2318. The Use of Restriction Spectrum Imaging in the Evaluation of Gleason Pattern 4 Cribriform Morphology Prostate Cancer
Authors * Denotes Presenting Author
  1. Abhinav Patel; University of Rochester
  2. Gary Hollenberg *; University of Rochester
  3. Eric Weinberg; University of Rochester
The use of multiparametric MRI (mpMRI) of the prostate and MR/US fusion biopsy (TB) have significantly improved detection of clinically significant prostate cancer (PCa) compared to (systematic biopsy) SB alone. However, a subtype of prostate cancer, Gleason pattern 4 cribriform morphology is not well visualized, with only approximately 17% of prostate tumors with pure cribriform pathology detected on mpMRI of the prostate. Restriction spectrum imaging (RSI) is a newer diffusion-weighted imaging technique that is currently performed by post-processing mpMRIs. The major advantage of RSI over most other diffusion techniques is that it utilizes diffusion data collected with multiple gradient directions and b-values. RSI provides a more physiologic representation of water molecule diffusion which can potentially deliver improved detection of cribriform pattern 4 subtype prostate cancer. The purpose of this project is to determine if RSI can improve the detection of prostate cancer with cribriform morphology. This would then potentially allow for earlier detection, improved treatment and better clinical outcomes in patients with PCa.

Materials and Methods:
An IRB approved dataset obtained at a university hospital included 150 patients who underwent mpMRI of the prostate and subsequent biopsy or prostatectomy. Inclusion criteria required pathology proven cribriform Gleason pattern 4 subtype. Two senior Radiology faculty, with significant clinical experience in reading prostate MRI studies, will independently analyze the standard mpMRI sequences (including routine high b value DWI) without and with the new RSI sequence. The cases will be randomized and evaluated at different times to reduce familiarity and recall from the cases. Results will then be compared to the original dictated reports for these cases. The DICOM images will be post-processed to generate the RSI dataset for each case.

Preliminary results are very promising after the review of initial cases. Two groups were examined: an experimental group of 4 patients with cribriform pathology not detected on conventional mpMRI and a control group with 4 patients with prostate cancer diagnosed on mpMRI. After the post processing RSI were reviewed independently by the Radiology faculty, all cases with cribriform pathology (which were initially not visualized) were successfully identified (100% sensitivity) and all cases of prostate cancer, regardless of subtype were successfully identified (100% sensitivity) based off RSI imaging alone. The final review, which will be conducted from August of 2020 to September of 2020, will include a cohort of 150 patients.

The utilization of the RSI imaging as part of mpMRI improves the detection of cribriform Gleason pattern 4 subtype prostate cancer when compared with standard high B value diffusion-weighted imaging. The addition of RSI to mpMRI of the prostate, could improve early tumor detection resulting in better clinical outcomes.