Comprehensive comparison of homologous recombination deficiency predictors in early-stage triple-negative breast cancer.
Summary
This study comprehensively compared seven homologous recombination deficiency (HRD) prediction methods in 235 patients with early-stage triple-negative breast cancer. The methods included sequencing-based, copy number-based, functional, mRNA-based, and image-based approaches, with an exploratory comparison to the FDA-approved Myriad myChoice CDx assay. Results revealed substantial concordance among methods, alongside specific discordances attributed to data preprocessing and training strategies. Despite these variations, all methods demonstrated similar prognostic performance in patients treated with adjuvant chemotherapy. The study emphasizes the critical need for rigorous optimization of data processing workflows and threshold definitions to ensure consistency and reproducibility of HRD classifications.
Analysis
🔴 CLINIQUE: This retrospective study on a population-representative cohort of early-stage triple-negative breast cancer patients evaluates the prognostic performance of HRD predictors. While all methods showed similar prognostic performance for invasive disease-free survival in patients receiving adjuvant chemotherapy, the variability among HRD classification methods highlights a challenge for clinical integration. Standardization is needed before these tools can reliably guide therapeutic decisions, particularly for identifying patients eligible for PARP inhibitors. The time to clinical impact is estimated at 3-5 years, pending prospective validation and test standardization. 🟢 BIOMOL: The study compared various approaches to assess HRD, including methods based on whole genome sequencing (WGS), RNA-seq, immunohistochemistry (RAD51 foci), and H&E imaging. Discordance between methods is often linked to technical aspects such as tumor cell content, sequencing depth, and fundamental data processing steps. mRNA- and image-based methods may incorporate signals not specific to HRD, raising questions about their analytical specificity. The study highlights the need for rigorous analytical validation for each method to ensure their robustness and translational relevance. 🔵 BIOINFO: Bioinformatic analysis revealed that sequencing-based methods (HRDetect, CHORD) and copy number-based methods (scarHRD) showed the greatest classification agreement. Discordances are attributable to model training strategies that insufficiently account for breast cancer heterogeneity, as well as data preprocessing steps like segmentation. The study emphasizes the importance of optimizing data processing workflows and threshold definitions to improve the reproducibility and comparability of HRD classification platforms. This is crucial for future clinical deployment, where these tools could help prioritize patients for targeted therapies.