Pathogenic Genomic Alterations in Circulating Tumor DNA Predict Overall Survival in Men with Metastatic...

Area of Science:

• Genomic Medicine
• Prostate Cancer Research
• Liquid Biopsy Analysis

Background:

• Existing prognostic models for metastatic castration-resistant prostate cancer (mCRPC) lack genomic biomarker integration.
• Circulating tumor DNA (ctDNA) offers a non-invasive source for genomic profiling, including aneuploidy and pathogenic genetic alterations (PGAs).

Purpose of the Study:

• To determine the prevalence of PGAs in ctDNA.
• To assess the correlation between PGAs and ctDNA aneuploidy fraction.
• To evaluate the association of PGAs with overall survival (OS) and develop a predictive clinical-genetic (CG) model.

Main Methods:

• Analysis of ctDNA from 776 patients in the Alliance phase 3 trial (A031201) using the AR-ctDETECT assay for PGAs.
• Random survival forest for feature selection to identify key genomic alterations.
• Evaluation of the CG model's added value using net reclassification improvement (NRI) and time-dependent area under the receiver operating characteristic curve (tAUC).

Main Results:

• Identified specific gene alterations (e.g., AR, MYC, PTEN, TP53) associated with survival outcomes.
• The CG model demonstrated superior OS prediction accuracy (tAUC 0.77) compared to the clinical model (tAUC 0.72), with a significant NRI of 0.29.
• Patients stratified by CG model risk scores showed distinct median OS: 19.6 months (poor), 33.6 months (intermediate), and 60.8 months (low).

Conclusions:

• Integrating ctDNA PGAs into a CG model substantially enhances OS prediction in mCRPC patients.
• The developed CG model effectively classifies patients into risk groups, aiding clinical decision-making.
• This genomic approach is valuable for patient selection in future mCRPC clinical trials.