SIU 2017: Radiogenomics, MRI & Prostate Cancer Imaging

Lisbon, Portugal ( Dr. Reiter presented on the use of genomics in conjunction with imaging to make management decisions in prostate cancer; this is a new field called radiogenomics. 

He started by reviewing some of the talks from earlier in the localized prostate cancer session, focusing on the use of markers. He coupled this with what we know already about the utility of mpMRI in the evalution and management of PCa. However, the combination of these two may provide greater utility than either alone. 

Questions that arise:

1) What is the role of genomic testing and how should it be utilized in the setting of mpMRI and targeted biopsy?
  • Does it help risk stratify?
  • Does it provide information independent of PIRADS scoring?

2) What are the biologic determinants that underlie MRI findings?
  • Can an integrated radiogenomic prognostic/predictive tool be created? Radiomics?
Oncotype Dx Genomic Prostate Score (GPS)
  • 17 gene RT-PCR based assay
  • Validated as an independent predictor of adverse pathology at the time RP – Gleason 4+3 or higher, pT3+
  • Dr. Reiter performed a study using after having had an mpMRI-guided systematic biopsy prior to RP, done in men with low/intermediate risk prostate cancer
           o   On MVA, GPS score (more than PIRADS score) was the strongest predictor of adverse pathology – independent of traditional clinicopathologic features
           o   Gleason score + GPS AUC 0.79 – essentially the same when you add MRI
                -Hence MRI didn’t necessarily add much (for low/intermediate risk disease)
           o   But GPS score did correlate with PIRADS score

Radiomics – association with quantitative imaging features with genomics
  • Two studies (UCLA and Miami)
  • UCLA study – MRI evaluation of the prostate and MRI-guided biopsy of the prostate in 6 patients. 
            o   IHC and gene expression analysis from the prostate tissue
            o   Correlate with detailed examination of the MRI imaging and mapping of the prostate
            o   WES of the normal appearing tissue and concerning areas
            o   They found that WES identified cancer related mutations both in the tumor and in areas thought to be normal – supports concept of “field effect” in prostate cancer
            o But no correlation with MRI features and specific mutations

  • Miami study –
            o   Prostate cancer associated signatures on Decipher, GPS and Prolaris correlated with radiomic features
            o   TZ and PZ have distinct radiomic features
            o   ADC values most significantly associated with distinct biologic processes

More work is needed in this features.

Presented by: Robert Reiter

Written by: Thenappan Chandrasekar, MD, Clinical Fellow, University of Toronto, Twitter: @tchandra_uromd at the 37th Congress of Société Internationale d’Urologie - October 19-22, 2017- Lisbon, Portugal
email news signup