Abstract 125P
Background
Presence of homologous recombination repair deficiency (HRD) in cancer cells is a biomarker for predicting responses to anticancer treatments acting through synthetic lethality, such as PARP inhibitors. HRD detection relies on functional or genomic approaches measuring genomic scars or genomic instability resulting from lack of functioning repair mechanisms. We evaluated 2 whole-genome microarray data analysis approaches for detecting HRD in cancer samples, compared to next-generation sequencing (NGS).
Methods
In total 212 FFPE samples were tested with OncoScan™ CNV Plus Assay for Research. Data was analyzed using 2 different approaches based on previously published methods: MTD1 (based on whole genome doubling events) and MTD2 (based on copy number analysis) with Chromosome Analysis Suite Version 4.3.0.71, Rstudio Version 2023.09.1+494, and R version 4.3.2. For all samples the HRD score obtained using a commercially available NGS-based assay was available for comparison. The sample cohort was split into the training (N=124 samples) and validation cohort (N=88 samples). Positive (PPA), negative (NPA) and overall percent agreements (OPA) between the microarray and commercial NGS-based approach were calculated.
Results
Analysis of 124 samples using MTD1 detected 30 samples with HRD. When compared to the commercially available HRD test the OPA was 88.7% (95% CI, 81.9-93.1%), with PPA of 80.8% (95% CI, 62.1-91.5%) and NPA of 90.82% (95% CI, 83.5-95.1%). MTD2 analysis approach of the same sample set detected HRD in 24 samples, reaching an OPA of 91.9% (95% CI, 85.8-95.6%), with PPA of 76.9% (95% CI, 57.9-89.0%) and NPA of 95.92% (95% CI, 90.0-98.4%). The analysis of the second sample cohort of 88 samples with MTD2 detected 41 samples with HRD and yielded similar agreement to the commercially available HRD test with OPA of 84.4% (95% CI, 75.1-90.3%), PPA of 80.0% (95% CI, 66.18-89.1%) and NPA of 88.37% (95% CI, 75.5-94.9%).
Conclusions
The Applied BiosystemsTM OncoScanTM CNV Plus Assay for Research offers a reliable and cost-effective alternative to NGS-based approaches for detection of HRD and can contribute to expanding access to HRD testing for cancer research.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Thermo Fisher Scientific.
Disclosure
C. Kidwell, A. Crawford: Financial Interests, Personal, Full or part-time Employment: Quantigen. M. Napier, A. Roter, D. Woo, J.S. Milosevic Feenstra: Financial Interests, Personal, Full or part-time Employment: Thermo Fisher Scientific.
Resources from the same session
112P - Reporting of molecular test results from cell-free DNA analyses: Expert consensus recommendations from the 2023 European Liquid Biopsy Society ctDNA workshop
Presenter: Vincent de Jager
Session: Poster session 08
114P - Prevalence and landscape of pathogenic or likely pathogenic germline variants in cancer predisposition genes among selected patients with lung adenocarcinoma
Presenter: Oscar Gerardo Arrieta Rodriguez
Session: Poster session 08
115P - Gene rearrangements, actionability and access to precision medicine: Results from the ARCAGEN study
Presenter: Marie Morfouace
Session: Poster session 08
116P - Single-cell RNA sequencing reveals a subset of FSIP1 cancer cells and verified its value of prognosis in lung adenocarcinoma
Presenter: Xiaochen Zhang
Session: Poster session 08
Resources:
Abstract
117P - Methylome and transcriptome profiling of hepatoid adenocarcinoma of the stomach
Presenter: Shirong Zhang
Session: Poster session 08
Resources:
Abstract
118P - Comparative analysis of DNA and RNA-based NGS for detecting MET exon 14 skipping mutation in pan-solid tumor samples
Presenter: Ruijun Cai
Session: Poster session 08
119P - Predicting the pathogenicity of novel fusion genes and explaining reasons using a large language model: A focused assessment
Presenter: Katsuhiko Murakami
Session: Poster session 08
120P - A prospective comparative evaluation of automatic trial match tools in a molecular tumor board
Presenter: Lilia GUEGUEN
Session: Poster session 08