2863. Concordance of PI-RADS v2.1 and PI-RR With PSMA PET After Radiation Therapy
Authors * Denotes Presenting Author
  1. Nathan Lee *; Cornell Medical College
  2. Nashwan Farooque; Cornell Medical College
  3. Elisabeth O'Dwyer; Cornell Medical College
  4. Sandra Huicochea Castellanos; Cornell Medical College
  5. Joseph Osborne; Cornell Medical College
  6. Daniel Margolis; Cornell Medical College
Prostate specific membrane antigen (PSMA) as a radiotracer for PET is increasingly being used in the setting of biochemical recurrence (BCR) or rising serum prostate specific antigen (PSA) after radiotherapy for prostate cancer. Multiparametric magnetic resonance imaging (mpMRI) has become a mainstay of detection of primary prostate cancer using prostate imaging reporting and data systems (PI-RADS) version 2.1, and a novel system has been proposed for the reporting of prostate mpMRI after radiation therapy: prostate imaging for radiation recurrence (PI-RR). The correlation between PSMA PET and both PI-RADS or PI-RR was investigated to determine the relative concordance.

Materials and Methods:
As part of a prospective clinical trial registered with ClinicalTrials.gov and with approval of the institutional review board (IRB) and compliance with HIPAA, the records of all subjects who underwent PSMA PET-MRI from October 5, 2017 through April 17, 2021 after radiation therapy with no prior prostatectomy were reviewed. Exclusion criteria were that either the PET or the mpMRI were not performed, or that images could not be interpreted by PI-RADS v2.1, PI-RR, or to determine standardized uptake value (SUV) for PET. Each case was prospectively assessed separately for PI-RADS and PI-RR by one radiologist blinded to the PET images, and the PET images were assessed by a second radiologist blinded to mpMRI data. The SUV maximum (SUVmax), PI-RADS, and PI-RR categories are recorded for each subject. Stepwise comparisons were performed using student’s t-test of SUVmax on PET versus separately grouped PI-RADS and PI-RR categories 1 - 2 versus 3 - 5, 1 - 3 versus 4 - 5, and 1 - 4 versus 5.

Of 57 records available, 3 were excluded for not having concomitant PSMA PET and mpMRI, for a total of 54 records analyzed. The average age was 69 years (range 47-89). PSA values were available for 33 subjects, with an average value of 18 ng/mL (range, 0.06-310). For PI-RADS, 5 were category 1 - 2, 10 were category 3, 16 were category 4, and 23 were category 5. For PI-RR, two were category 2, 4 were category 3, 17 were category 4, and 31 were category 5. Using t-test comparisons of grouped category cut-offs, there was a significant difference in the PSMA SUVmax for PI-RADS categories 1 - 2 vs 3-5 (average SUVmax 4.38 versus 20.89, p = 0.03) and between categories 1 - 3 versus 4 - 5 (4.63 versus 25.03, p = 0.03), and for PI-RR categories 1 - 2 versus 3 - 5 (4.19 versus 19.95, p = 0.03) and categories 1 - 3 versus 4 - 5 (6.20 versus 21.01, p = 0.05) but no significant difference for PI-RADS or PI-RR categories 1 - 4 versus 5.

There is a significant correlation between PSMA PET SUVmax and both PI-RADS and PI-RR. This suggests that PSMA PET may obviate the need for additional mpMRI in the case of BCR, and similarly, that salvage radiotherapy to the prostate could benefit from mpMRI for planning a boost in the absence of PSMA PET.