For a study, researchers sought to present the most recent evidence on interobserver agreement and accuracy; assess the strengths, weaknesses, and implications of use; and outline opportunities for improvement and future development of the Prostate Imaging Reporting and Data System version 2.1 (PI-RADS v2.1) for detection of prostate cancer (PCa) on multiparametric magnetic resonance imaging (mpMRI). According to the current data analysis, recent modifications to the PI-RADS system with PI-RADS v2.1 modestly improved interobserver agreement, with usually high sensitivity and moderate specificity for the diagnosis of clinically relevant PCa. Recent research also showed a significant increase in diagnostic specificity with PI-RADS v2.1 over PI-RADS v2. However, the research findings comparing the performance of v2.1 were hampered by small sample numbers and retrospective cohorts, which might introduce selection bias. Some studies found a significant difference between v2.1 and v2, while others found no statistically significant difference. Furthermore, in PI-RADS v2.1, the interpretation and reporting of specific findings, notably for category 2 lesions, remained subjective, and reader experience continued to vary significantly. The variables also contributed to the fact that there was still some interobserver variability and that more experienced readers did better.

Interobserver agreement, diagnostic performance, and sensitivity and specificity improved with PI-RADS v2.1, with more significant gains reported among more experienced readers. However, the therapeutic significance of the development might be minor given the declining character of the advances and the poor power of all trials analyzed. Despite advancements in PI-RADS v2.1, the most critical element in prostate cancer identification was practitioner skill in evaluating mpMRI of the prostate.

Reference:link.springer.com/article/10.1007/s11934-022-01084-y