Diagnostic accuracy of 18F-PSMA-1007 PET/CT compared to multiparametric prostate MRI for primary staging of prostate cancer using final histopathology as a reference standard: A preliminary analysis of the multi-center prospective Next Generation Trial.

Introduction: Sensitivity and specificity of prostate-specific membrane antigen (PSMA)-ligand PET/CT have been validated in local staging and metastatic screening of intermediate- to high-risk prostate cancer, although the novel 18F-PSMA-1007 ligand is not yet approved for clinical use in Canada. Prior studies have compared 68Ga-PSMA-11 PET/CT with multiparametric MRI (mpMRI) demonstrating similar accuracy in intraprostatic tumor detection and localization but with variable reports of accuracy in detecting extraprostatic extension and seminal vesicle invasion. We aim to compare the diagnostic accuracy of 18F-PSMA PET-CT (PSMA PET) versus mpMRI in primary PCa staging using final histopathology as our reference standard.

Methods: The Next Generation Trial (NCT05141760) is a multi-center prospective phase II validating-paired cohort study in progress with 94 participants currently actively enrolled. Inclusion criteria include males over 18 years of age with newly diagnosed biopsy-proven intermediate- to high-risk prostate cancer and a Gleason Grade Group of 2 or greater. 75 of these participants have undergone both 18F-PSMA PET/CT and 3T mpMRI within a 14 day period followed by radical prostatectomy with pelvic nodal dissection at least 5 days after last imaging.

We conducted a preliminary analysis of the first 10 cases with completed reports for each modality. Each case was reviewed independently by blinded readers and reported according to PIRADS v2.1 for mpMRI, PSMA-RADS v1.0 for PSMA PET as per PROstate cancer Molecular Imaging Standardized Evaluation (PROMISE) standardized reporting, and Gleason grading for pathology.

Results: Of the 10 cases reviewed, only 2 were accurately staged by mpMRI with the remaining 8 cases primarily understaging pT3 disease. 6 of 10 cases were accurately staged by PSMA PET, however there were 2 cases where pT2 disease was overstaged. Overall, there was excellent near-total agreement in nodule localization with respect to sector mapping between modalities (with near-total agreement defined as up to 1 discrepancy in any direction). PSMA PET demonstrated better total agreement with pathology in nodule localization (4/10 versus 1/10) but with a tendency to overcall apex and base involvement in midgland lesions. mpMRI tended to undercall intraprostatic extent based on sector mapping and was poor at identifying presence or absence of extraprostatic extension (3/10 versus 8/10). Accuracy of reporting seminal vesicle invasion was excellent for both modalities.

All 10 dominant nodules found on pathology were Gleason grade 2 lesions. There was excellent concordance with PI-RADS scores of 4/5, although there was an isolated case of a Gleason 2 dominant nodule not detected on mpMRI but identified on PSMA PET. This lesion demonstrated an SUVmax of 7.5, a PSMA expression score of 1 (equal to or above blood pool, less than spleen), and a PSMA-RADS score of 3 (equivocal). Dominant nodules displayed variable PSMA expression scores ranging from 1 to 3 (low to high) with PSMA-RADS scores of 4/5 (likely/definitively cancer). SUVmax of the Gleason grade 2 dominant nodules ranged from 6.5 to 76 (median of 21). SUVmax of pathology-proven non-dominant nodules ranged from 5.9 to 16.7 (median of 9.5). Nondetection rates of several Gleason grade 1 non-dominant nodules found on final pathology were similar between PSMA PET and mpMRI.

There were no cases of pelvic nodal involvement within this preliminary dataset. This was accurately reported with both modalities.

Conclusions: Preliminary results show comparable accuracy in prostate cancer localization between 18F-PSMA-1007 PET/CT and multiparametric MRI. However PSMA PET/CT demonstrates greater accuracy in T-staging including detection of extraprostatic extension, even when EPE is microscopic. There were similar nondetection rates of Gleason grade 1 non-dominant nodules for both PSMA PET and mpMRI. Interestingly, there was 1 case of a pathology-proven Gleason 2 lesion identified on PSMA PET/CT but occult on MRI.