Identifying Patient-Specific Neoepitopes for Cell-Based And

Poster: P2-04-26 Identifying patient-specific neoepitopes for cell-based and vaccine immunotherapy across breast cancer classifications reveals rarely shared recurrent neoepitopes Andy Nguyen,1 J Zachary Sanborn,1 Charles J Vaske,1 Shahrooz Rabizadeh,2 Kayvan Niazi,2 Patrick Soon-Shiong,2,3 Steven C Benz1 1NantOmics LLC, Santa Cruz, CA; 2NantOmics LLC, Culver City, CA; 3CSS Institute of Molecular Medicine, Culver City, CA Background Results Filtering High Quality Neoepitopes in TNBC Filtering High-Quality Neoepitopes Across Cancers • Immunotherapies such as checkpoint inhibitors, CAR T cells, Cancer Neoepitope Loads Across TCGA Dataset 20000 250 NK cells, and therapeutic vaccines are revolutionizing cancer Neoepitopes (via WGS) 18000 A*02:01 Restricted Samples Neoepitopes 250000 2000 medicine with remarkable responses in some patients. 16000 200 UCEC (14) 6% 1800 14000 THCA (10) 200000 1600 STAD (1) • Several cancers have FDA approval for the use of check point 12000 150 1400 READ (4) inhibitors. And publications have suggested that mutational 10000 150000 1200 PRAD (10) 8000 100 burden and neoepitope burden predict response to check LUSC (15) 1000 6000 LUAD (17) 94% point inhibitors. Neoepitope Counts 100000 800 4000 50 KIRC (13) 600 • Cancers such as TNBC have few treatment options and would 2000 HNSC (25) 50000 400 COAD (3) 0 0 greatly benefit from novel immunotherapy approaches. Detection: 200 BRCA (28) Non-Cancer Genes Coding Variants DNA + + + +

Neoepitope Counts per Cancer Neoepitope Counts per Cancer BLCA (3) RNA - + + + 0 0 • We analyzed whole genome sequencing (WGS) and RNA Detection: Cancer Driving netMHC - - + + DNA + + + + Genes sequencing (RNAseq) data from The Cancer Genome Atlas RNA HUGO Protein Bound Bind - + + + TCGA Barcode UCSC id TPM Neoepitope Normal netMHC - - + + (TCGA) to identify neoepitopes (tumor-speciﬁc antigens Gene Change Allele Strength

derived from somatic tumor mutations) that could be exploited Counts per MB Coding DNA TCGA-E2-A14X uc003ean.2 NAA50 229.85 PTDAHVLQK p.A145T PADAHVLQK A*11:01 146nM Shared Neoepitopes Across Cancers to develop next-generation, patient-specific cancer HUGO Protein TCGA Barcode UCSC id Neoepitope Normal Cancers immunotherapies. TCGA-E2-A1LL uc001asj.3 FBXO2 187.36 LLLHVLAAL p.R57H LLLRVLAAL A*02:01 18nM Gene Change TCGA-E2-A109, TCGA-CR-5249, (3) HNSC, uc001wxt.2 SOS2 YIHTHTFYV p.T390I YTHTHTFYV TCGA-BA-6872, BRCA Methods Filtering High-Quality Neoepitopes in HER2+ BRCA TCGA-CN-6989 • TCGA WGS and RNASeq data were obtained from the TCGA-EW-A1J5, LUAD, 12000 TCGA-21-1082, BLCA, University of California, Santa Cruz (UCSC) Cancer Genomics uc001zyl.4 USP8 SQIWNLNPV p.R763W SQIRNLNPV 10% 600 TCGA-GD-A2C5, LUSC, Hub (https://cghub.ucsc.edu/). 10000 500 TCGA-75-5147 BRCA • Neoepitopes were identified by creating all possible Normal Epitope 8000 400 Conclusions permutations of either 9-mer or 15-mer amino acid strings • Nearly all identified neoepitopes are patient-specific. TNBC samples do not Neoepitope 6000 derived from somatic single nucleotide variants (SNVs) or 300 share any common neoepitopes. 90% 4000 200 • Neoepitope-MHC interactions restrict more commonly shared mutations. insertions/deletions (indels) in coding regions. Neoepitope Counts • Potential neoepitopes were filtered against all possible 9-mer 2000 100 • Development of personalized immunotherapies is dependent on accurate Clonality of Neoepitopes across cancer classifications DNA and RNA sequencing. and 15-mer sequences from reference human coding genes, 0 0 Detection: • Checkpoint inhibitor markers such as neoepitope load or PD1 expression will in addition to all possible variation in dbSNP (http:// DNA + + + www.ncbi.nlm.nih.gov/SNP) sites. RNA - - + need to be screened before suggesting use of immunotherapy for breast netMHC - + + cancer patients. • In-silico HLA typing was performed using WGS and RNAseq A Single Recurrent Neoepitope in TCGA HER2+ BRCA Use a QR Scanner to data by alignment to the IMGT/HLA database. Typing results Acknowledgement download this Poster HUGO Protein Bound Bind TCGA Barcode UCSC id TPM Neoepitope Normal We would like to thank Kathryn Boorer, PhD, of were obtained for HLA-A, HLA-B, HLA-C, and HLA-DRB1. Gene Change Allele Strength NantHealth, LLC for writing assistance. TCGA-BH-A18R 21.39 C*03:03 131nM • NetMHC 3.4 (http://www.cbs.dtu.dk/services/NetMHC-3.4/) was Contact Copies of this poster obtained through Quick uc003ean.2 FANCD2 FAKDGGLVT P714L FAKDGGPVT FDA Approved use of Response (QR) Code are for personal use only used to predict MHC to neoepitope binding affinities. checkpoint inhibitors Corresponding Author: TCGA-AO-A0JM 14.12 C*05:01 851nM and may not be reproduced without [email protected] permission from the authors of this poster.