Metabolomic-driven prediction of the mutational status of healthy individuals with a family history of hereditary breast and ovarian cancer syndrome: the HRRmet study

Identifier:  imarina:9523789

Handle: https://hdl.handle.net/20.500.11797/imarina9523789

Authors:  Roig, Barbara; Fernandez-Castillejo, Sara; Guma, Josep; Badia, Joan; Mele, Mireia; Salvat, Monica; Querol, Montserrat; Cumeras, Raquel; Rodriguez-Balada, Marta

Abstract:
Pathogenic variants (PVs) identified in genes involved in the DNA homologous recombination repair (HRR) mechanism are the main cause of hereditary breast and ovarian cancer syndrome (HBOC). The main objective of this study was to identify differential plasma metabolomic profiles associated with the HRR genotype in healthy individuals. Cascade testing was performed by Sanger sequencing in healthy carrier and noncarrier individuals with a familial history of HBOC. PVs associated with HRR genes (BRCA1, BRCA2, PALB2, ATM, CHEK2 and RAD51) were identified. Untargeted metabolomics of plasma samples was performed by liquid chromatography coupled with mass spectrometry. Predictive models were developed using a machine learning approach. Thirty-one metabolites were selected to create the global predictive model (accuracy 61.9%), whereas the gene-specific models had a better performance (accuracy > 80%) and were constructed with fewer metabolites. The present study is the first to characterize the phenotype associated with the HRR-deficient genotype in healthy individuals with a familial history of HBOC. Metabolomic profiles may be useful for differentiating carriers from noncarriers of PVs in the HRR genes, and therefore, with potential predictive capacity of the HRR germline mutational status.