Related Experiment Video
Updated: Jun 19, 2026
MicroRNA Detection in Prostate Tumors by Quantitative Real-time PCR qPCR
Published on: May 16, 2012
A New Radiotranscriptomic Approach to Analyze Combined Sets of T3b Stage-Specific Genes and Radiomic Features in
Qian Yang1,2, Peng Tang3, Jiao Mo1
1Department of Ultrasound, Air Force Medical Center, Air Force Military Medical University, Beijing, China.
Background:
Current clinical staging of prostate cancer (PCa) using the tumor-node-metastasis (TNM) system and serum biomarkers remains limited in distinguishing locally advanced (T3b) PCa from organ-confined (T2c) disease.
Aims:
Building on our previous biomarker discovery in differentiating PCa from that of benign prostatic hyperplasia, this study pioneers a radiotranscriptomic model to distinguish T3b stage PCa from T2c stage PCa by integrating contrast-enhanced ultrasound (CEUS) radiomics with stage-specific transcriptomic signatures, addressing a critical knowledge gap in precision staging.
Methods And Results:
This prospective study was approved by the review board of Chinese PLA General Hospital (S2021-565-01), and all participants provided written informed consent. Transrectal B-mode ultrasound images and contrast-enhanced ultrasound images on two imaging planes were prospectively analyzed in 48 patients with biopsy-confirmed PCa (35 patients with stage T2c and 13 with stage T3b). Textural features were evaluated using microvascular ultrasonography and contrast-enhanced ultrasound. Radiomic data were then retrieved from all modes. An across-the-board investigation of mRNA and miRNA expressions was also performed in the two PCa stages. Six biomarkers (frizzled 4, ribosomal protein S7, ribosomal protein L29, miR-374c, miR-9, and miR-6510) were identified to differentiate T3b stage from T2c stage. The area under the curve (AUC) values of the combined set (AUC = 0.887, 0.956, and 0.996 for random forest, naïve Bayes, and support vector machine, respectively) and radiomic features alone (AUC = 0.921, 0.957, and 0.998, respectively) were found to be more accurate than those of the transcriptomic data alone (AUC = 0.583, 0.716, and 0.898, respectively) or clinical features alone (AUC = 0.585, 0.675, and 0.953, respectively). The PCa gene regulatory network comprised of four miRNAs (miR-148, miR-141, miR-342, and miR-210) may contribute to accelerating tumor progression.
Conclusion:
We established the new radiotranscriptomic signatures specifically optimized for differentiating T3b stage from T2c stage by decoding stage-specific imaging-genomic crosstalk. This new approach may overcome TNM staging limitations.
