#1024

Comparing accuracy in Multiparametric MRI of prostate and PSMA PET in the detection of Prostate Cancer and Correlation with Radical Prostatectomy

M. Mo¹, N. Sapre¹, J. Gleeson², D. Desai¹

¹Toowoomba Hospital, Urology, Toowoomba, Australia
²St Vincent's Hospital, Urology, Toowoomba, Australia

Introduction:

With guidelines being well formulated for prostate malignancy work up, we find an evolving use and dependence on imaging modality for gold standard prostate malignancy investigations.[1] This study examines the comparative accuracy of multiparametric MRI of the prostate (MRIp) and PSMA PET in detecting prostate cancer, particularly focusing on lesions that were initially missed by MRI but later confirmed via targeted biopsy or radical prostatectomy (RP) in a regional setting. With an increasing reliance on imaging for risk stratification, it is essential to determine the extent to which MRI accurately predicts ISUP grade groups compared to the gold standard RP findings. [2] Moreover, PSMA PET has emerged as a promising adjunct modality that may complement MRI, especially in cases with ambiguous imaging.[3] By assessing the rate at which PSMA PET aligns with RP outcomes in instances where MRI underperforms, our aim is to elucidate the potential role of MRI and PSMA PET in enhancing diagnostic precision in prostate cancer management.

Material and methods:

In this prospective, multi-institutional study, a single surgeon enrolled patients undergoing both multiparametric MRI (MRIp) and PSMA PET prior to radical prostatectomy (RP). Imaging findings were compared against final RP histopathology. The primary endpoint was the detection of clinically significant prostate cancer missed by MRIp or PSMA PET and later identified on RP histology, Patients having had low grade and later showed high grade on RP were considered missed.

Results:

In a prospective analysis of 56 patients (mean age: 67), the diagnostic accuracy of multiparametric MRI (MRIp) and PSMA PET was compared against final radical prostatectomy histopathology. MRIp was completely accurate in 26 patients (46.4%), partially accurate in 30 (53.6%), and completely inaccurate in 6 cases (10.7%). In contrast, PSMA PET was completely accurate in 41 patients (73.2%), partially accurate in 18 (32.1%), and inaccurate in 3 cases (5.4%). PSMA PET demonstrated significantly greater alignment with final ISUP grade and tumor localization compared to MRIp (p = 0.004 vs. p = 0.041). These findings highlight that while MRIp performs reasonably well—particularly in low-risk disease—its limitations in accurately grading and localizing high-risk or multi-zonal lesions reduce its standalone reliability. PSMA PET showed more consistent accuracy across all risk groups and prostate zones, reinforcing the benefit of a multimodal imaging strategy in preoperative prostate cancer assessment.