Abstract P6-21-09: muTarget.com: Linking gene expression and mutation status to identify patient cohorts eligible for targeted- and immunotherapy in breast cancer

Abstract

Abstract Selection of cancer patients for new targeted therapies reached its dead end as next generation sequencing based precision oncology approaches failed to deliver breakthrough improvements to oncology practice. The problem with current approaches is that the effect of a mutation can be indirect by influencing the expression of various other genes, which in turn can act as new therapy targets. A large-scale analysis of such cascades was not yet executed in breast cancer. Here, we developed an analysis tool to identify targetable genes showing an altered expression in relation to a mutation in other genes. The background database includes two independent large patient's cohorts, the TCGA and the Metabric datasets. Mutation status for each gene was determined using the VCF files from the TCGA repository. RNA-seq gene expression data for the same patients was re-normalized using a scaling normalization. Gene expression for the Metabric samples was determined using Illumina gene arrays and mutation status for the same patients is available for 174 selected genes. The Metabric database includes 1,981 patients and the TCGA breast cancer database contains 1,091 patients. Expression is linked with mutation status for each gene across all patients using Mann-Whitney test. A p<0.05 and a false discovery rate of <10% was accepted as significant. We demonstrate the utility of the analysis platform by using it to uncover patient cohorts with higher expression of PD1 (PDCD1) and PD-L1 (CD274). Immune checkpoint inhibitors permbrolizumab and nivolumab target PD1. PD-L1 inhibitors include atezolizumab, avelumab, and durvalumab. None of these immunotherapy agents is approved to be used in breast cancer. In both settings, only one gene reached statistical significance. For PD1, the best performing gene was NOP14. Patients with mutation in NOP14 (1.3% of patients) had a 2.08x increased expression (4.58 in mutated vs. 2.20 in wild type) of PD1 (p=8.4e-05, FDR=0.0239). For PD-L1, the strongest gene was CCDC88A (mutated in 2.6% of patients), which had a 2.03x increased expression (10.42 in mutated vs. 5.13 in wild type) of PD-L1 (p=6.2e-05, FDR=0.0147). Both NOP14 and CCDC88A have been linked to cancer development and progression, but have not been investigated in relation to immune therapies. One can anticipate that patients with mutation in these genes will be prone to respond to immune checkpoint inhibitors. In summary, an online portal was set up capable to identify genes with altered expression in relation to a given mutation. The presented approach can help to increase speed and reduce cost of development for future anticancer treatments. The analysis tool also enables identification of patient cohorts for new agents and is accessible at www.mutarget.com. Citation Format: Győrffy B, Nagy T. muTarget.com: Linking gene expression and mutation status to identify patient cohorts eligible for targeted- and immunotherapy in breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P6-21-09.