AACR 2022 Proceedings: Part A Online-Only and April 10

By CTI Meeting Technology

The AACR Annual Meeting is the focal point of the cancer research community, where scientists, clinicians, other health care professionals, survivors, patients, and advocates gather to share the latest advances in cancer science and medicine. From population science and prevention; to cancer biology, translational, and clinical studies; to survivorship and advocacy; the AACR Annual Meeting highlights the work of the best minds in cancer research from institutions all over the world.

• Language: English
• Publisher: CTI Meeting Technology
• Release date: May 13, 2022
• ISBN: 9781005372798

Book preview

Proceedings of the AACR

Volume 61 | April 2022

Part A: Online-Only Presentations and Sunday, April 10

TABLE OF CONTENTS

Online-Only Poster Presentations

BIOINFORMATICS / CONVERGENCE SCIENCE / SYSTEMS BIOLOGY

BIOINFORMATICS AND COMPUTATIONAL BIOLOGY

CONVERGENCE SCIENCE AND SYSTEMS BIOLOGY

CHEMISTRY

DRUG DISCOVERY, DESIGN, AND DELIVERY

PROTEOMICS AND MASS SPECTROMETRY

CLINICAL RESEARCH EXCLUDING TRIALS

BIOMARKERS

BIOSTATISTICS IN CLINICAL TRIALS

CLINICAL ENDOCRINOLOGY

CLINICAL OUTCOMES RESEARCH

CLINICAL RESEARCH IN SPECIAL POPULATIONS

IMMUNO-ONCOLOGY

PEDIATRIC CANCER: CLINICAL INVESTIGATIONS

PHASE I CLINICAL TRIALS

PRECISION ONCOLOGY

RADIATION ONCOLOGY

REAL-WORLD DATA (RWD) AND REAL-WORLD EVIDENCE (RWE)

SUPPORTIVE CARE AND SURVIVORSHIP RESEARCH

SURGICAL ONCOLOGY

TRANSLATIONAL RESEARCH

CLINICAL TRIALS

PHASE I TRIALS IN PROGRESS

PHASE II CLINICAL TRIALS

PHASE II TRIALS IN PROGRESS

PHASE III CLINICAL TRIALS

PHASE III TRIALS IN PROGRESS

COVID-19 AND CANCER

COVID-19 AND CANCER

ENDOCRINOLOGY

MOLECULAR, PRECLINICAL, AND CLINICAL ENDOCRINOLOGY

EXPERIMENTAL AND MOLECULAR THERAPEUTICS

BIOLOGICAL THERAPEUTIC AGENTS

DRUG DISCOVERY

DRUG RESISTANCE

GENE AND VECTOR-BASED THERAPY

MECHANISMS OF DRUG ACTION

MOLECULAR TARGETS

PHARMACOLOGY, PHARMACOGENETICS, AND PHARMACOGENOMICS

PRECLINICAL RADIOTHERAPEUTICS

SMALL MOLECULE THERAPEUTIC AGENTS

IMMUNOLOGY

PRECLINICAL IMMUNOTHERAPY

TUMOR IMMUNOBIOLOGY

MOLECULAR/CELLULAR BIOLOGY AND GENETICS

CELL CYCLE

CELL DEATH

CELL GROWTH SIGNALING PATHWAYS

CELLULAR STRESS RESPONSES

DNA DAMAGE AND REPAIR

EPIGENETICS AND EPIGENOMICS

GENE REGULATION AND TRANSCRIPTION FACTORS

GENOMICS

METABOLISM AND CANCER

MICRORNAS AND OTHER NONCODING RNAS

ONCOGENES AND TUMOR SUPPRESSOR GENES

MULTIDISCIPLINARY

AACR PROJECT GENIE USE CASES

POPULATION SCIENCES

POPULATION SCIENCES

PREVENTION / EARLY DETECTION / INTERCEPTION

CLINICAL PREVENTION, EARLY DETECTION, AND INTERCEPTION

IMPLEMENTATION SCIENCE

PRECLINICAL PREVENTION, EARLY DETECTION, AND INTERCEPTION

REGULATORY SCIENCE AND POLICY

REGULATORY SCIENCE AND POLICY / SCIENCE AND HEALTH POLICY

SURVIVORSHIP

ADVOCATES POSTER SESSION: ONLINE-ONLY

TUMOR BIOLOGY

ANGIOGENESIS

CARCINOGENESIS

IN VIVO IMAGING

METASTASIS

NONCLINICAL MODELS OF CANCER

PEDIATRIC CANCER: BASIC SCIENCE

RADIATION SCIENCE

STEM CELL BIOLOGY

TUMOR ADHESION

TUMOR EVOLUTION AND HETEROGENEITY

TUMOR MICROENVIRONMENT

PRESENTATIONS: SUNDAY, APRIL 10

BIOINFORMATICS / CONVERGENCE SCIENCE / SYSTEMS BIOLOGY

EMERGING TOPICS IN COMPUTATIONAL ONCOLOGY

ARTIFICIAL INTELLIGENCE AND DIGITAL PATHOLOGY

CANCER SYSTEMS BIOLOGY

CHEMISTRY

NANOTECHNOLOGY DRUG DELIVERY

CLINICAL RESEARCH EXCLUDING TRIALS

LATE-BREAKING RESEARCH: CLINICAL RESEARCH 1

BIOMARKERS 1

CLINICAL RESEARCH IN SPECIAL POPULATIONS / SUPPORTIVE CARE AND SURVIVORSHIP RESEARCH

PROGNOSTIC BIOMARKERS 1

PROGNOSTIC BIOMARKERS 2

CLINICAL TRIALS

BIOMARKER ADVANCES IN CLINICAL TRIALS

CLINICAL TRIALS OF CELLULAR IMMUNOTHERAPIES

CLINICAL TRIALS TARGETING THE DNA DAMAGE RESPONSE AND KRAS

COVID-19 AND CANCER

COVID-19 AND CANCER

ENDOCRINOLOGY

NEW APPROACHES TO TARGETING HORMONE DEPENDENT CANCERS

EXPERIMENTAL AND MOLECULAR THERAPEUTICS

ELUCIDATING DISEASE BIOLOGY AND DRUG RESISTANCE MECHANISMS

ANTIBODIES AND IMMUNE THERAPIES

BIOLOGICAL THERAPEUTIC AGENTS AND NOVEL DRUGS

CANCER DRUG RESISTANCE AND REVERSAL OF RESISTANCE

CELL DEATH AND OTHER PATHWAYS IN ONCOLOGY

NANOPARTICLES AND OTHER NOVEL STRATEGIES FOR DRUG DELIVERY

PROTEIN DEGRADERS AND PROTEASOME INHIBITORS

IMMUNOLOGY

IMMUNE CHECKPOINT AND IMMUNE MODULATORY THERAPY

ADOPTIVE CELL THERAPY 1

ADOPTIVE CELL THERAPY 2

IMMUNE CHECKPOINTS

IMMUNE RESPONSE TO THERAPIES 2 / IMMUNE MONITORING AND CLINICAL CORRELATES

MOLECULAR/CELLULAR BIOLOGY AND GENETICS

ONCOGENIC SIGNALING DRIVERS

TRANSCRIPTION FACTORS IN CANCER

CANCER GENOMICS 1

CANCER GENOMICS 2

CELL SIGNALING

CELLULAR STRESS RESPONSES 1: HYPOXIA AND OXIDATIVE STRESS

CELLULAR STRESS RESPONSES 2: SENESCENCE AND UNFOLDED PROTEIN RESPONSE

KINASES AND PHOSPHATASES

POPULATION SCIENCES

MOLECULAR GENOMICS OF CANCER RISK AND PROGNOSIS

DESCRIPTIVE EPIDEMIOLOGY COVERING CANCER INCIDENCE, MORTALITY, CLUSTERS, AND TRENDS

PREVENTION / EARLY DETECTION / INTERCEPTION

NEW TARGETS AND STRATEGIES FOR PREVENTION AND RISK REDUCTION

PRECLINICAL PREVENTION, EARLY DETECTION, AND INTERCEPTION 1

TUMOR BIOLOGY

LATE-BREAKING RESEARCH: TUMOR BIOLOGY 1

EX VIVO MODELS FOR UNDERSTANDING GENETIC VARIATION, METASTASIS, AND TREATMENT SENSITIVITY

FROM MECHANISMS TO NEW THERAPEUTIC TARGETS IN CHILDHOOD CANCER

3D MODELS AND MICROFLUIDICS

BIOLOGICAL MODIFICATION OF RADIORESPONSE

CARCINOGENESIS

DRUG TARGETING AND TREATMENT RESPONSE OF THE MICROENVIRONMENT

EVOLUTION, IMMUNE REGULATION, AND THE MICROENVIRONMENT OF METASTASIS

TISSUE RECOMBINANT, ORGAN SLICE, AND DECELLULARIZED MODELS

Friday, April 8, 2022

BIOINFORMATICS / CONVERGENCE SCIENCE / SYSTEMS BIOLOGY

Bioinformatics and Computational Biology

#5002

The prognostic relevance and underlying mechanisms of the novel oxygen sensor ADO in cancers.

Jie Huang,¹ Juan Shen,² Yao Qiu,³ Jiannan Qian,⁴ Jiawei Li,¹ Xueqin Chen,² Shenglin Ma¹. ¹Affiliated Hangzhou Cancer Hospital, Zhejiang University School of Medicine, Hangzhou, China; ²Affiliated Hangzhou First People's Hospital, Zhejiang University School of Medicine, Hangzhou, China; ³Nanjing Medical University Affiliated Hangzhou Hospital, Hangzhou, China; ⁴The Forth Clinical Medical College, Zhejiang Chinese Medical University, Hangzhou, China.

Purpose: Hypoxia is widely recognized as a determiner in cancer. Cysteamine dioxygenase (ADO) is demonstrated to be a novel oxygen sensor, globally expressed in animals and plants. However, the functions of ADO in human cancers remain unclear. This study was to identify the survival association of ADO in cancers and to excavate its underlying mechanisms.

Methods: Transcriptional expression profiles, clinical characteristics, single nucleotide variation (SNV) data, and 450K methylation data were downloaded from all cancer types in the cancer genome atlas (TCGA). Two independent liver cancer datasets GSE14520 and GSE36376 were downloaded for further validation.

Results: The expression of ADO gene was differentially expressed in various cancers. High expression of ADO indicated a worse OS and PFS in adrenocortical carcinoma (ACC), acute myeloid leukemia, liver hepatocellular carcinoma (LIHC), while predicted better OS and PFS in brain lower-grade glioma (LGG), and kidney renal clear cell carcinoma (KIRC). Across various cancers, ADO was positively correlated with tumor-associated pathways including G2/M_checkpoint, MYC_targets, and TGFB while negatively correlated with Genes_upregulated_by_reactive_oxygen_species (ROS) through ssGSEA analysis. The expression of ADO gene was highly correlated with hypoxic marker HIF1A, HIF2A, KDM5A, and KDM6A in almost all cancer types while poorly correlated with Hallmark_hypoxia geneset and Regulation_of_cellular response_to_hypoxia geneset, which indicated that ADO might work independently with HIF1A-mediated response. ADO was positively correlated with the methylation of most chromatin regulators in LGG and LIHC. ADO expression positively correlated with most immune checkpoint molecules in most cancers. Notably, a highly positive correlation between ADO and CD276, ENTPD1, or HMGB1 was observed in most cancer types. In liver cancer, ADO was positively correlated with the infiltration of B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages, myeloid dendritic cells. However, ADO expression was not strongly relevant to TMB in liver cancer. The increased expression, poor survival indication and immune checkpoint correlation of ADO in liver cancer were further validated by GSE14520 and GSE36376 datasets.

Conclusion: ADO was differentially expressed in various cancers and significantly associated with survival in some cancer types including liver cancer. The underlying mechanism might be associated with the G2/M checkpoint, MYC targets and TGFB pathways, and ADO-mediated HIF1A-independent hypoxic response. The immune cell infiltrates and immune checkpoints association hints that ADO might be a biomarker for immunotherapy in liver cancer.

#5003

Gene expression to copy number alterations analysis in colorectal cancers: Organ specificity of transcriptomic impact on copy number alterations.

Chinthalapally V. Rao,¹ Chao Xu,¹ Yuting Zhang,¹ Adam S. Asch,² Hiroshi Y. Yamada¹. ¹University of Oklahoma Health Sciences Center (OUHSC), Oklahoma City, OK; ²Stephenson Cancer Center, Oklahoma City, OK.

Background: Colorectal cancers (CRCs) is the third most commonly diagnosed cancer and the second most lethal cancer worldwide. CRCs carry high degrees of genomic instability (GI), which enables cancer evolution and makes prognosis poorer. Thus, GI is an exploitable target for therapy purposes. GI can be caused by gene mutation, alteration in gene expression, epigenetic modification, and other infectious or environmental factors. Several mutated genes involved in Chromosome instability and/or microsatellite instability have been identified. Yet it has not been technically feasible to modify functions of mutated genes (e.g., APC, TP53). From the drug development’ standpoint, inhibition of over-expressed target gene is preferred. However, specific genes involved in GI via expression alterations in CRCs have not been comprehensively identified. The gap in knowledge hinders development of CRC therapies targeting GI in CRC. In a previous study, we developed a data mining strategy (Gene Expression to Copy Number Alterations; GE-CNA) and comprehensively identified 1,578 genes that associate with CNA in lung adenocarcinoma, among which 39 were survival-critical (i.e., expression levels correlate with significant differences in patients’ survival).

Rationale: Causal genes for GI via expressions have not been comprehensively identified in CRCs. Thus, the mechanistic understanding of GI generation in CRC remains incomplete. Few attempts have been made to understand the organ-specificity of genomic instability in cancer.

Main Results: We applied the GE-CNA approach to 592 TCGA CRC datasets, and identified 513 genes whose expression levels associate with CNA. Among these, 27 were survival-critical. Comparison with previous results from lung adenocarcinoma indicated striking differences between lung adenocarcinoma and CRC: (a) overall CNA numbers are higher in lung adenocarcinoma than CRC in all stages, (b) 262 CRC-CNA facilitator genes did not show significant concentration in a specific pathway, and (c) 251 CRC-CNA suppressor genes are concentrated in the following pathways: Interferon Signaling, Antigen Presentation Pathway, Heme Biosynthesis II, Natural Killer Cell Signaling, Retinoic acid Mediated Apoptosis Signaling, JAK/Stat Signaling, Glucocorticoid Receptor Signaling, Heme Biosynthesis from Uroporphyrinogen-III I, and Glutathione Redox Reactions II.

Implications: The 27 survival-critical genomic instability genes are potential targets to suppress CRC genomic instability and therefore are targets for CRC drug development. Causal genes for genomic instability via expressions differ among organs. Hence, the targeting genomic instability and/or aneuploidy approach will need to be tailored for the specific target organ.

#5004

Integrative quality control of cancer somatic mutations with CaMutQC.

Xin Wang, Jian Ren, Qi Zhao. Sun Yat-sen University Cancer Center, Guangzhou, China.

The quality control of cancer somatic mutations is an essential step for eliminating false positive mutations from technical bias, and selecting key mutation candidates plays a crucial role in both downstream translation research and personal medical decisions. Existing tools, with complicated parameters and changeable filtering standards, may not be suitable specifically for cancer somatic mutations. In addition, previous genomic studies usually adopted different filtering criteria for processing raw sequencing datasets, which greatly restricts the efficiency of integrative exploration for cancer genomics.

In this study, we presented CaMutQC, a heuristic cancer somatic mutations (CAMs) quality control and filtration package for cancer genomic studies. CaMutQC provides two schemes, including filtration of false positive mutations generated by technical issues, and the screening of candidate cancer somatic mutations from single or multiple VEP-annotated files in VCF format. Specifically, for quality control, instead of directly discarding CAMs that failed to meet the thresholds of parameters collected from classic studies, CaMutQC labels them with customized tags to represent the filtering type, enabling portable subsequent analyses. And a vivid and well-structured filter report is generated after filtration or selection. In addition, we proved by applying an updated mutation QC strategy, which takes the union of CaMutQC-filtered cancer variants from multiple variant callers (eg. MuTect, MuSE and VarScan2) on published datasets that, CaMutQC can not only reduce the false positive CAMS, but also greatly rescue the false negative CAMS from each single caller or the overlap of multiple tools.

In summary, in this study, for the first time, we systematically implemented the parameters and criteria of CAMs quality control from published studies into the CaMutQC package, which we believe it would be a valuable tool for cancer genomic research. CaMutQC is implemented in R and is available at https://github.com/likelet/CaMutQC under the GPL-v3 license.

#5005

PSMB9/LMP2 as a controversial diagnostic biomarker for uterine leiomyosarcoma.

Raul Maia,¹ Georgia Kokaraki,² Jorge E. Souza,¹ Joseph W. Carlson,³ Tirzah B. Petta³. ¹University Federal of Rio Grande do Norte, Natal, Brazil; ²Karolinska Institutet, Stockholm, Sweden; ³University of Southern California, Los Angeles, CA.

Uterine leiomyosarcoma (uLMS) are rare and malignant tumors that arise in the cells of the myometrium. Diagnosis is based on histopathological features. Developing effective therapies against uLMS is a challenge due its resistance to conventional radiation and chemotherapy. In vivo and in vitro models for uLMS are urgently needed and it has been proposed that knockout mice for the gene PSMB9 (MIM:177045), also known by LMP2, could be a good model to study uLMS. Previous studies using mice homozygous deficiency for PSMB9 presented abnormalities in the biological functions of the immunoproteasome and spontaneously developed uLMS thus suggesting this gene as a tumor suppressor gene. Furthermore, expression of PSMB9 was suggested to be a specific biomarker for the diagnosis of uLMS. The goal of this study was to evaluate the role of PSMB9 in uLMS using in silico analysis of transcriptomics data from a patient cohort, combined with publicly available datasets. Molecular data were integrated and used for a meta-analysis of RNA-seq in order to find Differentially Expressed Genes (DEGs). A total of 68, 66 and 67 samples from normal myometrium (MM), uterine leiomyoma (LM) and uLMS, respectively, were analyzed. A quality control followed by adapter removal and alignment step was assessed using FastQC, Trim-galore and HISAT2 software respectively and counting reads was performed using htseq-count. Expression values were calculated as Transcripts Per Million (TPM) in order to remove libraries dependencies. Differential expression analysis was done using DESeq2 and edgeR and genes were classified as DEG when that had adjusted p-value < 0.01 and |log2FC| ≥ 2 in both packages. Gene Set Enrichment Analysis was performed using fgsea package with 100,000 permutations and MSigDB collections (KEGG and Reactome) as input parameters. Our results showed an over-expression of PMSB9 in uLMS samples and immune-related pathways were significantly enriched. Moreover, the overall survival analysis for the outcome of uLMS patients based on PSMB9 expression in TCGA showed no significant changes, further questioning the importance of PSMB9 in uLMS malignancy. Finally, based on phylogenetic analysis we showed that PSMB9 conservation between Mus musculus and Homo sapiens is not closely related. This divergence suggests that the biological role of PSMB9 may differ between M. musculus and H. sapiens. Human uLMS exhibits an over-expression in PSMB9 and these findings suggest PSMB9-deficient mice are not an appropriate diagnostic-biomarker for uLMS.

#5006

Computational analysis of hypoxia-inducible factor 1-alpha and UBS109 in hepatocellular carcinoma.

Neha Bahadur Merchant,¹ Santosh Kumar Behera,² Afroz Alam,¹ Ganji Purnachandra Nagaraju,³ Riyaz Basha⁴. ¹Banasthali University, Vanasthali, India; ²National Institute of Pharmaceutical Education and Research (NIPER), Ahmedabad, India; ³Winship Cancer Institute/Emory University, Atlanta, GA; ⁴UNT Health Science Center, Fort Worth, TX.

Hypoxia-inducible factor 1-alpha (HIF-1α) is a heterodimeric transcription factor, which is elevated during hypoxic conditions in several malignancies. HIF-1α is known to stimulate hypoxia-responsive genes that are linked with malignant properties viz.: progression, metastasis, and resistance to therapy. The objective of the present study is to target HIF-1α by potential phytodrug, curcumin and its monocarbonyl analog UBS109 in hepatocellular carcinoma (HCC). Curcumin has been known to exhibit therapeutic properties for the management of HCC due to its powerful anti-inflammatory and antioxidant properties as well as its ability to regulate multiple signaling pathways. UBS109 possesses comparable antitumor and anti-inflammatory properties as curcumin without any of its limitations. To analyze the binding capacity and ligand efficacy of HIF-1α against curcumin and UBS109, computational approaches were applied including molecular docking and dynamics simulation that can provide detailed understanding of its implication in HCC treatment and a deeper analysis of the protein-drug interaction.

#5007

Discovery of stromal targets in pancreatic cancer.

Shengjun Li, Jeong-Bin Park, Hyung-Jun Im. Seoul National University, Seoul, Republic of Korea.

Background: Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies with limited treatment options. Development of stroma targeting therapeutics for PDAC is gaining attention because PDAC has a unique tumor microenvironment with dense tumor stroma and the tumor stroma is a critical mediator of PDAC progression. Herein, we discovered pancreatic cancer stroma-specific cell surface markers by integrative analysis of single cell RNA sequencing and cell surface marker database.

Methods: Single cell RNA sequencing data of PDAC tissue (n = 3, GSE154778) and normal pancreas tissue (n = 3, GSE131886) was obtained from Gene Expression Omnibus database. Using Seurat package, the data from each sample integrated and the cells were clustered. We selected stroma cell cluster for differential analysis to obtain the genes enriched in PDAC stroma. Then, the intersection of the PDAC stroma specific genes with the gene set of SurfaceomeDB was obtained, which allowed us to obtain the stroma cell surface markers that are highly expressed in PDAC. The marker genes were validated in 1) 36 pairs of PDAC and normal tissue samples from a RNA sequencing data (GSE15471), and 2) 143 PDAC samples in The Cancer Genome Atlas (TCGA), 328 normal pancreatic tissues in Genotype-Tissue Expression (GTEx) database (https://www.gtexportal.org).

Results: The cells from PDAC and normal pancreas were clustered into ductal/tumor, stromal and immune cell clusters according to the expression of cell-type specific marker genes. We identified variably expressed genes between PDAC and normal cells within the stromal cell cluster. Among the PDAC stroma-specific genes, we could obtain 360 cell surface markers using SurfaceomeDB. The top ten PDAC stroma-specific cell surface marker genes were MXRA8, ANTXR1, THY1, SCD1, LY6E, BST2, GPNMB, CDH11, GJB2, and CD55. To validate the results, we compared the expression levels of the top 10 genes between pancreatic cancer tissue and adjacent normal pancreatic tissue. We found that the expression levels of the genes were significantly higher in the cancer tissue compared to the normal pancreatic tissue. We also confirmed that the expression levels of 10 genes were significantly higher in all stages of PDAC compared to normal tissues in a combined RNA sequencing dataset from healthy human pancreatic tissue from GTEx and pancreatic cancer patients from TCGA.

Conclusion: We identified pancreatic cancer stroma-specific cell surface markers using single-cell RNA sequencing data and SurfaceomeDB. The markers can be utilized in development of various pancreatic cancer stroma targeted therapeutics including antibodies, antibody-drug conjugates, peptides, cell therapeutic and nanoparticles.

#5008

The potential role of hsa-miR-3157-3p in prostate cancer pathogenesis.

Akhil Wali,¹ Hari Koul². ¹Morgan State University, Baltimore, MD; ²Louisiana State University, New Orleans, LA.

Prostate cancer remains a major area of concern as it is the second leading cause of cancer mortalities amongst American men. Our study aimed to find the role of differentially expressed microRNAs that play a role in prostate cancer etiopathology. With the use of the National Center for Biotechnology Information (NCBI), we narrowed down a desired dataset (GSE138740). While analyzing the dataset, we found the expression pattern of hsa-miR-3157-3p (adjusted p-value 0.04797) to be significantly upregulated in low grade tumor samples (Gleason Score 6 and n=88) compared to high grade tumor samples (Gleason Score 8,9 and n=13). Furthermore, the role of hsa-miR-3157-3p in prostate cancer remains a novel finding. We have identified target genes such as E2F3, UBE2M, ABCF1, PKN3 and VANGL1 using the RNA Central website portal. Further studies are underway to elucidate the expression pattern of this miRNA and it’s target genes that would correlate with the disease progression observed among tumor samples in The Cancer Genome Atlas (TCGA) database. The Ingenuity Pathway Analysis (IPA) software generated report will be presented at the meeting to describe gene networks and pathways implicated in Prostate Cancer. Overall, these studies are designed to further validate whether this miRNA and its target genes can serve as potential biomarkers to predict the onset and progression from low to high grade prostate cancer.

#5009

DKK1 is a potential target for enhancing the efficiency of radioimmunotherapy in head and neck squamous carcinoma.

Xinyu YE,¹ JIAN Zhang,¹ Yi Lu,¹ Yingxuan Gong,¹ Xinyu Jiang,¹ Yueming Li,¹ Yinghao Wang,¹ Xiuzhi Li,¹ Jingwen Liu,² Rencui Quan². ¹SUSTECH, Shenzhen, China; ²Shenzhen People’s Hospital, Shenzhen, China.

Immune checkpoint inhibitors (ICIs) have been approved to treat various types of tumors, including head and neck squamous carcinoma (HNSC). However, most HNSC patients do not respond to ICIs. Radioimmunotherapy has been proposed to enhance the immune response rate in HNSC. Herein, we aimed to explore the effect of DKK1 on radiotherapy and immunotherapy in HNSC. We collected the gene expression profile and clinical information of HNSC patients from TCGA, GEO, and ICGC. The immune cell infiltration, level of related pathways and biological processes, signature scores of gene sets of interest were assessed using CIBERSORT, GSEA or GSVA respectively. One of the genes, Dickkopf-1 (DKK1), a secreted protein which plays important roles in the Wnt signaling pathways, was identified because of its significantly higher expression in HNSC. Then, we confirmed the DKK1 expression at protein level by immunohistochemistry of HNSC tissue microarray. CCK8, apoptosis detection and survival curves were performed to investigate the response of DKK1-knockdown HSC-3 cell line to radiation. We found the mRNA and protein expression level of DKK1 was significantly higher in HNSC tissue. The DKK1high group had a shorter survival time in both disease-free survival and overall survival than the DKK1low group. The DKK1high group showed a more immunosuppressive microenvironment with lower infiltration of T cells and higher infiltration of MDSCs. Furthermore, the DKK1high group was enriched in radioresistance related to wound healing, angiogenesis, and autophagy. Patients in the DKK1high group who received radiotherapy had much poorer disease-free survival and overall survival than those in the DKK1low group. HSC-3 cell line was more sensitive to radiation after knockdown of DKK1 by siRNA. Finally, we established a risk score system based on the 9 genes from 41 DKK1-associated immunomodulators and created a nomogram model to predict the prognosis of the patients. Our data show that DKK1 can affect both radiotherapy and immunotherapy in HNSC, suggesting that DKK1 can be a potential target for radioimmunology in HNSC.

#5010

Comprehensive characterization of 11 prognostic alternative splicing events in ovarian cancer interacted with the immune microenvironment.

Guixi Zheng. Qilu Hospital of Shandong University, Jinan, China.

Alternative splicing (AS) events play a crucial role in the tumorigenesis and progression of cancer. Transcriptome data and Percent Spliced In (PSI) values of ovarian cancer patients were downloaded from TCGA database and TCGA SpliceSeq. Totally we identified 1,472 AS events that were associated with survival of ovarian serous cystadenocarcinoma (OV) and exon skipping (ES) was the most important type. Univariate and multivariate Cox regression analysis were performed to identify survival-associated AS events and developed the prognostic model based on 11-AS events. The immune cells and different response to cytotoxic T lymphocyte associated antigen 4 (CTLA-4) and programmed cell death protein 1 (PD-1) blockers in low-risk and high-risk group of OV patients were analyzed. Ten kinds of immune cells were found up-regulated in low-risk group. Activated B cell, natural killer T cell, natural killer cell and regulatory T cell were associated with survival of OV. The patients in low-risk group had good response to CTLA-4 and PD-1 blockers treatment. Moreover, a regulatory network was established according to the correlation between AS events and splicing factors (SFs). The present study provided valuable insights into the underlying mechanisms of OV. AS events that were correlated with the immune system might be potential therapeutic targets.

#5011

Meta-analysis of human cancer single cell RNAseq datasets using the fully integrated IMMUcan database.

Jordi Camps,¹ Floriane Noel,² Robin Liechti,³ Lou Goetz,⁴ Caroline Hoffmann,⁵ Lucile Massenet-Regad,² Elise Amblard,² Melissa Saichi,² Mahmoud M. Ibrahim,¹ Jack Pollard,⁶ Jasna Medvedovic,² Helge G. Roider,¹ Vassili Soumelis². ¹Bayer AG, Berlin, Germany; ²INSERM U976, Paris, France; ³SIB Swiss Institute of Bioinformatics, Lausanne, Switzerland; ⁴Vital-IT group, SIB Swiss Institute of Bioinformatics, Lausanne, Swaziland; ⁵Institut Curie, Paris, France; ⁶Sanofi, Cambridge, MA.

The development of single cell RNA-sequencing (scRNAseq) technologies has greatly contributed to deciphering the tumor microenvironment (TME) landscape, and a wealth of biological data is now publicly accessible. This represents a very valuable resource to researchers in the field, offering a reference for comparison of novel results, as well as opportunities for original meta-analysis studies. However, the massive amount of biological information renders its exploitation difficult in the absence of a well-structured and annotated resource. Marked heterogeneity and variability between studies in terms of cancer type, clinical context, technological platform, data quality, number and type of cells, create additional bottlenecks. We have developed a fully integrated scRNAseq database exclusively dedicated to human cancer. It gathers 134 datasets on 51 different cancer types, annotated in 50 fields containing precise clinical, technological and biological information. We developed an original data processing pipeline organized in 4 steps: first, data collection; second, data processing, which includes quality control, sample integration, cell clustering; third, cell ontology tree of the TME, built and used to annotate the clusters in a supervised and manual manner; and fourth, interface to analyze TME in a cancer type-specific or global manner. This integrated, accessible and user-friendly resource should be of great value to the biomedical community. It represents an unprecedented level of detailed annotation, offering vast possibilities for downstream exploitation of human cancer scRNAseq data for discovery and validation studies. The database is freely accessible at: https://immucanscdb.vital-it.ch.

#5012

Deep Gaussian process with uncertainty estimation Improves microsatellite instability prediction based on whole slide image: A retrospective multicenter and multiethnic cohort study.

Sunho Park,¹ Hongming Xu,² Sung Hak Lee,³ Jeonghyun Kang,⁴ Tae Hyun Hwang¹. ¹Mayo clinic, Jacksonville, FL; ²Dalian University of Technology, Dalian, Liaoning, China; ³The Catholic University of Korea, Seoul, Republic of Korea; ⁴Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea.

Background: Microsatellite Instability (MSI) is a prognostic and predictive biomarker which can guide treatments including immunotherapy for patients with gastrointestinal cancer. Universal MSI testing would benefit to patients to receive a better therapy, but many patients remain untested. To deliver broadly accessible MSI testing, deep learning models have demonstrated their feasibility for MSI prediction using H&E-stained whole-slide images (WSIs). However, to be deployed into clinical routine care, prediction models should be validated in datasets from multicenter and multiethnic groups. In addition, these models should provide uncertainty of prediction to help clinicians to make informed decisions.

Method: We develop a MSI prediction model using WSIs based on Deep Gaussian process (DGP), a Bayesian model being able to model the uncertainty in prediction. We implement a DGP model in the transfer learning, where a WPI is decomposed into multiple non-overlap image patches which are converted to feature vectors by a pretrained convolutional neural network (CNN). Then, the DGP model makes a prediction for an image by averaging score function values at all the feature vectors in the image in weakly supervised learning (WSL). To test our method, we collected H&E stained colorectal/stomach cancer slides (n=1,619) from multi-institutions and multiethnic groups, including TCGA Colorectal (CRC; n=351) and Stomach Adenocarcinoma (STAD; n=174) and six datasets from tertiary hospitals in Korea (i.e., Yonsei-1 (n=174), Yonsei-1-remade (n=146), Yonsei-2 (n=95), St. Mary-1 (n=48), St. Mary-2 (n=50), and Yonsei-CLASSIC trial (n=581)). We also compare our method to multiple deep learning methods, including Densenet, Resnet, Shufflenet, Googlenet, Squeezenet and two WSL models, MIL (Campanell et al., 2019) and CLAM (Lu et al., 2021).

Result: In each cancer type, we train each model using TCGA dataset as discovery cohort and tested on the rest of datasets as validation cohort. We use the area under of roc curve (AUC) for the evaluation metric. For the CRC cohorts, our DGP method achieves overall best performance with 0.81 (Yonsei-1), 0.82 (Yonsei-1-remade), 0.89 (Yonsei-2), 0.85 (St. Mary-1) 0.75 (St. Mary-2) AUCs compared to the state-of-the art methods. For the STAD cohorts, our method also achieves overall best performance with 0.74 AUC (Yonsei-classic). In addition, incorporating the uncertainty in prediction measured by our method improves the model’s performance to predict MSI. Specifically, removing cases with prediction results with high uncertainty, which could lead false positive prediction, significantly improves MSI prediction performance. Taken together, these results demonstrate the robustness and generalisability of the DGP model for MSI prediction across multicenter and multiethnic datasets.

#5013

Integrated analysis of intragastric microbiome and human gene expression uncovers genetic, microbial, and immunological associations in gastritis and gastric cancer.

Chan Hyuk Park,¹ Changjin Hong,² A-reum Lee,¹ Jaeyun Sung,³ Tae Hyun Hwang². ¹Hanyang University College of Medicine, Seoul, Republic of Korea; ²Mayo Clinic, Jacksonville, FL; ³Mayo Clinic, Seoul, MN.

Background: Gastric carcinogenesis has been found to be associated with intragastric microbiome, gastric mucosal gene expression, and immune cells comprising the tumor microenvironment separately in independent studies. However, to date, there has been no multi-omic investigations analyzing these biological traits simultaneously. Thus, there is a paucity in our understanding of the intricate relationship between intragastric microbes and the host in gastric disease. To address this gap, we aimed to comprehensively investigate the microbiome, host gene expression, and immune cells in the stomach across a range of different states along the gastric carcinogenesis pathway.

Methods: Gastric biopsy samples were obtained from healthy individuals or patients with gastritis or gastric cancer (biopsy cohort, n=30). Surgical samples (of cancer and adjacent severe gastritis tissues) were obtained from gastric cancer patients (surgery cohort, n=40). Using 16S rRNA gene sequencing, RNA-seq, and cell-type enrichment analysis of each participant’s biospecimens, we identified associations between the microbiome and host transcriptome or inferred immune cell types.

Results: Microbiome composition and gene expression patterns showed significant differences between disease states (healthy, gastritis, or cancer). Expression of ‘Cell cycle’ pathway genes was enriched in gastric cancer in both biopsy and surgery cohorts. In the biopsy cohort, Helicobacteraceae, which was abundant in gastritis, was highly correlated with FAM3D, which was previously found to be involved in gastrointestinal inflammation. In the surgery cohort, Lachnospiraceae, which was abundant in gastric cancer, was highly correlated with UBD, which is known to regulate mitosis and reduce cell cycle time. Cell-type enrichment analysis revealed lower B cell infiltration in gastric cancer than in gastritis (both cohorts). Tissue-infiltrating B cells were associated with Helicobacteraceae.

Conclusions: Associations between the intragastric microbiome, human gene expression, and tissue-infiltrating immune cells are unique to gastric disease state. The findings herein may aid the discovery of novel biomarkers or therapeutic targets of gastric cancer.

#5014

The landscape of ERBB2 mutations in Chinese patients with solid tumors.

Xiangyu Su,¹ Hongmei Zhao². ¹Cuhk Hospital affiliated to Southeast University, Nanjing, China; ²Chosen Med Technology (Beijing) Co., Ltd., Beijing, China.

Background: Receptor tyrosine kinase (ERBB2) is altered by mutations and amplifications in various cancer types. Meanwhile, FDA approved some drugs to use for variants in ERBB2. However, the landscape of ERBB2 mutations and copy number variants (CNV) in pan-cancer is unknown in Chinese.

Methods: 599 gene-panel by next-generation sequencing (NGS) was performed on 2,570 patients with solid tumors including non-small cell lung cancer (NSCLC) (n = 660), esophagogastric (n = 89), colorectal (n = 966), breast (n = 156) and gastric cancer (n = 699). The mutations and CNV in ERBB2 were detected for characterize the landscape in Chinese patients with solid tumors.

Results: The frequency of single nucleotide variants (SNV) and CNV in ERBB2 was 9% and 4.3% in pan-cancer and the rate of variants was 13.3%. The most frequency of CNV was occurred in breast cancer and the last was in non-small cell lung cancer (13.5% vs 1.1%). Nevertheless, the rate of SNV in different cancer species was almost the same as in pan-cancer (mean: 8.8%). The higher rate of variants in ERBB2 was breast cancer and the lower was NSCLC. The cancer types with the highest incidence of ERBB2 variants were the following: breast (22.4%), gastric adenocarcinoma (18.3%), esophagogastric cancer (11.2%), colorectal cancer (10.8%) and NSCLC (9.7%).

Conclusion: Based on the results, the profile of variants in ERBB2 revealed more patients may profit from FDA-approval drugs in China. The patients with somatic mutations in ERBB2 are treated with trastuzumab deruxtecan and adotrastuzumab emtansine in NSCLC. The patients with amplification are treated with trastuzumab deruxtecan in most tumors, but margetuximab and chemotherapy can cure breast cancer, according to FDA-approval.

Keywords: ERBB2 mutations, CNV, SNV, solid tumors

#5015

Precision profile simulation study for a next generation sequencing bTMB assay.

Kevin Doubleday, Daniel Gaile, Ravi Vijaya-Satya, Xianxian Liu, Kevin D'Auria, Soni Shukla, Han-Yu Chuang, Katie Quinn, Darya Chudova. Guardant Health, Inc, Redwood City, CA.

Background: Precision profile simulations (PPS) can be used to assess variability of biomarker profiles and provide valuable insight into assay performance, especially when reliable precision estimates can not be obtained empirically due to scarcity of representative samples or insufficient materials per sample. A PPS was conducted for the GuardantOMNI assay to characterize the expected variability in blood tumor mutational burden (bTMB) score across a representative range of expected bTMB scores in clinical samples. The simulations were aligned to, but not completely prescribed by, the PPS guidance provided in Guidance for Industry and and Food and Drug Administration Staff Class II Special Controls Guidance Document: Ovarian Adnexal Mass Assessment Score Test System. A sample’s bTMB score is a real valued quantity (e.g., bTMB = 21.04 mut/Mb) that is derived by multiplying the number of qualified mutations observed within a targeted panel by a scaling factor. Variability in observed bTMB scores for a given blood sample is governed primarily by sample coverage, tumor shedding level, and the assay somatic variant detection probabilities (a function of underlying variant allele frequencies, VAFs).

Methods: The relationship between site-specific total molecule counts and coverage was modeled utilizing a composite dataset consisting of both clinical and contrived samples. Sample coverage was modeled using variance component estimates from Precision Study data (18 cancer samples each with 6 to 18 replicates).

The reference, single-strand mutant, and double-strand mutant molecule counts for a somatic variant site detected in at least one sample replicate were modeled utilizing a bias corrected Dirichlet Multinomial model.

The variants with the simulated VAF and coverage levels were processed with the GuardantOMNI germline/somatic classifier to account for the uncertainty in germline/somatic classification at lower coverage values.

Results: Precision profiles consisting of simulation derived %CV estimates for 18 clinical samples with a representative set of mean bTMB scores were generated. The PPS bTMB score distributions were consistent with the bTMB scores observed in the Precision Study, supported by visualization and confidence intervals at level 0.05 margins of equivalence for the empirical mean bTMB and standard deviation estimates.

The sample specific %CV estimates were observed, in most instances, to decrease with increasing input levels for matched targeted LoD (Limit of Detection) simulation results.

Precision profile %CV estimates were observed to be inversely related to mean bTMB scores.

Conclusions: The results provide proof of principle that estimation of GuardantOMNI bTMB score precision via an intuitive and interpretable simulation model is viable. The simulation results were consistent with empirical data and general expectations regarding the precision of the bTMB scores.

#5017

A single-cell atlas of tumor microenvironment defines the continuum of gastric adenocarcinoma tumorigenesis and progression.

Ruiping Wang,¹ Shumei Song,¹ Jiangjiang Qin,² Yuan Li,³ Yibo Fan,¹ Deyali Chatterjee,¹ Ghia Tatlonghari,¹ Zhiyuan Xu,² Can Hu,² Shaowei Mo,² Matheus D. Sewastjanow,¹ Ahmed Adel Fouad Abdelhakeem,¹ Zhenning Wang,³ Xiangdong Cheng,² Jaffer A. Ajani,¹ Linghua Wang¹. ¹The University of Texas MD Anderson Cancer Center, HOUSTON, TX; ²Cancer Hospital of The University of Chinese Academy of Sciences, Hangzhou, China; ³First Hospital of China Medical University, Shenyang, China.

Gastric adenocarcinoma (GAC), a global health burden, lacks detail understanding of the evolution-driven cellular/molecular programs that lead to GAC tumorigenesis followed by progression/metastases. The definition of the interplay(s) between immune/stromal cells and premalignant/malignant GAC cells could propel us into a new dimension of understanding and therapeutics. Here, we performed a comprehensive single-cell profiling of 68 specimens collected from 43 subjects including non-neoplastic tumor adjacent tissue, precancerous lesions, localized, and metastatic GACs. We profiled a total of 77,392 high-quality cells which revealed 62 unique cell states uncovering varying profiles. We define alterations in TMEs that underscore initiation of tumorigenesis to eventual progression. For instance, we found a striking preponderance of B lineage cells, primarily the IgA+ plasma cells, in TMEs of the precancerous lesions, whereas 3 immunosuppressive myeloid subsets with high expression of genes signature including SPP1, LAIR1, SIRPA, TIM-3, TGFB1, and MARCO dominated in advanced GACs. We observed that fractions of GZMK+ effector CD8 T cells and progenitor exhausted CD8 T cells gradually increased as GACs progressed to advanced stages with highest abundant in metastatic GACs. In addition, our analysis revealed extensive stromal remodeling along the GAC continuum, which may have contributed to enhanced angiogenesis and progressive immune suppressive signaling. Notably, we uncovered 3 unique TME interactomes that are defined by 6 cellular environtypes that provide context-dependent definition and granularity to GAC networks inhabited by 62 TME cell subsets giving GAC to a novel landscape not yet defined. The two distinct environtypes in GAC primaries are validated in three independent large-scale GAC cohorts, giving credence and definition to previously established histopathological variables, genomic/molecular subtypes and clinical outcomes. The analysis of tumor associated stromal cells discovered SDC2 as an exploitable target to pursue. SDC2 was not only abundant in the stroma, but the abundance is validated in 3 independent single-cell GAC cohorts as well as at the protein level using independent approaches. Overexpression of SDC2 formed a gradient from early GAC to metastatic and is prognostic in large-scale GAC cohorts examined. This study provides an atlas of GAC TMEs from tumorigenesis to advanced GAC that could be further developed for novel therapeutics but also serves as a community resource.

#5018

A comprehensive analysis of genomic determinants of response to immune checkpoint inhibitor-based immunotherapy.

Jing Yang, Qi Liu, Yu Shyr. Vanderbilt University Medical Center, Nashville, TN.

A low response rate is the major challenge for existing cancer immunotherapies. A number of biomarkers have been reported to be associated with the likelihood of patient responses to immune checkpoint inhibitors (ICI), including the expression of PD-L1, tumor mutational burden (TMB), tumor infiltration lymphocytes, and T cell repertoire. However, these biomarkers showed variable performance even conflicting conclusions in different cohorts. And a large proportion of patients exhibited an immune-exclusion or immune-desert phenotype that cannot be explained by current biomarkers. In addition, establishing a cohort with adequate sample size requires multiple years of multi-center efforts. Therefore, it is still a challenge and a lack for the evaluation of efficacy of known biomarkers and the discovery of new signature(s) in a large-scale study. We hypothesized that a data-driven meta-analysis approach integrating multi-cohorts with multi-omics data will help to characterize the predictive genomic features and discover the mechanisms of antitumor immune responses.

In this work, we collected 3,037 ICI therapy samples with genetic and transcriptomics data for 14 cancer types. Then a data-driven strategy was designed to explore each type of genomic feature which included gene mutations, pathway mutations, TMB, mutational signatures, gene expression, pathway expression, interaction among checkpoint blockade, and immune cell components. Then a comprehensive analysis, which contained a meta-analysis for the identification of consistent biomarkers across multiple cohorts and a data aggregation-analysis for the identification of rare biomarkers on aggregated data, was performed to explore the associations between ICI response and omics features. Furthermore, we validated these signatures as biomarkers in three independent cohorts. Finally, we identified five types of biomarkers, including 1) two biomarkers regarding source of antigen (TMB and mutations of MYO7B), 2) one feature of responding to antigen (cytotoxic T lymphocytes pathway), 3) three biomarkers indicating macrophages M1 infiltrations (macrophages M1 component, and two marker genes of macrophages M1, CXCL11 and CCL19), 4) one biomarker regarding immune interactions (expression relations among immune checkpoint genes), and 5) two biomarkers about T cell-inflamed tumor microenvironment (IFNG and ligands genes of chemokine receptor CXCR3). In addition, a predictive module was built based on these signatures and outperformed than TMB or PD-L1 expression which were approved molecular biomarkers by FDA. We found high TMB patients had more likelihood to active the cytotoxic T lymphocytes which secret IFNG to kill tumor cell and promote the recruitment of macrophage M1. And the T cells were stimulated by macrophage M1 via the binding of CXCR3 of T cell surface and receptor of CXCR3 of macrophage M1 surface.

#5019

Eliminating copy number alteration effects in the gene essentiality data from the Cancer Dependency Map project.

Arkajyoti Bhattacharya, Carlos G. Urzúa-Traslaviña, Marcel A. T. M. van Vugt, Rudolf S. N. Fehrmann. University Medical Center Groningen, Groningen, Netherlands.

The Cancer Dependency Map project (DepMap) aims to identify cancer vulnerabilities through the in vitro study of genetic dependencies in cancer cell lines using CRISPR/Cas9 loss-of-function screens. Cas9-mediated DNA break formation induces toxic effects that are proportional with the number of cuts. Consequently, the DepMap data displays a ‘copy number alteration effect’ (CNA effect), in which sgRNAs targeting highly amplified regions of the genome produce depletion effects. Indeed, when sgRNAs target intergenic regions of amplified regions, the observed effects are comparable to the targeting of essential genes. Hence, correcting the CNA effect is crucial to reduce false positive (and increase true negative) co-dependencies between genes. Currently, the CERES methodology is used to correct for CNA effects. By applying a consensus-independent component analysis-based (consensus ICA) algorithm on gene essentiality data after CERES correction, we observed that many CNA effects are still present in the CERES-corrected data. Subsequently, we developed a methodology that outperforms CERES in removing the CNA effect. For example, we identified a set of genes mapping to chromosome 8q of sample ACH-000542 (ovarian adenocarcinoma cell line HEYA8 having amplification at chromosome 8q) having significantly lower essentiality levels compared to the rest of the genome. Whereas the CERES algorithm did not correct this effect of the chromosome 8q amplification, it was captured by consensus ICA in a consensus estimated source and was corrected by our novel methodology. After removing all CNA effects from gene essentiality data, we observed that 12.07% of co-dependency scores reported by CERES-corrected data decreased by 0.2 (on a range of -1 to 1), and 11.34% co-dependency scores increased by 0.2. In conclusion, consensus ICA-based methodology improves the correction of the CNA effect on gene essentiality data. Application of our improved method of CNA effect correction could reduce the number of false positive targets in validating biological hypotheses leading to novel therapeutic strategies in cancer.

#5020

Identification of transcriptional regulatory networks associated with papillary thyroid carcinoma.

Niradiz Reyes,¹ Stephanie Figueroa,² Christian Figueroa,³ Jan Geliebter⁴. ¹University of Cartagena, Cartagena, Colombia; ²Northeastern University, Boston, MA; ³Ossining High School, Ossining, NY; ⁴New York Medical College, Valhalla, NY.

Background: Identification of critical transcription factors (TFs) required by cancer cells to sustain biological processes that support their growth and survival is urgently needed to develop strategies targeting them. Papillary thyroid carcinoma (PTC) is the commonest thyroid malignancy making about 80% of all thyroid cancers cases. Gene expression of a number of thyroid-specific TFs has been found deregulated in thyroid carcinomas. Therefore, molecular approaches targeting overexpressed oncogenes are desirable therapeutic methods to improve the treatment of cancer patients. In this study, we used Transcription Factor Enrichment Analysis (TFEA) to detect positional motif enrichment associated with changes in transcription observed in this cancer type.

Methods: We previously identified differentially expressed genes between PTC and normal thyroid tissue using DNA microarrays. In the current study, the former overexpressed genes were subjected to TFEA using the web-based tool ChEA3. For comparison purposes, the Gene Expression Profiling Interactive Analysis (GEPIA2) tool was used to identify the most differentially overexpressed genes in thyroid carcinoma versus paired normal samples from patients in the TCGA database, and the differentially overexpressed gene set was also subjected to TFEA.

Results: TFEA using the ChEA3 web-server identified several TFs associated with overexpressed genes observed in PTC. The top-ranked TFs found in both the gene set identified in PTC patient samples using microarrays and the gene set identified in the TCGA database using GEPIA2 were: EHF, POU2F3, KLF5, ELF3, HES2, GRHL3, and HNF1B. Several of these TFs have previously been identified in thyroid carcinomas, while others are newly identified. EHF and ELF3 are known to be overexpressed in thyroid carcinogenesis; KLF5 is a zinc-finger transcriptional factor recently found to be highly expressed in a subset of PTC patients with aggressive behavior. Bioinformatics analysis has previously shown that HNF1B is up-regulated in several cancers including thyroid carcinomas. The role of the remaining identified TFs in thyroid carcinogenesis has not been described. POU2F3 is a member of the POU domain family of transcription factors that bind to a specific octamer DNA motif and regulate cell type-specific differentiation pathways. Hes2 encodes a mammalian basic helix-loop-helix transcriptional repressor. GRHL3 (Grainyhead-like 3) is a transcription factor involved in epithelial morphogenesis.

Conclusions: Bioinformatics analysis of differentially expressed genes in tumor versus normal paired thyroid tissue samples from PTC patients and in thyroid carcinoma versus paired normal samples from patients in the TCGA database allowed the identification of TFs that may play a role in PTC. The group of TFs identified in this study may represent potential therapeutic targets for this cancer type.

#5021

Accurate quantification of infiltrating B cell composition and clone diversity in tumor samples.

Fabio Navarro, Eric Levy, Pamela Milani, Qiang Li, Shruti Bhide, Upasana Dutta, Charles W. Abbott, Jose Jacob, Rena McClory, John West, John Lyle, Sean Boyle, Richard O. Chen. Personalis, Inc., Menlo Park, CA.

Tumors harbor a complex ecosystem of malignant, immune, and stromal cells. While malignant cells dictate much of the tumor biology, there is evidence that the tumor microenvironment (TME) also plays a major role in disease etiology. Given the complexity and abundance of the TME cellular composition, investigating the role of immune cell types will yield novel biomarkers for tumor progression and response to therapies.

The role of B cells as a prognostic biomarker remains elusive. For instance, infiltrating B cells in CRC have both positive and negative prognostic value. Thus, a scalable approach to quantify B cells and the B-cell receptor (BCR) repertoire could yield novel insights into the role of B cells in tumor biology. To address this, we have developed immune cell quantification (InfiltrateID™) and immune receptor repertoire profiling (RepertoireID™) methods as part of the ImmunoID NeXT Platform®, an augmented, immuno-oncology-optimized exome/transcriptome platform.

We estimate B cell abundance and BCR repertoire by profiling FFPE and PBMC samples using ImmunoID NeXT™. In expanding upon InfiltrateID to further estimate B cell abundance, here we regress the bulk RNA-seq readout from a reference signature from purified immune cell types. We also generate orthogonal quantifications of B cell abundance by profiling samples with cytometry by time of flight, single-cell RNA-seq, flow cytometry, and immunohistochemistry (IHC). We compare BCR results from ImmunoID NeXT to a standalone sequencing approach to evaluate the concordance of top clones. We then utilize BCR profiling from ImmunoID NeXT to analyze clonality and isotype composition in tumor samples.

We first use InfiltrateID to estimate absolute B cell fractions in over 50 samples. Overall, we observe a high correlation between InfiltrateID results and orthogonal data sets in both PBMC and tumor FFPE samples (R²=0.90). When comparing BCR results from RepertoireID to a standalone BCR sequencing method that profiles IgM and IgG, we identify 475 and 387 of the top 500 clones in IgG and IgM, respectively, with highly concordant abundances across all clones (R²>0.72 and R²>0.82 in IgM and IgG, respectively). Next, we use InfiltrateID to estimate absolute B cell fractions in over 650 samples from 14 tumor types. On average, samples display B cell fractions in agreement with the literature and IHC quantifications, with higher B cell fractions in lung, breast, and cervical tumors. We also observe a range of BCR clonality values across tumor types. Finally, we observe differences in B cell composition and repertoire diversity in tumor samples from patients who underwent checkpoint blockade therapy.

We show that InfiltrateID and RepertoireID accurately capture the composition and clone diversity of infiltrating B cells in tumor samples.

#5024

Aging patterns in five disparate cancer entities indicate a novel senescent cell population with therapeutic targetability.

Dominik Saul, Robyn Laura Kosinsky. Mayo Clinic, Rochester, MN.

Human aging is associated with molecular changes and cellular degeneration, resulting in a significant increase in cancer incidence with age. Whether aging as its own entity can be initiator of putative cancerous developments, the transcriptional changes underlying the cancerous development itself are largely unknown. Here, we analyzed publicly available bulk RNA sequencing data of 328 control vs. 453 cancer samples in combination with single cell sequencing datasets from 49 control vs. 105 cancer patients. Using widely accepted aging-associated gene sets, we assembled key features of aging and their reflection in cancer development in five distinct cancer entities with an age-dependent increase of incidence, i.e. chronic myeloid leukemia (CML), colorectal cancer (CRC), hepatocellular carcinoma (HCC), lung cancer (LC) and pancreatic ductal adenocarcinoma (PDAC). After confirmation of the aging/senescence-induced gene (ASIG) expression to be upregulated in malignant diseases compared to healthy controls, we elucidated the importance of ASIGs during single cell development, making use of pseudotime analyses. In each entity, distinct temporally late enrichment states were carved out, and their transcriptional connectivity next to epigenetic control verified. Remarkably, we were able to demonstrate that all cancer entities analyzed in this study comprised cell populations expressing ASIGs. Except for pancreatic cancer, all entities accounted for subpopulations with aging associated patterns. Using a novel pipeline to repurpose FDA-approved drugs and combinations on a single cell level, we discovered a path for novel treatment regimen on a tumor-specific basis. These respective cellular subpopulations demonstrate pillars of susceptibility within the complex tumor microenvironment and serve as a basis for future studies on the role of aging and senescence in human malignancies.

#5025

Single cell RNA sequencing reveals degenerative mechanisms in osteoblastic osteosarcoma and targetability of senescent subpopulations.

Dominik Saul, Robyn Laura Kosinsky. Mayo Clinic, Rochester, MN.

Osteosarcoma is the most common primary solid tumor of the bone, typically affecting children and adolescents. Surgery and adjuvant chemotherapy result in an unsatisfying 5-year survival rate of <25%. Although immune adjuvant strategies are promising, recent developments targeting the specialized tumor microenvironment by aiming at osteoclasts and checkpoint inhibitors focus on the individual immune biology. We aimed to mirror the tumor microenvironment and pseudotemporal development of osteoblasts during their transformation to osteoblastic osteosarcoma cells. For this purpose, single cell RNA-sequencing (scRNA-seq) profiles of 166,012 osteoblastic osteosarcoma and 49,555 healthy bone cells isolated from 28 patients were analyzed. In pseudotemporal analyses, we identified developmental divergence points with distinct transcriptional patterns purporting paths of devolution or giving rise to proliferative arrest. We found the biologically predicted clinging conjunction between osteoclasts, osteoblasts and B cells as regulatory T cells. In addition, we discovered intratumoral cellular communication patterns with ligand-receptor interactions and secreted signaling analysis between certain pillars of heterogeneous subpopulations, illustrating the importance of the local immune-genomic landscape. Upon several experimentally verified pathways, the TGF-β interaction was paramount. A mathematical assembly of degenerated cell populations revealed canonical markers of an osteoblastic cancerous state, out of which the vast majority has been associated with tumorigenesis in earlier studies. Intriguingly, nonnegative matrix factorization gave rise to a spatially localized senescent cell state, potentially accessible to new senolytic compounds which are currently in clinical trials for aging-related diseases. This senescent state with proliferative arrest and the secretion of a senescence-associated secretory phenotype (SASP), has not been described in osteosarcoma so far. In addition, we identified novel pathways of intertumoral communication in these distinct clusters, which can be exploited therapeutically, and validated these with a single-cell guided pipeline to repurpose drugs and their combinations. Thus, we were able to define a temporal pattern for the development of osteoblastic osteosarcoma. A local microenvironment with gradual states of devolution, and distinct cell-cell communicational patterns was established, out of which TGF-β warrants further experimental as therapeutical exploitation. Finally, we curtailed degenerated and, surprisingly, senescent subpopulations, each of which were predicted to vary in therapeutic response, mirroring the clinical observed intratumoral heterogeneity and potentially guiding the way to an individualized treatment regimen.

#5026

Dynamic changes in circulating protein levels reveal an association between ipilimumab and nivolumab combination treatment (SWOG Lung-MAP S1400I trial) with outcomes in squamous cell lung cancer.

Edgar Gonzalez-Kozlova,¹ Hsin-Hui Huang,¹ Mary Redman,² Roy Herbst,³ Scott Gettinger,⁴ Luda Bazhenova,⁴ Hui Xie,¹ Manishkumar Patel,¹ Kai Nie,¹ Jocelyn Harris,¹ Kimberly Argueta,¹ Karen Kelly,⁵ Ethan Cerami,⁵ James Lindsay,⁵ Joyce Yu,⁵ Roshni Biswas,⁵ Stephen Van Nostrand,⁵ Radim Moravec,⁶ Diane Marie Del Valle,¹ Seunghee Kim-schulze,¹ Sacha Gnjatic¹. ¹Icahn School of Medicine, New York, NY; ²SWOG Statistical Center; Fred Hutchinson Cancer Research Center, Seattle, WA; ³The University of Texas MD Anderson Cancer Center, Houston, TX; ⁴Yale Cancer Center, New Heaven, CT; ⁵Dana-Farber Cancer Institute, Boston, MA; ⁶NCI, Bethesda, MD.

Squamous cell lung cancer (SqCLC) is a type of non-small cell lung cancer strongly associated with cigarette smoking and with known targetable mutations, however some patients do not have matching targeted drugs. The S1400I Phase III randomized LungMap sub-study of nivo+ipi versus nivo accrued 275 (252 eligible) previously-treated patients with stage IV SqCLC and absence of matched mutations (NCT02154490).Here, to investigate the longitudinal (baseline, C2 week 3, C4 week 7, C5 week 9) serum protein changes associated with treatment or response, we performed Olink proximity extension assay in 160 patients (561 samples) using an immuno-oncology panel of 92 analytes. We utilized mixed linear and joint models to identify differentially expressed proteins between treatment arms (nivo vs. nivo+ipi), response and quantify the effect of other potential prognostic factors. This approach quantifies the effect of covariates such as smoking status, demographics, prior radiation therapy, and metastases. We jointly modeled expression and survival to identify changes in proteins over time. The thresholds for significance in differentially expression tests were a log fold change of at least 0.5 and a false discovery rate of under 0.05 (FDR as a multiple testing adjustment method). The joint models used thresholds of log fold change of more than 1 and hazard ratio of more than 1. Results revealed increases in 40 serum proteins after treatment with either nivo alone or combined with ipi, including CXCL9, CXCL10, CXCL11, CXCL13, IL6, IL8, IL10, IFNg, and soluble PD-L1 and PDCD1. nivo+ipi treatment showed greater increases in IL8 and CXCL13 post-treatment compared to nivo alone. In addition, circulating IL6, CXCL13, and MIC-A/B were higher in patients with stable or progressive disease compared to those with objective response after treatment. The joint model revealed a worse hazard ratio with increases in 10 soluble proteins, including IL6, IL8, and CXCL13. Finally, for patients with similar values of IL8, those treated by combination had a better survival outcome than those treated by nivo alone. Using a data-modeling approach, we identified significant longitudinal changes in 40 serum proteins out of 92 tested in patients with SqCLC treated with nivo or nivo+ipi. Increases in serum inflammatory markers IL6 and CXCL13 were associated with worse disease. Although overall survival was not statistically different between arms, our modeling approaches suggested that IL-8 monitoring may help identify patients benefiting more from the combination vs. nivo alone.

#5027

PTMcosmos: A web portal of post-translational modifications and proteogenomic resources in cancer.

Liang-Bo Wang, Akshay Govindan, Song Cao, Li Ding. Washington University in St. Louis, Saint Louis, MO.

PTMcosmos is a comprehensive database with an interactive web portal designed to catalog and visualize post-translational modifications (PTMs) in humans. It contains 469,183 experimentally-validated PTM sites and their supporting evidence from UniProt Knowledge Base, PhosphoSitePlus, and the Clinical Proteomic Tumor Analysis Consortium (CPTAC). PTMcosmos summarizes the entire spectrum of CPTAC proteomics data on human cancer patients, including protein and PTM peptide abundance data from 10 different cancer types. Additionally, PTMcosmos contains cancer somatic mutations from The Cancer Genome Atlas (TCGA), thus allowing for the collective integration and analysis of different data types. In PTMcosmos, we have built an ensemble of interactive visualization tools that allow researchers to investigate altered PTM functions due to genetic alterations in close proximity. The database is live at https://ptmcosmos.wustl.edu. We used PTMcosmos to investigate PTM regulation across cancer types. First, we examined the differential abundance of the PTM sites of cancer driver genes, focusing primarily on phosphorylation of the tumor suppressor retinoblastoma protein (encoded by RB1) and acetylation of the histone acetyltransferase E1A Binding Protein P300 (EP300) across cancer types. We analyzed the association between these PTM events and downstream targets, as well as with tumor subtypes, significantly mutated gene (SMG) mutation status, and clinical features. Second, we investigated the association of the protein abundance of cancer driver genes with ubiquitylsites in lung squamous cell carcinoma (LSCC) to nominate potentially novel modes of regulation of these proteins’ activities. We further analyzed the tumor subtype specificity and tumor-normal abundance changes of these ubiquitylsites and their corresponding substrate proteins, identifying several EGFR ubiquitylsites which may regulate EGFR abundance in LSCC. Finally, we identified the linear and spatial clustering of mutations and PTM sites, identifying multiple mutation-PTM clusters in cancer related genes, including TP53, PIK3CA, CTNNB1, EGFR, and IDH1. We envision that PTMcosmos will serve both the CPTAC consortium and the wider research community to better understand the role of PTMs in cancer.

#5028

Integration of molecular data into cancer patient database.

Basil H. Chaballout,¹ Avery T. Funkhouser,¹ Bailey B. Blair,¹ Jane L. Goodwin,¹ Alexander M. Strigenz,¹ Connie M. Arthur,² Julie C. Martin,³ W. Jeffery Edenfield,³ Anna V. Blenda¹. ¹University of South Carolina School of Medicine Greenville, Greenville, SC; ²Brigham and Women’s Hospital, Harvard Medical School, Boston, MA; ³Prisma Health Cancer Institute, Greenville, SC.

Background: Precision medicine holds promise of being a more effective method for treating complex diseases such as cancer. Nuanced care is needed, as the development and clinical course of cancer is multifactorial with influences from the general health status of the patient, germline and neoplastic mutations, co-morbidities, and environment including lifestyle. To tailor an individualized treatment to the patient, such multifactorial data must be presented in an easy-to-use, easy-to-analyze fashion for providers to use effectively.

Purpose: To address the need, we have built a searchable database integrating cancer-critical gene mutation status, serum galectin protein markers, serum and tumor glycomic profiles, with clinical, demographic, and lifestyle data points of individual patients.

Methods: The initial data was acquired from breast, colon, and lung cancer patients’ serum and biopsy samples from the Prisma Health Cancer Institute Biorepository. The acquired data contains the status of 2,800 COSMIC cancer-critical gene mutations, individual patient profiles of five serum galectins, and serum and biopsy glycan structures from each patient’s glycomic profile. DNA from tumor cells was used to screen the regions frequently mutated in human cancer genes. Multiplex PCR using Ion AmpliSeq™ Cancer Hotspot panel v2 was performed by Precision Genetics. Enzyme-linked immunosorbent assay was employed to perform galectin profiling of cancer patient serum samples. Glycomic profiling of serum and biopsy samples was performed by the Emory Comprehensive Glycomic Core. In addition, healthy control values for galectin and glycomic profiles were obtained and added to the patient database for reference. The data is being stored using Microsoft SQL servers and is fed into an interactive web application using RStudio.

Results: Our interactive database allows care providers to amalgamate cohorts from these groups to find correlations between different data types with the possibility of finding a stage signature based upon a combination of genetic mutations, galectin serum levels, glycan signatures, and patient clinical data and lifestyle choices.

Conclusion: Our project provides a framework for an integrated interactive database to analyze molecular and clinical patterns across cancer stages and provides opportunities for increased diagnostic and prognostic power.

#5029

A computational analysis of NEK10 and its novel protein-protein interaction with HspB1.

Andriele Silva, Shaneen Singh. CUNY Brooklyn College, CUNY Graduate Center, New York, NY.

The NEK kinase family of proteins consists of 11 serine/threonine kinases that participate in the disjunction of the centrosome, mitotic spindle assembly, and primary cilium formation. NEK10 is the most divergent member of the NEK family. It is unique in having a catalytic domain that is centrally positioned and flanked by two coiled-coil domains, while all the other NEKs have their catalytic domain near the N-terminus. Instead, NEK10 has four armadillo repeats of unexplored function in its N-terminus. NEK10 seems to play a key role in carcinogenesis and has been linked in melanoma, breast cancer, and a variety of ciliopathies. As part of our long-term goal to build interactomes of all the NEK members, we have previously reported data for known and predicted NEK10 interacting proteins, including novel protein-protein interactions such as HspB1 and MAP3K1, which have not been previously reported in the literature. In this study, we focused on understanding the molecular mechanism underlying NEK10’s interaction with HspB1 and its functional consequences. HspB1 is involved in various cellular functions such as maintaining cytoskeletal integrity and cell death. It is also a well-studied cancer protein and contains a highly conserved α-crystallin domain flanked by disordered N- and C-terminal domains. Many proteins are targeted by HspB1 to promote resistance to cell death, and malignant phenotypes; high levels of HspB1 have been identified in many cancer stem cells, such as those from lung and breast cancers. We used state-of-the-art computational approaches to model and characterize the full-length NEK10 as well as HspB1 protein. Our results offer robust full-length three-dimensional models of NEK10 and HspB1 and their biophysical characterization. In addition, we delineate the role of the armadillo repeats that are unique to NEK10 in the NEK family, in its interaction with HspB1. Our docking