Evaluative Methodology for HRD Testing: Development of Standard Tools for Consistency Assessment

Area of Science:

• Oncology
• Genomics
• Biomarker Development

Background:

• Homologous recombination deficiency (HRD) is a key prognostic and predictive biomarker in cancer.
• Current HRD testing methods, particularly targeted panels, yield inconsistent results, necessitating standardized benchmarks.
• There is an urgent need for reliable evaluation tools for HRD assay performance.

Purpose of the Study:

• To develop and validate standardized reference materials and datasets for homologous recombination deficiency (HRD) testing.
• To establish benchmarks for evaluating the accuracy and reproducibility of HRD assays.
• To identify challenges in HRD detection and improve assay consistency.

Main Methods:

• Development of ten pairs of DNA reference materials from cancer and normal cell lines with varying cancer-to-normal mass ratios.
• Creation of reference datasets for allele-specific copy number variations (ASCNVs) and HRD scores using multiple sequencing methods and analytical pipelines.
• Validation of four HRD panel assays using the established reference materials and datasets.

Main Results:

• Reference datasets for ASCNVs and HRD scores were established, covering a significant portion of the human genome.
• Genomic Instability Scores (GIS) ranged from 11 to 96, suitable for various validation thresholds.
• Significant variability was observed in HRD panel assays regarding ASCNV detection and HRD scores, with limited concordance to reference data.

Conclusions:

• The developed HRD reference materials and datasets provide a robust toolkit for objective HRD assay evaluation.
• The study highlights significant variability in current HRD testing, underscoring the need for standardization.
• Addressing challenges in ASCNV detection and the impact of high ploidy is crucial for improving HRD assay consistency.