Radiological T-staging in prostate Cancer: Towards a universal MRI-based scoring system

Abstract

Magnetic resonance imaging (MRI) is recommended for local prostate cancer tumor (T) staging. While MRI offers superior anatomical resolution compared to digital rectal examination, its accuracy in distinguishing organ-confined (T2) from locally advanced (T3a) tumors remains low. This narrative review critically examines the current MRI-based T-staging performance, its limitations, and the clinical implications of image scoring systems to assess local tumor extent. We highlight the evolution from qualitative assessments toward semi-quantitative and quantitative systems, each integrating features like tumor capsular contact length, bulging, capsular irregularity, neurovascular bundle asymmetry, and capsular enhancement. Emerging data suggest that tumor volume and curvilinear capsular contact length exhibit high specificity, however, most features lack sufficient sensitivity and inter-reader agreement to independently guide treatment management. We argue for favoring specificity over sensitivity in MRI-based T-staging, considering that false positive MRI readings carry negative side effects for patients following overtreatment. Finally, we propose the development of a universal MRI-based scoring system tailored to clinical relevance, prioritizing reproducible guidance for treatment decisions over traditional staging accuracy. Such a scoring system better aligns with contemporary therapeutic goals, reduces overtreatment with its negative side effects, and facilitates standardized reporting in clinical trials and practice.