AUA 2019: Correlation Between MRI Phenotypes and a Genomic Classifier of Prostate Cancer

Chicago, IL ( In this study the authors attempted to evaluate the association between MRI phenotypes of prostate cancer as defined by the Prostate Imaging Reporting and Data System version 2 (PI-RADS v2) and Decipher Genomic Classifier (used to estimate the risk of early metastases).

This was a single-center, retrospective study with 72 prostate cancer patients who underwent 3T multiparametric MRI (mpMRI) before radical prostatectomy between April 2014 and August 2017.  In all patients, the lesions were micro-dissected from the radical prostatectomy specimens and tested with Decipher (a total of 89 lesions; 23 mpMRI invisible [PI-RADS v2 scores <= 2] and 66 MRI visible [PI-RADS v2 scores >= 3]). Linear regression analysis was used to assess clinicopathologic and MRI predictors of Decipher results, and correlation coefficients (r) were used to quantify these associations. Area under the receiver operating characteristic curve (AUC) was used to determine whether PI-RADS v2 could accurately distinguish between low-risk and intermediate-/high-risk lesions (using a DECIPHER cutoff score of 0.45).

The median age of the included patients was 63 years (range: 42-76), and median PSA was 9.8 ng/mL (range: 1.2-69). The lesions' Gleason grade groups (GGG) were 44% GGG 2 and 56% GGG>=3. It was shown that mpMRI-visible lesions had higher Decipher scores than mpMRI-invisible lesions (mean difference 0.22; 95% CI 0.13, 0.32; p < 0.05). Most MRI-invisible lesions (82.6%) were low risk. The AUC for PIRADS score (0.776) was slightly higher than the AUC for GGG (0.739), p=0.446, to distinguish between low and intermediate/high Decipher group lesions. The AUC for PIRADS (0.863) was significantly higher than the AUC for GGG (0.780), p=0.018 for peripheral zone lesions. The correlation (r) between the lesion’s characteristics and Decipher risk group is shown in table 1.

Table 1 – Correlation between the lesion’s characteristic and Decipher risk group:

AUA 2019 Decipher risk group
In conclusion, mpMRI phenotypes of prostate cancer are positively associated with Decipher risk groups. Most, but not all mpMRI-invisible lesions had a low risk for metastasis. mpMRI could be used in conjunction with genomic assays to identify lesions that carry the biological potential for early metastasis.

Presented by: Andrei Purysko, MD, Cleveland Clinic, Cleveland, Ohio 

Written by: Hanan Goldberg, MD, Urologic Oncology Fellow (SUO), University of Toronto, Princess Margaret Cancer Centre @GoldbergHanan at American Urological Association's 2019 Annual Meeting (AUA 2019), May 3 – 6, 2019 in Chicago, Illinois