Luke, J. J. | bioRxiv Science

bioRxiv2020

Tumor Neoantigenicity Assessment with CSiN Score Incorporates Clonality and Immunogenicity to Predict Immunotherapy Outcomes

Lack of responsiveness to checkpoint inhibitors is a central problem in the modern era of cancer immunotherapy. Tumor neoantigens are critical mediators of host immune response and immunotherapy treatment efficacy. Current studies of neoantigens almost entirely focus on total neoantigen load, which simplistically treats all neoantigens equally. Besides, neoantigen loads have been linked with treatment response and prognosis only in some studies, but not others. We developed a Cauchy-Schwarz index of Neoantigens (CSiN) score to characterize the degree of concentration of immunogenic neoantigens in truncal mutations. Unlike simple neoantigen loads, CSiN incorporates the effect of both clonality and MHC-binding affinity of neoantigens when characterizing patient neoantigen profiles. By exploiting the clinical responses in 501 treated patients (mostly by checkpoint inhibitors) and the overall survival of 1,978 baseline patients, we showed that CSiN scores predict treatment response to checkpoint inhibitors and prognosis in melanoma, lung cancer, and kidney cancer patients. CSiN substantially outperforms prior genetics-based prediction methods of responsiveness. Overall, our work fulfilled an important gap in current research involving neoantigens. One Sentence SummaryThe quality of tumor neoantigens predicts response to immunotherapy

bioinformatics↗

bioRxiv2019

Molecular correlates and therapeutic targets in T cell-inflamed versus non-T cell-inflamed tumors across cancer types

The T cell-inflamed tumor microenvironment, characterized by CD8 T cells and type I/II interferon transcripts, is an important cancer immunotherapy biomarker. Tumor mutational profile may also dictate response with some oncogenes (i.e. WNT/{beta}-catenin) known to mediate immuno-suppression. Building on these observations we performed a multi-omic analysis of human cancer correlating the T cell-inflamed gene expression signature with the somatic mutanome and transcriptome for different immune phenotypes, by tumor type and across cancers. Strong correlations were noted between mutations in oncogenes and non-T cell-inflamed tumors with examples including IDH1 and GNAQ as well as less well-known genes including KDM6A, CD11c and genes with unknown functions. Conversely, we observe many genes associating with the T cell-inflamed phenotype including VHL and PBRM1, among others. Analyzing gene expression patterns, we identify oncogenic mediators of immune exclusion broadly active across cancer types including HIF1A and MYC. Novel examples from specific tumors include sonic hedgehog signaling in ovarian cancer or hormone signaling and novel transcription factors across multiple tumors. Using network analysis, somatic and transcriptomic events were integrated, demonstrating that most non-T cell-inflamed tumors are influenced by multiple pathways. Validating these analyses, we observe significant inverse relationships between protein levels and the T cell-inflamed gene signature with examples including NRF2 in lung, ERBB2 in urothelial and choriogonadotropin in cervical cancer. Finally, we integrate available databases for drugs that might overcome or augment the identified mechanisms. These results nominate molecular targets and drugs potentially available for immediate translation into clinical trials for patients with cancer.

cancer biology↗