBx and TBx was 35.7 (n=191) and 9.5% (n=51), respectively. The median number of SBx and TBx cores was 12 (IQR: 10-13) and 4 (IQR: 3-6), respectively. At MVA, age at biopsy was an independent predictor of csPCa at both SBx (OR: 1.03) and TBx (OR: 1.05), after accounting for confounders (all p < 0.04). In men aged 40-50 years, SBx had higher probability to detect csPCa relative to TBx (25 vs. 18% at 40 years). Conversely, in patients aged >50 years the probability of csPCa was higher in TBx vs. SBx (Figure 1), reaching the highest difference for very elderly patients (50 vs. 70% at 80 years). Positive predictive value of MRI for csPCa in men aged < 50 vs â‰¥50 years was 21% vs 50% (p=0.001).
Conclusions: In men aged < 50 years the performance of SBx in detecting csPCa is higher relative to TBx, reflecting a lower accuracy of MRI in this population. These results reinforce the need for a patient risk assessment in addition to MRI in young men and the need for concomitant SBx together with TBx in this patient group.