130 The GLIMPSE Study: Global Variation in the Quality of Multiparametric Magnetic Resonance Imaging of the Prostate

Aim

Multiparametric MRI (mpMRI) is standard-of-care in patients with clinical suspicion of prostate cancer (PCa). However, high variability in MRI quality remains. We aim to offer the first global 'GLIMPSE' into prostate mpMRI quality, analysing PRIME Trial data, to determine whether modifications to MRI protocols improve diagnostic quality.

Method

PRIME (NCT04571840) is an international non-inferiority trial comparing biparametric MRI to mpMRI in detecting clinically significant PCa.

For each scanner, centres were invited to submit MRI scans of patients with suspected PCa, detailed MRI protocols, PI-RADS v2.1 scores and pathology results. Expert radiologists evaluated these in consensus, blinded to MRI and biopsy results. MRI quality was assessed using the PI-QUAL scoring system — using PI-RADS v2.1 recommendations and each MR sequence (DWI, T2-WI and DCE). PI-QUAL score 5/5 demonstrates scans of optimal diagnostic quality. Recommendations were provided for scanners not reaching PI-QUAL 5. Centres were invited to re-submit using modified protocols for re-evaluation.

Results

71 scanners were evaluated from 42 centres across 5 continents. On initial review, 13% scanners scored PI-QUAL 3, 55% scored PI-QUAL 4 and 32% scored PI-QUAL 5. 100% scanners were of adequate diagnostic quality for T2-WI and DWI, whilst only 82% were for DCE.

Recommendations focused on DCE (62%), followed by DWI (35%), then T2-WI (25%). Overall MRI quality increased from 32% to 87% PI-QUAL 5 between initial and final review.

Conclusions

GLIMPSE offers the first global overview into prostate MRI quality. Initial MRI quality needed improvement, particularly with DCE imaging. Basic changes derived from PI-RADS recommendations can improve global MRI quality.