Developing a Biomarker of Homologous Recombination Deficiency for Enhanced PARP Inhibitor Efficacy Across Cancers

Homologous recombination deficiency (HRD) occurs when double-strand breaks in the genome are not repaired properly due to a defect in the homologous recombination repair pathway. As HRD confers sensitivity to poly(adenosine diphosphate [ADP]-ribose) polymerase (PARP) inhibitors and platinum-based chemotherapy, there is a great deal of interest in the accurate identification of patients who harbor HRD. We have previously developed an algorithm called SigMA (Signature Multivariate Analysis), which can potentially predict HRD status more accurately and less expensively compared to commercially available tests. In this project, we aim to (i) improve the accuracy of the algorithm and extend its applicability to more tumor types by training on new datasets (a 10-fold increase compared to the datasets used for the current version) and incorporating additional genomic information (indel signatures as well as single base substitution signatures); (ii) make the software more robust by implementing automated tuning of parameters for different sequencing platforms; and (iii) incorporate SigMA in CLIA-certified laboratories at HMS-affiliated and other hospitals, and engage in collaborations with diagnostic and pharmaceutical companies interested in supplementing their panels with additional biomarkers or extending the applicability of PARP inhibitors to other HRD indications.