Related Experiment Video
Updated: Jul 25, 2025
11:29
miRNA Expression Analyses in Prostate Cancer Clinical Tissues
Published on: September 8, 2015
10.9K
Radiogenomics Analysis Linking Multiparametric MRI and Transcriptomics in Prostate Cancer.
Catarina Dinis Fernandes1, Annekoos Schaap1, Joan Kant2
1Electrical Engineering Department, Eindhoven University of Technology, 5600 MB Eindhoven, The Netherlands.
Cancers
|June 28, 2023
Summary
This study links advanced MRI imaging features with gene activity in prostate cancer (PCa). Quantitative imaging, particularly MRDI A, correlates with transcription factors STAT6 and TFAP2A, aiding in non-invasive assessment of PCa aggressiveness.
Area of Science:
- Oncology
- Radiology
- Genomics
Background:
- Prostate cancer (PCa) prognosis is heterogeneous, necessitating accurate aggressiveness assessment for tailored treatments.
- Current methods rely on invasive biopsies, highlighting the need for non-invasive approaches.
- Radiogenomics offers a promising avenue by integrating imaging and genomic data.
Purpose of the Study:
- To identify imaging and transcriptomic features associated with clinically significant PCa (ISUP grade ≥ 3).
- To evaluate the correlation between these imaging and transcriptomic features.
- To explore the potential of radiogenomics for non-invasive PCa aggressiveness assessment.
Main Methods:
- Parallel analysis of multi-parametric MRI (T2W, DWI, DCE) textural features and MRDI-derived pharmacokinetic maps with RNA sequencing data.
- Machine learning models trained to classify PCa into clinically insignificant or significant categories.
- Correlation analysis between selected imaging and transcriptomic features.
Main Results:
- Five significant correlations (p < 0.05) were identified between imaging and transcriptomic features.
- The MRDI A median (perfusion-based imaging feature) showed strong negative correlations with STAT6 (-0.64) and TFAP2A (-0.50) activity.
- A T2W textural feature also correlated significantly with STAT6 activity (-0.58).
Conclusions:
- Quantitative imaging features, including texture and perfusion-based MRDI A, can predict clinically significant PCa.
- The MRDI A feature demonstrates a strong link to underlying transcriptomic information, specifically STAT6 and TFAP2A.
- These findings support radiogenomics as a tool for non-invasive PCa aggressiveness evaluation.
