Based on the EAU guidelines recommendation, patients with PIRADS 3 and PSAd less than 0.1 are considered of very low risk of harboring clinically significant prostate cancer (CsPca), 0.1-0,15 have intermediate low risk, and above 0.15 have risk above 20% of CsPca. Several biomarkers have been tested to increase detection rate, but their use is not widely extended nor validated.
In the present study, we examined further clinical and radiologic characteristics in men with PI-RADS 3 index lesions that can predict csPCa on mpMRI-target biopsy, in addition to PSAD for decision making.
Table 1. Adapted from Cornford, P. et al. EAU-EANM-ESTRO-ESUR-ISUP-SIOG Guidelines on Prostate Cancer—2024 Update. Part I: Screening, Diagnosis, and Local Treatment with Curative Intent. European Urology S0302283824022541 (2024) doi:10.1016/j.eururo.2024.03.027. Risk data table of clinically significant prostate cancer, related to PI-RADS score and PSA-D categories in biopsy-naive men.
For this aim, we performed a revision of a prospective database with patients who underwent targeted and systematic biopsies from 2015 to 2023 for PI-RADS 3 lesions identified on mpMRI. Baseline variables were collected, such as PSA density (PSAd), 4Kscore, prostate size, and the apparent diffusion coefficient (ADC) value of the lesion on mpMRI. Dedicated genitourinary radiologists reviewed all MRIs. Logistic regression, receiver operating characteristic (ROC), and decision curve analyses (DCA) assessing the association between clinic-radiologic factors and csPCa were performed.
Overall, 230 patients were included in the study and the median age was 65 years. The median prostate size and PSA were 50 g and 6.26 ng/mL, respectively. 17.4% of patients had csPCa, while 27.5% had Gleason group 1. In univariable logistic analyses, we found that age, BMI, prostate size, PSAd, ADC, and 4Kscore were significant csPCa predictors (P<0.05). PSAd showed the best prediction performance in terms of AUC (=0.679), as compared to 4k (AUC= 0.667) and ADC (AUC=0.61). On multivariable analysis, PSAd and 4Kscore were associated with csPCa (OR of 1.59 [1.23 – 2.06] and 1.1 [1.02 – 1.19], respectively). The net benefit of PSAd combined with clinical features was superior to those of other parameters (4k or ADC). Using a threshold of PSA density of 0.1 ng/ml/cm3 or 4k score of 10% would help reduce unnecessary biopsies without missing a significant proportion of clinically significant prostate cancer. Within patients with PSAd < 0.15, 4Kscore was a statistically significant predictor of csPCa (OR=3.25, p=0.032).
Figure 1. AUC plot (A) and decision curve analysis plot (B) for three selected univariable logistic models (PSAd, ADC, and 4Kscore, respectively) assessing association between baseline clinic-radiologic factors and csPCa.
To our knowledge, this is the largest series exploring the predictor rates of clinically significant prostate cancer in patients with PIRADS 3 lesions, exploring variables such as PSAD, 4k, and ADC.
In conclusion, PIRADS 3 lesion represents a “gray zone” where further tools are needed for decision making for risk stratification and prostate biopsy. PSAd and 4Kscore are better predictors of csPCa in patients with PIRADS 3 lesions compared to ADC. The predictive role of 4Kscore is higher in patients with low PSAd; 4k score has better performance as compared to patients with high PSAD. These results can assist practitioners in the risk stratification of patients with equivocal lesions to determine the need of biopsy.
Written by: Tarek Ajami, MD, Urologist, Desai Sethi Urology Institute, University of Miami, Miller School of Medicine, Miami, FL
References:
Read the Abstract