Comprehensive Summary
In this single-center study, Kang et al developed new machine learning (ML) models to analyze the intricacies of histological and molecular heterogeneity. U4tilizing twin biopsies from patients with primary prostate cancer, this group created a transcriptomics-based, cluster prediction model to better determine risk stratification of tumors and highlight drug vulnerabilities. Samples were processed in halves, one for histology and omics, with the other half processed for cell derivation, PDO establishment, genomic characterization, and drug screening.
Outcomes and Implications
A significant contribution of this study is the furthered understanding of multifocal prostate cancer, through this model’s interpretation of key tumor markers and activity levels. Determining prostate cancer qualities through this ML model could lead to improved patient-centered treatment. This experimental model can be employed in clinical trials, with the goal of supporting more efficient and efficacious therapeutic studies. Subsequent therapies and treatment plans developed with this growing database through the categorization of tumor qualities and drug responses can greatly improve patient outcomes.