WO2024077037A1 - Methods and compositions related to non-coding variants for the prediction of response to cancer immunotherapy - Google Patents

Abstract

The present disclosure relates to compositions and methods for the detection of 3' UTR somatic mutations and/or variants for use as prognostic assay for response to cancer immunotherapy. The present disclosure allows for the use of one or more non-coding mutations alone or in combination with non-coding mutations in genic (e.g. 5' UTR, introns, promoters) or intergenic loci, as well as coding variants, for the prediction of response to cancer immunotherapy.

Description

METHODS AND COMPOSITIONS RELATED TO NON-CODING VARIANTS FOR THE PREDICTION OF RESPONSE TO CANCER IMMUNOTHERAPY

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to U.S. Provisional Application No. 63/378,392, filed on October 5, 2022, the entire contents of which are incorporated herein.

FIELD

The present technology- relates to one or more non-coding variants in 3’ UTR regions for the prediction of a response to cancer immunotherapy.

GOVERNMENT INTEREST

This invention was made with government support under grant CA258776 and CA232105 awarded by the National Institutes of Health. The government has certain rights in the invention.

SEQUENCE LISTING

The instant application contains a sequence listing, which has been submitted in XML format via EFS-Web. The contents of the XML copy named “BID-013PC BIDMC 2020- 007_Sequence Listing"’, which was created on August 19, 2022 and is 16,384 bytes in size, the contents of which are incorporated herein by reference in their entirety.

BACKGROUND

Only a portion of cancer patients respond favorably to immunotherapy indicating the need for new biomarkers and targets. The landscape of non-coding mutations in cancer progression and immune evasion is largely unexplored. The small number of currently available cancer samples with combined whole genome sequencing (“WGS”) and RNAseq data prohibits the use of WGS variant calling data for analyses that require high statistical power, such as QTL analysis.

Accordingly, there is a need for new compositions and methods that provide improved leverage of transcriptomic data to identify mutations and biomarkers for effective therapies, as well as to understand non-coding mutations in cancer. SUMMARY

Therefore, the present disclosure provides, in aspects, a method of selecting a patient for treatment with an immunotherapy, comprising: (a) obtaining a biological sample from the patient; (b) assaying the biological sample for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; (c) classifying the patient as a likely responder or a likely non-responder to the immunotherapy based on the presence, absence or level of the one or more non-coding somatic and/or germline mutations in the 3' UTR regions; and (d) selecting the patient for treatment with the immunotherapy if the patient is classified as a likely responder.

In aspects, the present disclosure provides a method of predicting a patient response to treatment with an immunotherapy, comprising: (a) obtaining a biological sample from the patient; (b) assaying the biological sample for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; (c) classifying the patient as a likely responder or a likely non-responder to the immunotherapy based on the presence, absence or level of the one or more non-coding somatic and/or germline mutations in the 3' UTR regions.

In aspects, the present disclosure provides a method of preventing or treating cancer in a patient, comprising: (a) selecting the patient for treatment with an immunotherapy, the selection being based on classifying a patient as suitable for treatment with an immunotherapy based on assaying a biological sample from the patient for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; and (b) administering an effective amount of the immunotherapy to the patient.

In aspects, the present disclosure provides a method of modulating an immune landscape phenotype in a patient, comprising: (a) selecting the patient for treatment with an immunotherapy, the selection being based on classifying a patient as suitable for treatment with an immunotherapy based on assaying a biological sample from the patient for one or more noncoding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; and (b) administering an effective amount of the immunotherapy to the patient. In some embodiments, the modulation of the immune landscape phenoty pe comprises one or more of modulating immune cell infiltration and modulating T cell and B cell receptor diversity, and modulating antigen presentation or processing. In any of the above aspects, the patient is afflicted with a cancer.

In some embodiments, the method further comprises obtaining a biological sample from the patient, sending the biological sample to a diagnostic laboratory', and receiving from the diagnostic laboratory' a report that states whether the patient is a likely responder to an immunotherapy.

In some embodiments, the biological sample is selected from blood, plasma, serum, mucus, stool, sputum, saliva, nasal secretion, lavage fluid, respiratory fluid, and combinations thereof. In some embodiments, the biological sample is selected from a tumor biopsy, tumor resection, frozen tumor tissue specimen, lymph node, bone marrow, circulating tumor cells, cultured cells, a formalin-fixed paraffin embedded tumor tissue specimen, bronchoalveolar lavage, skin, hair, urine, and combinations thereof. In some embodiments, the tumor biopsy is selected from a core biopsy, needle biopsy, surgical biopsy, and an excisional biopsy.

In some embodiments, the assaying comprises one or more of polymerase chain reaction (PCR) amplification reaction, reverse-transcriptase PCR analysis, quantitative realtime PCR, single-strand conformation polymorphism analysis (SSCP), mismatch cleavage detection, heteroduplex analysis, deoxyribonucleic (DNA) acid sequencing, ribonucleic acid (RNA) sequencing. Northern blot analysis, in situ hybridization, array analysis, restriction fragment length polymorphism analysis, and combinations thereof. In some embodiments, the method comprises detection with one or more next generation sequencing (NGS) methods. In some embodiments, the method comprises ribonucleic acid (RNA) sequencing, whole genome sequencing, or targeted (probe or amplification-based) genome sequencing, such as panel sequencing and/or whole exome sequencing.

In some embodiments, the mutations comprise one or more of single-nucleotide variants (SNVs) and/or short insertions-deletions (indels), optionally wherein the SNVs and/or indels is located within the 3’ UTR. In some embodiments, the mutations or site of mutations are putative or experimentally-supported miRNA and/or RBP binding sites. In some embodiments, the mutations are selected from one or more of the entries of Table 4 below, or are located in one or more of the entries of Table 5, or are located in one or more of the entries of Table 7.

In some embodiments, the mutations are in one or more of those in FIG. 4B. In some embodiments, the mutations are in one or more of B2M. HLA genes, CANX, I.DHA. PSMB2, HNRNPR, WARS, APOBEC3C, STAT1, and ADAR. In some embodiments, the mutation is chrl: 154583325, T-to-C in ADAR.

In some embodiments, the classifying comprises one or more of gene expression quantitative trait loci (eQTL) and immune landscape QTL (ilQTL) analysis. In some embodiments, the classifying comprises comparing the results of the assaying to one or more reference datasets that are indicative of patient response or non-response to the immunotherapy. In some embodiments, the classifying comprises combining the results of the assaying into a polygenic risk score, signature, mathematical model, statistical model, or machine learning classifier. In some embodiments, the classifying comprises combining the results of the assaying and one or more reference datasets that are indicative of patient response or nonresponse to the immunotherapy into a polygenic risk score, signature, mathematical model, statistical model, or machine learning classifier. In some embodiments, the classifying comprises one or more steps of FIG. 1A.

In some embodiments, the method further comprises assaying one or more non-coding mutations in genic loci or intergenic loci of the patient's genome. In some embodiments, the genic loci are one or more of a 3'UTR, 5' UTR, intron, and promoters. In some embodiments, the method further comprises assaying one or more coding mutations of the patient’s genome.

In some embodiments, the method comprises merging and/or comparing (i) one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome, with (ii) one or more non-coding mutations in genic loci or intergenic loci of the patient’s genome, and/or (iii) one or more coding mutations of the patient’s genome.

In some embodiments, the method comprises merging and/or comparing (i) one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome, with (ii) one or more non-coding mutations in genic loci or intergenic loci of the patient’s genome, and/or (iii) one or more coding mutations of the patient’s genome with (iv) one or more reference datasets that are indicative of patient response or non-response to the immunotherapy.

In some embodiments, the method further comprises assaying whether the biological sample is positive or negative for gene expression of one or more checkpoint marker. In some embodiments, the checkpoint marker is selected from programmed death 1 (PD-1), programmed death-ligand 1 (PD-L1), programmed death-ligand 2 (PD-L2), Cytotoxic T Lymphocyte Antigen-4 (CTLA-4), Lymphocyte Activation Gene-3 (LAG-3), T cell Immunoglobulin and Mucin protein-3 (TIM-3), B7-H3, B7-H4, inducible T cell co-stimulator (ICOS), Sialic acid-binding immunoglobulin-type lectin 7 (SIGLEC7), Sialic acid-binding immunoglobulin-type lectin 9 (SIGLEC9), and V-domain Ig suppressor of T cell activation (VISTA).

In some embodiments, the biological sample has been immunohistochemically stained for the expression of PD-L1, PD-L2 or both PD-L1 and PD-L2.

In some embodiments, the immunotherapy is an agent that modulates one or more checkpoint biomarker, selected from PD-1, PD-L1, PD-L2, CTLA-4, LAG-3, TIM-3, B7-H3, B7-H4, ICOS, SIGLEC7, SIGLEC9, and VISTA. In some embodiments, the immunotherapy is an agent that modulates one or more PD-1, PD-L1, PD-L2. and CTLA-4. In some embodiments, the immunotherapy is a cellular immunotherapy treatment. In some embodiments, the agent that modulates: PD-1 is an antibody or antibody format specific for PD-1; PD-L1 is an antibody or antibody format specific for PD-L1; PD-L2 is an antibody or antibody format specific for PD-L2; or CTLA-4 is an antibody or antibody format specific for CTLA-4.

In some embodiments, the antibody or antibody format is selected from one or more of a monoclonal antibody, polyclonal antibody, antibody fragment, Fab, Fab', Fab'-SH, F(ab')2, Fv, single chain Fv, diabody, linear antibody, bispecific antibody, multispecific antibody, chimeric antibody, humanized antibody, human antibody, and a fusion protein comprising the antigen-binding portion of an antibody. In some embodiments, the antibody or antibody format specific for PD-1 is selected from nivolumab, pembrolizumab, and pidilizumab. In some embodiments, the antibody or antibody format specific for PD-L1 is selected from atezolizumab, avelumab. durvalumab. and BMS-936559. In some embodiments, the antibody or antibody format specific for CTLA-4 is selected from ipilimumab and tremelimumab.

In some embodiments, the method further comprises administering an agent that modulates one or more PD-1, PD-L1. PD-L2, and CTLA-4. In some embodiments, the administration is sequential or simultaneous. In some embodiments, the administration is by intratumoral, intradermal, subcutaneous, intramuscular, intraperitoneal or intravenous injection, or direct injection into cancer tissue. In some embodiments, the administration is intratumor al.

In some embodiments, the treatment with the immunotherapy comprises reducing or eliminating immune evasion.

In some embodiments, the treatment with the immunotherapy comprises modulation to enable successful subsequent or concurrent treatment with immunotherapy.

In some embodiments, the cancer is selected from one or more of basal cell carcinoma, biliary tract cancer; bladder cancer; bone cancer; brain and central nervous system cancer; breast cancer; cancer of the peritoneum; cervical cancer; choriocarcinoma; colon and rectum cancer; connective tissue cancer; cancer of the digestive system; endometrial cancer; esophageal cancer; eye cancer; cancer of the head and neck; gastric cancer (including gastrointestinal cancer); glioblastoma; hepatic carcinoma; hepatoma; intra-epithelial neoplasm; kidney or renal cancer; larynx cancer; leukemia; liver cancer; lung cancer (e.g., small-cell lung cancer, non-small cell lung cancer, adenocarcinoma of the lung, and squamous carcinoma of the lung); melanoma; myeloma; neuroblastoma; oral cavity cancer (lip, tongue, mouth, and pharynx); ovarian cancer; pancreatic cancer; prostate cancer; retinoblastoma; rhabdomyosarcoma; rectal cancer; cancer of the respiratory system; salivary gland carcinoma; sarcoma; skin cancer; squamous cell cancer; stomach cancer; testicular cancer; thyroid cancer; uterine or endometrial cancer; cancer of the urinary' system; vulval cancer; lymphoma including Hodgkin's and non-Hodgkin's lymphoma, as well as B-cell lymphoma (including low grade/follicular non-Hodgkin’s lymphoma (NHL); small lymphocytic (SL) NHL; intermediate grade/follicular NHL; intermediate grade diffuse NHL; high grade immunoblastic NHL; high grade lymphoblastic NHL; high grade small non-cleaved cell NHL; bulky disease NHL; mantle cell lymphoma; AIDS-related lymphoma; and Waldenstrom's Macroglobulinemia; chronic lymphocytic leukemia (CLL); acute lymphoblastic leukemia (ALL): Hairy cell leukemia; chronic myeloblastic leukemia; as well as other carcinomas and sarcomas; and post-transplant lymphoproliferative disorder (PTLD), as well as abnormal vascular proliferation associated with phakomatoses, edema (e.g. that associated with brain tumors), and Meigs’ syndrome. In some embodiments, the cancer is melanoma. In some embodiments, the cancer is a gastric cancer. In some embodiments, the method identifies a patient for immunotherapy treatment who would not otherwise be identified for immunotherapy treatment using a IHC test for expression of a checkpoint biomarker, optionally selected from PD-1, PD-L1, PD-L2, CTLA- 4, LAG-3, TIM-3, B7-H3, B7-H4, ICOS, SIGLEC7, SIGLEC9, and VISTA.

In aspects, the present disclosure provides a method of identifying an immune cell gene signature profile which is correlative with responsiveness or non-responsiveness to an immunotherapy, the method comprising one or more steps of FIG. 1 A and/or FIG. 4A.

In aspects, the present disclosure provides a companion diagnostic, complementary diagnostic, or codiagnostic test kit, comprising: (a) an array of nucleic acids suitable for detection of one or more non-coding somatic and/or germline mutations in 3' UTR regions of a patient’s genome, the one or more non-coding somatic and/or germline mutations in 3 ' UTR regions being indicative of a patient response or non-response to an immunotherapy; and (b) instructions for use.

In aspects, the present disclosure provides a companion diagnostic, complementary diagnostic, or codiagnostic test kit. comprising reagents and instructions for use in one or more of any of the preceding aspects and/or embodiments.

The details of one or more examples of the disclosure are set forth in the description below. Other features or advantages of the present disclosure will be apparent from the following drawings, detailed description of several examples, and also from the appended claims. The details of the disclosure are set forth in the accompanying description below. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present disclosure, illustrative methods and materials are now described. Other features, objects, and advantages of the disclosure will be apparent from the description and from the claims. In the specification and the appended claims, the singular forms also include the plural unless the context clearly dictates otherwise. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

DESCRIPTION OF THE DRAWINGS

FIG. 1 A shows an image of the approach from biological sample analysis, identification of 3’ UTR variants, and modeling, to the extraction of actionable outcomes. FIG. IB is a graph showing the distribution of the relative position of single nucleotide variant calls along the 3’ UTR as identified by analysis of TCGA WES, RNAseq and WGS data. FIG. 1C is a graph showing the number of RNAseq-derived somatic 3’ UTR single nucleotide variants per sample, grouped by “Mutation Rate Category” as defined by TCGA WES variant-calling data. FIG. ID shows overlap between variant calls in the analysis of RNAseq data with GATK and Strelka2 (GATK=5431118, Strelka2=10175223, Overlap=4692062).

FIG. 2A and FIG. 2B are graphs and images showing the analysis of PD-L1 3’ UTR variants in TCGA STAD. FIG. 2A is a graph showing normalized PD-L1 expression (log2(read count+1)) in patients homozygous to the reference vs the alternative allele. For all three SNPs studied, patients homozygous for the alternative allele show higher levels of PD- L1 expression than those homozygous for the reference allele. FIG. 2B is an image showing the results of Luciferase assays performed with the reference vs alternative allele full-length PD-L 1 3 ' UTR psicheck2 reporter. Shown is luc expression relative to the full length reference PD-L1 3’ UTR. Significance was assessed by Student's t-test. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001

FIG. 3A to FIG. 3D are graphs and images showing the identification and characterization of 3' UTR cis-eQTLs in TCGA STAD. FIG. 3 A is a waffle plot show ing the percentage of eQTLs that mapped to each of the indicated gene biotypes. A waffle plot consists of 100 squares and the number of colored squares represents the percentage of eQTLs mapping to each gene biotype. FIG. 3B is a barplot showing the number of protein-coding cis-eQTLs that reside in each genic region (5’ UTR, CDS or 3’ UTR). FIG. 3C is a graph showing the number of 3’ UTR cis-eQTLs that overlap with miRNA or RBP binding sites in the indicated databases. TarBase is a database of experimentally validated miRNA binding sites, microT includes predicted miRNA binding sites and POSTAR2 contains predicted RBP binding sites based on CLIP data. FIG. 3D is a graph showing KEGG pathway enrichment analysis of the 500 topmost significant 3’ UTR cis-eGenes. X-axis represents the number of genes bearing a c .s-eQTL present in the relevant gene set..

FIG. 4A and FIG. 4B are graphs and images showing the massive parallel reporter assay for the characterization of 3’ UTR immune-related cis-eQTLs. FIG. 4A is an image showing a protocol followed for 3’ UTR MPRA assay. Ref: Reference allele, Alt. allele: Alternative allele, Syn. pA terminator: Synthetic polyadenylation terminator, oligodT: primer containing 20 T nucleotides. FIG. 4B shows a Manhattan plot of the top hits from the MPRA assay in both gastric cancer cell lines. The black line (p.adj.=0.05) separates significant from nonsignificant calls. Blue dots represent likely germline variants, while red dots represent likely somatic variants.

FIG. 5 shows a graph of the identification and characterization of 3’ UTR immune- related ilQTLs in TCGA STAD. The left-hand side of the dashed black line shows the distribution of significant ilQTLs along genic regions (5’ UTR, CDS or 3’ UTR), showing enrichment in 3’ UTR variants. The right-hand side of the dashed black line shows the overlap of 3’ UTR ilQTLs with databases for miRNA response elements (mRE) and RBP binding sites. For each region (3' UTR, 5’ UTR. CDS, Predicted mRE, Validated mRE, and RBP), the order of the bars on the graph is the following from left to right: BCR shannon, CD8+ T cells. Leukocyte ratio, Marcophages Ml, Neutrophils, NK cells (activated), NK cells (resting), TCR shannon, and Treg (regulatory⁷ T cells).

FIG. 6A to FIG. 6C shows graphs of the identification of causal 3‘ UTR eQTLs and ilQTLs in ADAR. FIG. 6A is a plot showing the distribution of CDS and 3’ UTR eQTLs and ilQTLs for the ADAR gene. The location of the exon where the 3’ UTR is found is indicated with the arrow. FIG. 6B is a graph showing a cohort of gastric cancer and melanoma patients was classified into R (responders) vs NR (non-responders) according to response efficacy with anti-PD-1 immunotherapy. Primary cancer RNAseq data from these patients were analyzed. Differential expression analysis revealed increased expression of ADAR in R compared to NR. A Wilcoxon rank sum test was used to assess the significance of the comparison. FIG. 6C is a graph showing correlation analysis between ADAR1 and TARDBP log2 TPM normalized expression performed using TIMER v2.0 (timer.cistrome.org).

FIG. 7A and FIG. 7B are graphs showing ilQTL polygenic risk score for predicting response to immunotherapy in melanoma and gastric cancer patients. FIG. 7A is a graph showing the comparison of the polygenic risk score (PRS) distribution in the non-responder (NR) and responder (R) groups of the testing population. A Student’s t-test was used to assess the significance of the comparison. FIG. 7B is a graph showing a Receiver Operator Characteristic (ROC) curve and the ability of the PRS score and PD-L1 expression classifications to distinguish between R and NR patients in the testing population. An Area Under the Curve (AUC) score is reported for both classifiers. FIG. 8A and FIG. 8B are graphs showing the per-sample number of 3‘ UTR variants. The distribution of the number of 3’UTR (FIG. 8A) SNVs. and (FIG. 8B) short indels per sample called by the analysis of TCGA STAD RNAseq data are shown.

FIG. 9 is a chart showing enrichment of significant cis-eQTL variants in 3’ UTRs. Comparing the average length of each genic region (5’UTR, CDS, 3'UTR) to the number of significant cis-eQTLs residing in those regions, reveals an enrichment of significant variants in the 3’UTR.

FIG. 10A and FIG. 10B are graphs showing the comparison of the MPRA results from AGS and SNU719 cell lines. FIG. 10A shows volcano plots for the results from the MPRA assay with SNU719 (first column) and AGS (second column) cells. The dotted line defines the cutoff for a significant call, any dots above the line have an adjusted p value lower than 5e-2. Significant calls from the SNU719 assay are colored red in the first row, while significant calls from the AGS assay are shown in red in the second row. FIG. 10B shows a Manhattan plot of the top hits from the MPRA assay, with the AGS cell line shown in the left plot, and the SNU719 cell line show n in the right plot.

FIG. 11 is a graph showing enrichment analysis in top CD8+ T cell infiltration QTL variants. KEGG pathway analysis in genes with significant variants as CD8+ T cell infiltration QTLs reveals enrichment in ribosome-related pathways.

FIG. 12A and FIG. 12B are graphs showing how ADAR eQTL and ilQTL variants appear as hits in the MPRA assay. Manhattan plots of the results of the 3’ UTR MPRA assay in the tw o gastric cancer cell lines, (FIG. 12A) AGS cell line, and (FIG. 12B) SNU719 cell line. The point size for each variant increases with increasing absolute log2FC. The ADAR chrl: 154583325 (T-to-C) variant is shown in the text box (green), while all other variants are colored grey. Dots above the dashed red line (p.adjust=log2(0.05)) represent significant variants.

DETAILED DESCRIPTION

The present disclosure is based, in part, on the discovery that one or more non-coding mutations in 3' untranslated regions (3’ UTRs) alone or in combination with non-coding mutations in genic (e.g. 5’ UTR, introns, promoters) or intergenic loci as w ell as coding variants are particularly useful for the prediction of response to cancer immunotherapy. The present disclosure identifies, inter alia, common somatic and germline 3' untranslated region (3'UTR) variants across the human transcriptome from 375 gastric patients from The Cancer Genome Atlas. By performing gene expression quantitative trait loci (eQTL) and immune landscape QTL (ilQTL) analysis, disclosed herein are 3' UTR variants with cis effects on expression, as well as effects on immune landscape phenoty pes, including immune cell infiltration and T cell and B cell receptor diversity. The majority of eQTLs and ilQTLs overlapped with predicted or experimentally-supported microRNA (miRNA) and RNA-binding protein (RBP) response elements, revealing key regulatory' regions in the target genes. To distinguish between causal and correlative effects of 3' UTR eQTLs in immune-related genes, the experiments disclosed herein include a massive parallel reporter assay (MPRA). The experiments confirmed how 3' UTR eQTLs. such as those in Programmed death-ligand 1 (PD-L1), and numerous new sites (e.g., in ADAR1 provide a unique resource for the identification of novel immunotherapeutic targets and biomarkers. As disclosed herein, the prioritized ilQTL variant signature predicted response to immunotherapy better than standard of care PD-L1 expression in independent patient cohorts, showcasing the untapped potential of non-coding mutations in cancer.

Messenger RNA 3' Untranslated Regions (3’ UTRs) are primary sites for post- transcriptional regulatory' events. These processes account for -60% of the variation in protein expression, while about 20% of germline expression quantitative trait loci (eQTLs) are located in 3 ' UTRs, which are more conserved than other noncoding loci, suggesting selective pressure. 3 ' UTRs are the most common targets of key regulatory molecules, such as RNA binding proteins (RBPs) and microRNAs (miRNAs).

MiRNAs are potent post-transcriptional regulators, implicated in the control of numerous cellular mechanisms, as well as of all cancer hallmarks, hence their role in cancer immune surveillance has become a research hotspot. miRNAs have been found to efficiently regulate Programmed death ligand 1 (PD-L1), other B7 family members, cytokines and numerous immune genes. On the other hand. RBPs have been shown to regulate mRNA processing, localization, interactions, and stability', while different RBPs such as Mex3B, Mex3C, and HNRNPR have been show n to regulate key antigen presentation mechanisms.

However, tumors mutate or truncate their 3" UTRs to escape this tight regulatory control. Unfortunately, the current reliance of variant-calling pipelines on whole exome sequencing (WES) data, which do not include probes for 3 ' UTR regions, has resulted in a lack of understanding of the 3' UTR variant’s role in cancer progression. Small-scale, targeted studies have identified individual 3 ' UTR somatic mutations that associate with changes in cis- gene expression and immune phenotypes, especially for PD-L1. For instance, a common somatic mutation in the PD-L1 3' UTR has been shown to disrupt miR-570 binding leading to increased expression.

Disclosed herein is, inter aha, a comprehensive mutational analysis on raw RNAseq data from hundreds of stomach adenocarcinoma (STAD) samples in TCGA to identify 3 ' UTR germline and somatic single-nucleotide variants (SNVs), as well as short insertions-deletions (indels). By performing a quantitative trait loci (QTL) analysis, c/s-acting gene expression QTLs (cA-eQTLs) were identified, as well as variants associated with changes in immune phenotypes, herein termed immune landscape QTLs (ilQTLs). A massively parallel reporter assay (MPRA) was designed and implemented to validate at scale cA-eQTLs in immune- related genes directly affecting post-transcriptional stability of respective genes. MPRAs have been utilized successfully in the past for the functional validation of non-coding variants, such as promoter and UTR germline variants. Further, apart from the 1.188 PCAWG samples overlapping with TCGA, WGS data are not available for about 90% of TCGA subjects, limiting large scale 3' UTR variant investigations. Specifically for gastric cancer, only 40 samples comprise WGS data. The only study to date which attempted to analyze 3' UTR variants in TCGA in non-WGS samples, mistakenly considered that 3' UTR regions were captured in the WES probe sets used in the study. As disclosed herein, these regions are not covered in the TCGA WES kits.

The present disclosure is the first to investigate the translational potential of somatic 3' UTR variants for their ability to inform patient stratification for immunotherapy and to predict outcomes across diverse cohorts of immune checkpoint inhibition. Utilizing the prioritized ilQTLs, a polygenic risk score was established, which surprisingly proved more accurate in predicting response to checkpoint inhibition in melanoma and gastric cancer patients than PD- L1 expression, providing the first direct support of the potential utility of UTR variants in predictive modeling in immunotherapy. The present disclosure establishes the tools and apply them to unbiasedly identify transcriptome- wide 3' UTR variants associated with changes in cA-gene expression and immune phenotypes in cancer and lay the foundations for similar 3' UTR-focused studies in other cancer types.

By repurposing RNAseq data available for hundreds of gastric adenocarcinoma patients in TCGA, disclosed herein is the first study to deeply investigate 3 ' UTR somatic and germline variants in cancer and their ability to affect the tumor immune landscape. The present reveals the importance of 3'UTR variants in driving czs-gene expression in cancer and provides a framework for incorporating 3'UTR variant-calling data in TCGA and other cohorts. Functional variants were prioritized by performing a transcriptome-wide c/s-eQTL analysis in the TCGA STAD cohort and identified significant variants across 1117 eGenes. Around 90% of the 3'UTR eQTLs overlap with putative or experimentally-supported miRNA and RBP binding sites, providing a potential functional relevance for those variants. The enrichment in immune-related pathways in the topmost significant 3'UTR cA-eQTLs indicates the importance of 3 'UTR variants in controlling cancer immunogenicity. Since 3’ UTR regulatory roles go beyond post transcriptional gene expression regulation and include localization, translation rate control, and even protein-protein interactions and liquid organelle formation, a transcriptome-wide ilQTL analysis was performed. In this analysis, the majority of significant variants resided near and/or within 3’ UTR regions. As disclosed herein, this is the first time that somatic 3'UTR variants have been shown to correlate with immune phenotype changes in an unbiased large-scale study. The present disclosure is based, in part, on the discovery’ of significant 3'UTR ilQTL vanants in widely studied immunoregulatory genes, such as ADAR and STAT1.

In addition to validating previously described immune-related 3'UTR eQTLs, the present disclosure also identified novel 3' UTR variants and genes with immune-related functions in cancer. Interestingly, among the significant hits from the MPRA assay, there w’ere multiple genes encoding ribosomal subunits. Pathway analysis in the top significant CD8+ T cell infiltration QTL variants revealed an overall enrichment in ribosome-related proteins (FIG. H).

To investigate the clinical relevance of the ilQTL analysis disclosed herein, the experiments of the present disclosure demonstrate that non-coding 3'UTR ilQTL variants predict response rates to immunotherapy treatments (FIG. 7A). In these experiments, the previously unexplored space of non-coding variants were investigated, and showed that a signature of ilQTL variants has stronger predictive powder for drug response than PD-L1 expression in a cohort of melanoma and gastric cancer patients (FIG. 7B). This is the first application of non-coding mutations to predict response to immunotherapy, providing a potentially strong reason to include these important regulatory regions in WES investigations. Therefore, the present disclosure provides, in aspects, a method of selecting a patient for treatment with an immunotherapy, comprising: (a) obtaining a biological sample from the patient; (b) assaying the biological sample for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; (c) classifying the patient as a likely responder or a likely non-responder to the immunotherapy based on the presence, absence or level of the one or more non-coding somatic and/or germline mutations in the 3' UTR regions: and (d) selecting the patient for treatment with the immunotherapy if the patient is classified as a likely responder.

In aspects, the present disclosure also provides a method of predicting a patient response to treatment with an immunotherapy, comprising: (a) obtaining a biological sample from the patient; (b) assaying the biological sample for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; (c) classifying the patient as a likely responder or a likely non-responder to the immunotherapy based on the presence, absence or level of the one or more non-coding somatic and/or germline mutations in the 3' UTR regions.

In aspects, the present disclosure provides a method of preventing treating cancer in a patient, comprising: (a) selecting the patient for treatment with an immunotherapy, the selection being based on classifying a patient as suitable for treatment with an immunotherapy based on assaying a biological sample from the patient for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; and (b) administering an effective amount of the immunotherapy to the patient.

In aspects, the present disclosure further provides a method of modulating an immune landscape phenofype in a patient, comprising: (a) selecting the patient for treatment with an immunotherapy, the selection being based on classifying a patient as suitable for treatment with an immunotherapy based on assaying a biological sample from the patient for one or more noncoding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient's genome; and (b) administering an effective amount of the immunotherapy to the patient. In some embodiments, the modulation of the immune landscape phenotype comprises one or more of modulating immune cell infiltration and modulating T cell and B cell receptor diversity, and modulating antigen presentation or processing. In some embodiments, the mutations comprise one or more of single-nucleotide variants (SNVs) and/or short insertions-deletions (indels). optionally wherein the SNVs and/or indels is located within the 3’ UTR. In some embodiments, the mutations or site of mutations are putative or experimentally-supported miRNA and/or RBP binding sites. In some embodiments, the mutations are selected from one or more of the entries of Table 4 below, or are located in one or more of the entries of Table 5 below, or are located in one or more of the entries of Table 7 below. In some embodiments, the mutations are selected from about 1, about 5, about 10, about 20, about 25, about 30, about 35, about 40, about 45, about 50, about 60, about 70, about 80, about 90, or about 100 entries of Table 4. In some embodiments, the mutations are selected from about 1, about 5, about 10, about 20, about 25, about 30, about 35, about 40. about 45, about 50, about 60, about 70. about 80, about 90, or about 100 entries of Table 5. In some embodiments, the mutations are in one or more of B2X HLA genes, CANX, LDHA, PSMB2, HNRNPR, WARS. APOBEC3C, STATE and ADAR.

In some embodiments, the mutation is chrl: 154583325. T-to-C in ADAR.

In some embodiments, the classifying comprises one or more of gene expression quantitative trait loci (eQTL) and immune landscape QTL (ilQTL) analysis.

In some embodiments, eQTL analysis includes gene-level expression. In some embodiments eQTL analysis includes gene-level expression from TCGA. In some embodiments, a mixed model or a linear model such as FastQTL is utilized to call eQTLs. In some embodiments, gene expression across libraries is normalized using trimmed mean of m- values as implemented in edgeR. In some embodiments, genes are selected based on an expression threshold of 1 read in at least 80% of the samples. In some embodiments, an inverse quantile normal transformation was performed on the expression values prior to their inclusion into the linear model. In some embodiments, a nominal p-value cutoff of le-7 identifies significant calls. In some embodiments, eQTL calls are mapped to transcript annotations (Gencode v32), while relative genomic locations (5 'UTR, CDS, 3 'UTR) are assigned using annotatr.

In some embodiments, iQTL analysis includes immune data per sample obtained from the Cancer Research Institute (CRI) i Atlas project. In some embodiments, immune data is calculated directly from next generation sequencing data, and/or other modalities, such as imaging, tissue staining, multiplexed immunofluorescence, or spatial tissue profiling. In some embodiments, QTL analysis includes the use of a linear or mixed model such as the model including genotype and covariate data as described above, while using immune profile estimates or quantile-normalized immune profile estimates as phenotypes. ilQTL selection as well as genomic and regulatory annotation are performed as for eQTLs.

In some embodiments, the classifying comprises comparing the results of the assaying to one or more reference datasets that are indicative of patient response or non-response to the immunotherapy. In some embodiments, the classifying comprises combining the results of the assaying into a polygenic risk score, signature, mathematical model, statistical model, or machine learning classifier (supervised or unsupervised). In some embodiments, the classifying comprises combining the results of the assaying and one or more reference datasets that are indicative of patent response or non-response to the immunotherapy into a polygenic risk score, signature, mathematical model, statistical model, or machine learning classifier (supervised or unsupervised). In some embodiments, the classifying comprises one or more steps of FIG. 1A.

In some embodiments, the method further comprises assaying one or more non-coding mutations in genic loci or intergenic loci of the patient's genome. In some embodiments, the genic loci are one or more of a 3'UTR, 5' UTR, intron, and promoters. In some embodiments, the method further comprises assaying one or more coding mutations of the patient’s genome.

In some embodiments, the method comprises merging and/or comparing (i) one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome, with (ii) one or more non-coding mutations in genic loci or intergenic loci of the patient’s genome, and/or (iii) one or more coding mutations of the patient’s genome.

In some embodiments, the method comprises merging and/or comparing (i) one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome, with (ii) one or more non-coding mutations in genic loci or intergenic loci of the patient’s genome, and/or (iii) one or more coding mutations of the patient’s genome with (iv) one or more reference datasets that are indicative of patient response or non-response to the immunotherapy.

In some embodiments, the method further comprises assaying whether the biological sample is positive or negative for gene expression of one or more checkpoint markers. In some embodiments, the checkpoint marker is selected from programmed death 1 (PD-1), programmed death-ligand 1 (PD-L1), programmed death-ligand 2 (PD-L2), Cytotoxic T Lymphocyte Antigen-4 (CTLA-4), Lymphocyte Activation Gene-3 (LAG-3), T cell Immunoglobulin and Mucin protein-3 (TIM-3), B7-H3, B7-H4, inducible T cell co-stimulator (ICOS), Sialic acid-binding immunoglobulin-type lectin 7 (SIGLEC7), Sialic acid-binding immunoglobulin-type lectin 9 (SIGLEC9), and V-domain Ig suppressor of T cell activation (VISTA).

In some embodiments, the biological sample has been immunohistochemically stained for the expression of PD-L1, PD-L2 or both PD-L1 and PD-L2.

In some embodiments, the immunotherapy is an agent that modulates one or more checkpoint biomarker, selected from PD-1, PD-L1, PD-L2, CTLA-4, LAG-3, TIM-3, B7-H3, B7-H4, ICOS, SIGLEC7, SIGLEC9, and VISTA. In some embodiments, the immunotherapy is an agent that modulates one or more PD-1, PD-L1, PD-L2. and CTLA-4. In some embodiments, the agent that modulates: PD-1 is an antibody or antibody format specific for PD-1; PD-L1 is an antibody or antibody format specific for PD-L1; PD-L2 is an antibody or antibody format specific for PD-L2; or CTLA-4 is an antibody or antibody format specific for CTLA-4. In some embodiments, the immunotherapy is a cellular immunotherapy treatment.

In some embodiments, the antibody or antibody format is selected from one or more of a monoclonal antibody, polyclonal antibody, antibody fragment, Fab, Fab', Fab'-SH, F(ab')2, Fv, single chain Fv, diabody, linear antibody, bispecific antibody, multispecific antibody, chimeric antibody, humanized antibody, human antibody, and a fusion protein comprising the antigen-binding portion of an antibody. In some embodiments, the antibody or antibody format specific for PD-1 is selected from nivolumab, pembrolizumab, and pidilizumab. In some embodiments, the antibody or antibody format specific for PD-L1 is selected from atezolizumab, avelumab. durvalumab, and BMS-936559. In some embodiments, the antibody or antibody format specific for CTLA-4 is selected from ipilimumab and tremelimumab.

In some embodiments, the method further comprises administering an agent that modulates one or more PD-1, PD-L1, PD-L2, and CTLA-4. In some embodiments, the administration is sequential or simultaneous. In some embodiments, the administration is by intratumoral, intradermal, subcutaneous, intramuscular, intraperitoneal or intravenous injection, or direct injection into cancer tissue. In some embodiments, the administration is intratumoral. In any of the above aspects, the patient is afflicted with a cancer.

As disclosed herein, administering, or administering a treatment/therapy, refers to a treatment/therapy from which a subject receives a beneficial effect, such as the reduction, decrease, attenuation, diminishment, stabilization, remission, suppression, inhibition or arrest of the development or progression of cancer and/or a genetic disease or disorder, or a symptom thereof.

In some embodiments, the treatment/therapy that a subject receives, or the prevention in the onset of cancer and/or a genetic disease or disorder results in at least one or more of the follow ing effects: (1) the reduction or amelioration of the severity of cancer and/or a genetic disease or disorder, and/or a symptom associated therewith; (2) the reduction in the duration of a symptom associated with cancer and/or a genetic disease or disorder; (3) the prevention in the recurrence of a symptom associated with cancer and/or a genetic disease or disorder; (4) the regression of cancer and/or a genetic disease or disorder, and/or a symptom associated therewith; (5) the reduction in hospitalization of a subject; (6) the reduction in hospitalization length; (7) the increase in the survival of a subject; (8) the inhibition of the progression of cancer and/or a genetic disease or disorder and/or a symptom associated therewith; (9) the enhancement or improvement the therapeutic effect of another therapy; (10) a reduction or elimination in the cancer cell population, and/or a cell population associated with a genetic disease or disorder; (11) a reduction in the growth of a tumor or neoplasm; (12) a decrease in tumor size; (13) a reduction in the formation of a tumor; (14) eradication, removal, or control of primary, regional and/or metastatic cancer; (15) a decrease in the number or size of metastases; (16) a reduction in mortality; (17) an increase in cancer-free survival rate of a subject; (18) an increase in relapse-free survival; (19) an increase in the number of subjects in remission; (20) a decrease in hospitalization rate; (21) the size of the tumor is maintained and does not increase in size or increases the size of the tumor by less 5% or 10% after administration of a therapy as measured by conventional methods available to one of skill in the art, e.g., X-rays, MRI, CAT scan, ultrasound etc; (22) the prevention of the development or onset of cancer and/or a genetic disease or disorder, and/or a symptom associated therewith; (23) an increase in the length of remission for a subject; (24) the reduction in the number of symptoms associated with cancer and/or a genetic disease or disorder; (25) an increase in symptom-free survival of a cancer subject and/or a subject associated with a genetic disease or disorder; and/or (26) limitation of or reduction in metastasis. In some embodiments, the treatment/therapy that a subject receives does not cure cancer, but prevents the progression or worsening of the disease. In certain embodiments, the treatment/therapy that a subject receives does not prevent the onset/development of cancer, but may prevent the onset of cancer symptoms.

In some embodiments, '“preventing” an onset or progression of cancer in a subject in need thereof, or “preventing” an onset or progression of a genetic disease or disorder associated with immune dysregulation in a subject in need thereof, is inhibiting or blocking the cancer or genetic disease or disorder. In some embodiments, the methods disclosed herein prevent, or inhibit, the cancer or genetic disease or disorder at any amount or level. In some embodiments, the methods disclosed herein prevent or inhibit the cancer or genetic disease or disorder by at least or about a 10% inhibition (e.g., at least or about a 20% inhibition, at least or about a 30% inhibition, at least or about a 40% inhibition, at least or about a 50% inhibition, at least or about a 60% inhibition, at least or about a 70% inhibition, at least or about a 80% inhibition, at least or about a 90% inhibition, at least or about a 95% inhibition, at least or about a 98% inhibition, or at least or about a 100% inhibition).

In some embodiments the methods disclosed herein enable the prevention, stratification, or prognosis of side-effects for cancer immunotherapy.

In some embodiments, the method further comprises obtaining a biological sample from the patient, sending the biological sample to a diagnostic laboratory', and receiving from the diagnostic laboratory' a report that states whether the patient is a likely responder to an immunotherapy.

In some embodiments, the biological sample is selected from blood, plasma, serum, mucus, stool, sputum, saliva, nasal secretion, lavage fluid, respiratory fluid, and combinations thereof. In some embodiments, the biological sample is selected from a tumor biopsy, tumor resection, frozen tumor tissue specimen, lymph node, bone marrow, circulating tumor cells, cultured cells, a formalin-fixed paraffin embedded tumor tissue specimen, bronchoalveolar lavage, skin, hair, urine, and combinations thereof. In some embodiments, the tumor biopsy is selected from a core biopsy, needle biopsy, surgical biopsy, and an excisional biopsy.

In some embodiments, the assaying comprises one or more of polymerase chain reaction (PCR) amplification reaction, reverse-transcriptase PCR analysis, quantitative realtime PCR, single-strand conformation polymorphism analysis (SSCP), mismatch cleavage detection, heteroduplex analysis, deoxyribonucleic (DNA) acid sequencing, ribonucleic acid (RNA) sequencing, Northern blot analysis, in situ hybridization, array analysis, restriction fragment length polymorphism analysis, and combinations thereof. In some embodiments, the method comprises detection with one or more next generation sequencing (NGS) methods. In some embodiments, the method comprises whole genome sequencing or panel sequencing, including whole exome sequencing enriched with 3’UTR regions, or ribonucleic acid (RNA) sequencing.

In various embodiments, biological sample refers to a sample obtained or derived from a source of interest (e.g., a cell, tissue, blood, bone marrow, or other biological source), as described herein. In certain embodiments, a source of interest comprises an organism, such as an animal or human. In certain embodiments, a biological sample is a biological tissue or fluid. Non-limiting examples of biological samples include bone marrow, blood, blood cells, ascites, (tissue or fine needle) biopsy samples, cell-containing body fluids, free floating nucleic acids, sputum, saliva, urine, cerebrospinal fluid, peritoneal fluid, pleural fluid, feces, lymph, gynecological fluids, swabs (e.g., skin swabs, vaginal swabs, oral swabs, and nasal swabs), washings or lavages such as a ductal lavages or bronchioalveolar lavages, aspirates, scrapings, specimens (e.g., bone marrow specimens, tissue biopsy specimens, and surgical specimens), feces, other body fluids, secretions, and/or excretions, and cells therefrom, etc.

In some embodiments, the treatment with the immunotherapy comprises reducing or eliminating immune evasion.

In some embodiments, the treatment with the immunotherapy prevents an onset, or progression, of cancer, and is assessed using the overall stage grouping as a non-limiting example: Stage I cancers are localized to one part of the body, typically in a small area; Stage

II cancers are locally advanced and have grow n into nearby tissues or lymph nodes, as are Stage

III cancers. Whether a cancer is designated as Stage II or Stage III can depend on the specific ty pe of cancer. The specific criteria for Stages II and III can differ according to diagnosis. Stage IV cancers have often metastasized, or spread to other organs or throughout the body. The onset or progression of cancer can be assessed using conventional methods available to one of skill in the art, such as a physical exam, blood tests, and imaging scans (e.g., X-rays, MRI. CT scans, ultrasound etc.), optionally in combination with the present methods.

In some embodiments, the cancer is selected from one or more of basal cell carcinoma, biliary tract cancer; bladder cancer; bone cancer; brain and central nervous system cancer; breast cancer; cancer of the peritoneum; cervical cancer; choriocarcinoma; colon and rectum cancer; connective tissue cancer; cancer of the digestive system; endometrial cancer; esophageal cancer; eye cancer; cancer of the head and neck; gastric cancer (including gastrointestinal cancer); glioblastoma; hepatic carcinoma; hepatoma; intra-epithelial neoplasm; kidney or renal cancer; larynx cancer; leukemia; liver cancer; lung cancer (e.g., small-cell lung cancer, non-small cell lung cancer, adenocarcinoma of the lung, and squamous carcinoma of the lung); melanoma; myeloma; neuroblastoma; oral cavity cancer (lip, tongue, mouth, and pharynx); ovarian cancer; pancreatic cancer; prostate cancer; retinoblastoma; rhabdomyosarcoma; rectal cancer; cancer of the respiratory system; salivary gland carcinoma; sarcoma; skin cancer; squamous cell cancer; stomach cancer; testicular cancer; thyroid cancer; uterine or endometrial cancer; cancer of the urinary' system; vulval cancer; lymphoma including Hodgkin's and non-Hodgkin's lymphoma, as well as B-cell lymphoma (including low grade/follicular non-Hodgkin’s lymphoma (NHL); small lymphocytic (SL) NHL; intermediate grade/follicular NHL; intermediate grade diffuse NHL; high grade immunoblastic NHL; high grade lymphoblastic NHL; high grade small non-cleaved cell NHL; bulky disease NHL; mantle cell lymphoma; AIDS-related lymphoma; and Waldenstrom's Macroglobulinemia; chronic lymphocytic leukemia (CLL); acute lymphoblastic leukemia (ALL); Hairy cell leukemia; chronic myeloblastic leukemia; as well as other carcinomas and sarcomas; and post-transplant lymphoproliferative disorder (PTLD), as well as abnormal vascular proliferation associated with phakomatoses, edema (e.g. that associated with brain tumors), and Meigs’ syndrome. In some embodiments, the cancer is melanoma. In some embodiments, the cancer is a gastric cancer.

In some embodiments, the method identifies a patient for immunotherapy treatment who would not otherwise be identified for immunotherapy treatment using a IHC test for expression of a checkpoint biomarker, optionally selected from PD-1, PD-L1, PD-L2, CTLA- 4, LAG-3, TIM-3, B7-H3, B7-H4, ICOS, SIGLEC7, SIGLEC9, and VISTA.

In aspects, the present disclosure provides a method of identifying an immune cell gene signature profile which is correlative with responsiveness or non-responsiveness to an immunotherapy, the method comprising one or more steps of FIG. 1 A and/or FIG. 4A.

In aspects, the present disclosure provides a companion diagnostic, complementary diagnostic, or codiagnostic test kit, comprising: (a) an array of nucleic acids suitable for detection of one or more non-coding somatic and/or germline mutations in 3' UTR regions of a patient’s genome, the one or more non-coding somatic and/or germline mutations in 3' UTR regions being indicative of a patient response or non-response to an immunotherapy; and (b) instructions for use.

In aspects, the present disclosure provides a companion diagnostic, complementary diagnostic, or codiagnostic test kit, comprising reagents and instructions for use in one or more of any of the preceding aspects and/or embodiments.

In various embodiments, the “subject” refers to any animal (e.g., a mammal), including, but not limited to, humans, and non-human animals (including, but not limited to, non-human primates, dogs, cats, rodents, horses, cows, pigs, mice, rats, hamsters, rabbits, and the like (e.g., which is to be the recipient of a particular treatment, or from whom cells are harvested)). In some embodiments, the subject is a human.

It will also be understood that, although the terms first, second, etc., may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first subject could be termed a second subject, and. similarly, a second subject could be termed a first subject, without departing from the scope of the present disclosure. The first subject and the second subject are both subjects, but they are not the same subject. Furthermore, the terms “subject,” “user,” and “patient” are used interchangeably herein.

As used herein, the word “include,” and its variants, is intended to be non-limiting, such that recitation of items in a list is not to the exclusion of other like items that may also be useful in the materials, compositions, devices, and methods of this technology. Similarly, the terms “can” and “may” and their variants are intended to be non-limiting, such that recitation that an embodiment can or may comprise certain elements or features does not exclude other embodiments of the present technology' that do not contain those elements or features. Although the open-ended term “comprising,” as a synonym of terms such as including, containing, or having, is used herein to describe and claim the disclosure, the present technology, or embodiments thereof, may alternatively be described using more limiting terms such as “consisting of’ or “consisting essentially of’ the recited ingredients.

Unless defined otherwise, all technical and scientific terms herein have the same meaning as commonly understood by one of ordinary' skill in the art to which this disclosure belongs. Although any methods and materials, similar or equivalent to those described herein, can be used in the practice or testing of the present disclosure, the preferred methods and materials are described herein. All publications, patents, and patent publications cited are incorporated by reference herein in their entirety’ for all purposes.

This disclosure is further illustrated by the following non-limiting examples.

EXAMPLES

Example 1: Characterization of the somatic and germline 3 ' UTR variant landscape in gastric adenocarcinoma

In the experiments of this example, next generation sequencing and specifically RNAseq data were utilized to call variants. Sequencing data from 375 gastric cancer patients (Table 2, below), including 375 primary gastric cancer samples and 40 matched controls, were analyzed following a comprehensive approach using GATK best practices. These experiments led to the identification of a high number of expressed variants and indels per sample (FIG. 8A and FIG. 8B). Analysis of the distribution of called variants along the length of the 3' UTR revealed that the RNAseq-derived calls matched the distribution of WGS calls from PCAWG (FIG. IB). Analysis of the distribution of TCGA WES-derived variants along the length of the 3 ' UTR showed that most variants fall in the beginning of the 3 ' UTR, proximal to coding sequences, likely representing sequences captured by coding region probes (FIG. IB).

Variants detected in the gnomAD database, or present in at least 2 out of the 40 TCGA gastric normal controls, were defined as germline. Out of 5,431,118 variants identified across the genome by the RNAseq GATK analysis, 42.9% overlapped with the gnomAD database, while atotal of 73.4% were assigned a “likely germline" categorization. The rest (26.6%) were treated as “likely somatic”. Samples deemed as ultramutated by TCGA, based on WES-derived variant calls, exhibited high frequency of RNAseq-derived somatic SNVs (FIG. 1C). All TCGA STAD RNAseq samples were also analyzed using Strelka2, an orthogonal variant calling algorithm, and identified a -90% concordance (FIG. ID).

Example 2: High-throughput capture of known functioned variants in PD-L1 3 ' UTR and validation

In the experiments of this example, the high-throughput approach was evaluated to determine whether the high-throughput approach could capture the functional impact of the small number of 3’ UTR SNVs that have been previously associated with changes in PD-L1 expression in gastric cancer, namely rs2297136 and rs4143815, or non-small cell lung cancer, such as rs 4742098. Indeed, the analysis identified all three variants and when comparing PD- LI expression in samples with or without the 3‘ UTR variants, a significant increase in expression levels in patients carrying the alternative allele was observed (FIG. 2A).

To validate the potential function of these alleles, luciferase assays were performed with reference (GRCh38 genome assembly) or variant allelic PD-L1 3’ UTR reporters in three gastric cancer cell lines and HEK293T cells. In all cases except one. the variant allele was associated with higher luciferase (ZMC) expression, with the most consistent increase in PD-L1 3’ UTR reporter activity observed with the rs4143815 allele (FIG. 2B).

Example 3: Prioritization of 3 ’ UTR germline variants and somatic mutations controlling, cis gene expression in gastric adenocarcinoma

In the experiments of this example, transcriptome-wide 3' UTR variants associated with czs-gene expression changes in gastric cancer were prioritized. Variants that were present in 5 or more samples were investigated, corresponding to 1.3% or higher minor allele frequency (MAF) in the tested population. An eQTL analysis was performed with the dosage of each variant as the genotype variable, and the inverse quantile-normalized expression of the corresponding gene as the phenotype variable. To remove unwanted variation from the model, sex and age were used along with the top 5 genetic principal components (PCs) and expression surrogate variables (SVs) (76) as covariates, as shown in the following model used for eQTL analysis: gene expression ~ sex + age + genetic PCs (1 - 5) + SVs (1 - 5)

With a cutoff of a nominal p-value of le-7, -3,000 cA-eQTLs were identified in proteincoding genes, accounting for 75% of all eQTLs (FIG. 3A). Interestingly, the highest percentage (60%) of significant c/.s -eQTLs in protein-coding genes were found to reside in the 3’ UTR genic region (FIG. 3B), reflecting the importance of 3' UTR civ-acting elements in controlling gene expression, often post-transcriptionally. Comparing the number of 3’ UTR variants to the average relative length of the 3’ UTR in protein coding genes revealed a significant enrichment (FIG. 9) (chi-square p-value < le-5).

Overlapping the 3’ UTR cA-eQTLs with databases of predicted and experimentally supported miRNA and RBP binding sites showed that around 90% of the variants reside in functionally relevant regulatory' elements (FIG. 3C). Finally, gene-set enrichment analysis (GSEA) of the top significant 3‘ UTR c/.s-eQTLs revealed enrichment in immune-related pathways (FIG. 3D), indicating that 3' UTR variants could have an impact on immune phenotypes in gastric cancer. Significant 3’ UTR cis -eQTLs were also identified in important gastric cancer oncogenes, including KRAS, PDGFRA, and CCDN1 among others.

Example 4: Massive parallel reporter assay validation of cis-eQTLs residing, on cancer immunoediting genes

Considering the significant enrichment of immune-related pathways in the top 3’ UTR c/.s'-eQTLs, as well as the importance of immune escape in cancer progression, and as a target for novel therapeutics, the experiments of this example proceeded with a functional validation of prioritized 3 ’ UTR eQTLs on a compiled list of immune-related genes (Table 3, below). The manually curated list incorporates immune checkpoint and known or suspected immunomodulatory genes, MHC machinery, genes used in signatures for response to immunotherapy, and significant hits from hypothesis-free CRISPR-Cas9 screens for CD8+ T- cell effector function and in vivo screening of transplantable tumors in mice treated with ICI.

The top 750 variants were selected that resided in 299 prioritized genes based on the curated list described above (Table 4, below). A massive parallel reporter assay (MPRA) was developed and performed in two gastric cancer cell lines to assess the effect of each variant on the post-transcriptional stability of a reporter gene (FIG. 4A). Briefly, a reporter plasmid library containing barcoded reference and alternative alleles for the 750 eQTL variants was transfected into AGS and SNU719 cells. The effect of the variant on post-transcriptional expression of the reporter was assessed by barcode quantification from amplicon sequencing of RNA extracted from transfected cells.

The two cell lines used in the MPRA assay yielded similar outcomes (FIG. 10A and FIG. 10B). Approximately 15% of eQTLs (128 variants) showed a significant causative effect on the expression of the reporter (FDR-adjusted p-value < 0.05) in at least one of the two cell lines (Table 5, below). A subset of genes with causal regulatory variants (FIG. 4B, Table 5) are involved in antigen processing and presentation (e.g., HLA genes, CTSB, CTSS, LGMN, GUTA, TAPBP) as well as RNA-editing enzymes such as ADAR.

Example 5: Uncovering transcriptome-wide 3 'UTR variants regulating the gastric adenocarcinoma immune landscape 3' UTR regions not only regulate gene expression, but can also influence mRNA localization, protein-protein interactions and other post-transcriptional, translational and post- translational functions. Therefore, 3’ UTR variants can affect immune phenotypes in cancer gene expression ~ sex + age + genetic PCs (1 - 5) + S Vs (1 - 5) independently of their effect on expression. To unbiasedly associate 3’ UTR variants with changes in the immune landscape of gastric cancer, the experiments of this example performed an immune landscape (il)QTL analysis using a similar model as above, focusing on immune phenotypes instead of czs-gene expression as the dependent variable. Immune phenotypes for the TCGA STAD cohort were obtained from the Cancer Research Institute (CRI) iAtlas project and included expression-based immune cell infiltration estimates. TCR/BCR entropy and leukocyte ratio were calculated by combined imaging, methylation, and expression-based analyses.

For almost all immune features, the majority⁷ of ilQTLs resided in the 3’ UTR region of protein-coding genes, and a large percentage of those are predicted to reside in regions of miRNA/RBP binding, similarly to eQTLs (FIG. 5). Significant ilQTLs were identified in immune-relevant genes, including B2M HLA genes, CANX, LDHA, PSMB2, HNRNPR, and ADAR, which are known to affect the tumor immune landscape. The top hits also included WARS' and APOBEC3C, showing the potential of this approach for prioritization of novel cancer-specific immunotherapeutic targets.

The experiments subsequently focused on CD8+ T cell ilQTLs since the level of CD8+ T cell infiltration in a tumor is an important determinant of cancer immunotherapy response. A significant CD8+ T cell fraction of QTL variants in 467 genes was identified. GSEA analysis showed that the most enriched "cellular component” gene sets. In the topmost significant CD8+ T cell infiltration QTL variants are the ribosome and ribosomal subunits (FIG. 11). In addition, immunoregulatoiy CRISPR hits (Table 3, below) w ere enriched in significant CD8+ T cell ilQTL 3‘ UTR variants compared to all 3‘ UTR variants (Fisher’s exact test p- value=0.0057). showing concordance between the two orthogonal approaches of key gene prioritization. In addition, 5% of those ilQTLs overlapped with eQTLs, showing the ability of ilQTLs to capture associations beyond gene expression regulation.

Example 6: Novel functional 3 ' UTR cis-eOTL and ilQTL variants in ADAR

Significant functional variants were identified in all of the high throughput investigations (eQTLs, ilQTLs, MPRA validated variants) in the Adenosine Deaminase RNA Specific (ADAR) gene. ADAR encodes an enzyme that catalyzes A-to-I editing in RNA and has been implicated in promoting cancer hallmarks in multiple cancer types, including breast, thyroid and gastric malignancies. In some CRISPR screens for genes that sensitize tumors to immunotherapy, ADAR is the 4th most enriched out of -20,000 genes, and ADAR loss has been shown to lead to an increase in tumor inflammation and elevated sensitivity to PD-1 blockade therapy, highlighting the potential of ADAR as a cancer immunotherapy target. ADAR downregulation has also been shown to induce inflammatory signaling in gastric cancer specifically. To investigate the immunoediting role of ADAR in gastric cancer further, the experiments of this example queried the iAtlas portal, where CNVs on ADAR were reported to exhibit high effect sizes on Leukocyte Fraction (Amp: p=10-4 to 10-10 (multiple groups)), Lymphocyte Infiltration Score, (Amp: p=10-6 / Del: p=10-5), and CD8+ T Cell content (Del: p=10-4). Moreover, ADAR appeared as a hotspot for 3’ UTR eQTL and ilQTL variants in gastric cancer (FIG. 6A), while ADAR was found to significantly overexpressed in responders (n=55) compared to the non-responders (n=80) to immune-checkpoint inhibitors, in a cohort consisting of gastric cancer and melanoma patients (FIG. 6B).

Through the QTL and MPRA analysis, the experiments in this example identified a novel somatic 3’ UTR cis-eQTL variant (chrl : 154583325, T-to-C) in ADAR (FIG. 6A) with causal effects on post-transcriptional regulation (FIG. 12A and FIG. 12B). The same variant was also found as an ilQTL for multiple immune features, including CD8+ T cell infiltration and TCR diversity (FIG. 6A). Based on a meta-analysis of CLIP data from the POSTAR2 project, the ADAR variant is predicted to overlap with multiple RBP binding sites (Table 6, below ). One of those RBPs, TARDBP, has been shown to directly regulate ADAR1 expression in liver cancer and leukemia cell line models. Indeed, correlation analysis in gastric cancer patients from TCGA revealed a strong association in the expression of the two proteins (FIG. 6C), suggesting that the regulation axis between TARDBP and ADAR could be functional in gastric cancer as well.

Example 7 3 ’ UTR variants as an effective means for immunotherapy response prediction

The experiments in this example investigated the translational potential of 3' UTR ilQTL variants and their ability to predict therapeutic outcomes to cancer immunotherapy. To this end, the experiments analyzed the cohort of responders (R) vs non-responders (NR) to immune checkpoint inhibitors described above, where tumors were subjected to whole transcriptome sequencing thus enabling the detection of 3‘ UTR variants. Significant ilQTL variants in those samples were identified in these experiments. Samples were separated into a training (n=68, % Responders) and a test (n=67, % Responders) set, and 28 3' UTR ilQTL variants were selected and enriched in the R vs NR samples of the training set (Fisher’s exact test p-value < 0.05, Table 7). The variants were used to devise a polygenic risk score for the potential prediction of response to immune checkpoint inhibition. When tested on the orthogonal test set (n=67), the polygenic risk score was significantly increased in the responders (T-test p-value = 0.0019, FIG. 7A), and exhibited a higher area under the receiver operating characteristic curve (AUC, ROC) than PD-L1 expression (FIG. 7B). The experiments of this example demonstrated the first time that noncoding variants have been used to predict immunotherapy treatment outcomes in cancer.

Methods

Cell Culture

The AGS (ATCC CRL-1739) and HEK293T (ATCC CRL-3216) cell lines were purchased from ATCC, the SNU-719 (KCLB-00719) cell line was purchased from the Korean Cell Line Bank (KCLB), and YCCEL1 was a gift from Erik Flemington (Tulane School of Medicine). AGS, SNU719 and YCCEL1 were maintained in RPMI-1640 (Gibco, Gaithersburg, MD, USA) with 10% FBS (Invitrogen, Carlsbad, CA, USA). HEK293T cells were maintained in DMEM (Gibco) with 10% FBS (Invitrogen). All cell lines were incubated at 37°C and 5% CO2.

Luciferase assays

The wild-ty pe (WT) PD-L1 3‘ UTR psicheck2 reporter was cloned as described previously. Targeted mutagenesis for the generation of single-mutant PD-L1 3’ UTR reporters was also performed as described previously. Cells were grown to 80% confluence in 6-well plates and were transfected with 1 pg psicheck2 vector per well using Lipofectamine 3000 (ThermoFisher Scientific, Waltham, MA, USA), as per the manufacturer’s protocol. Fresh media was added to the transfected wells 24 hr later. Cells were harvested 48 hr posttransfection and the Dual -Luciferase Reporter Assay System (Promega, Madison, WI, USA) kit was used to measure Firefly and Renilla luciferase activity levels on the GloMax explorer (Promega).

RNAseq Variant Calling

Raw RNAseq data for 375 TCGA STAD primary cancer and 40 matched normal samples were obtained from Genomic Data Commons (GDC) following NIH dbGAP approval. Reads were mapped against the human genome (hg38) using STAR. Mapped reads were deduplicated and short variants/indels were called using Mutect2 following GATK best practices. The Mutect2 output was converted to gVCF by using the region coverage statistics. Since Mutect2 cannot perform genotype calling and does not distinguish between homozygous reference and no-call regions, HaplotypeCaller was also run in parallel by following GATK best practices for RNA. For samples lacking a Mutect2 call at a specific variant position, HaplotypeCaller was used to distinguish whether the lack of a Mutect2 call was because of no coverage in that region or a homozygous reference genotype. A mutation was characterized as likely somatic by calculating the posterior probability of the event, while using variant call statistics, clonality in tumor samples, matched healthy tissues and gnomAD variants as priors. Calls were filtered using the FilterMutect2 tags “base quaK, “map qual”, "n ratio” and “slippage”. Only biallelic variants present in at least 5 out of the 375 (>1.3%) samples were pursued further. Strelka2 was also run following the same preprocessing steps as for GATK callers. eQTL analysis

Gene-level expression in TCGA STAD samples was calculated using Salmon v0.91 and Ensembl genome annotation v77. A linear model was utilized to call eQTLs w ith FastQTL. In brief, gene expression across libraries was normalized using trimmed mean of m-values as implemented in edgeR. Genes w ere selected based on an expression threshold of 1 read in at least 80% of the samples. An inverse quantile normal transformation was performed on the expression values prior to their inclusion into the linear model. Mutect2/HaplotypeCaller alternative allele dosage w as utilized as genotype input, while age at diagnosis, sex, the top 5 genetic principal components (PCs) and expression surrogate variables (SVs) w ere covariates. Genetic PCs were calculated using SmartPCA on WES-derived germline variants from the same TCGA STAD patients. Using the Bonferroni correction method, a nominal p-value cutoff of le-7 was used to identify significant calls. eQTL calls w ere mapped to transcript annotations (Gencode v32), while relative genomic locations (5’ UTR, CDS, 3’ UTR) w ere assigned using annotatr. Only eQTLs with significant cis effects were retained for further analysis. Variant annotation for potential overlap with post-transcriptional regulatory regions was performed using the GenomicRanges package in R (v 1.38.0). Experimentally-supported miRNA binding site coordinates were obtained from TarBase, predicted miRNA binding sites w ere acquired from microT-CDS, and CLIP-based predictions of RBP binding sites were obtained from the POSTAR2 database. ilQTL analysis

Immune data per sample were obtained from the Cancer Research Institute (CRI) i Atlas project. QTL analysis was performed with FastQTL, using the same genotype and covariate data as above, while using quantile-normalized immune profile estimates as phenotypes. ilQTL selection as well as genomic and regulatory annotation were performed as for eQTLs.

Pathway Enrichment and Over-representation analysis

The top 500 QTL genes, ranked based on QTL p-value. were investigated by pathway enrichment analysis using ClusterProfiler (v3.12.0). Pathway information was obtained from the Gene Ontology' Resource, and the Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway database. Plots w ere generated in R using ggplot2 (v3.3.3).

Cancer Immunology-related genes

A collection of more than 2,500 immune-related genes was manually curated from the literature and experimental resources. Specifically, the list includes immune checkpoint and immunomodulatory genes, genes involved in the MHC machinery and microsatellite instability, cytokines and chemokines, gene markers for metabolic reprogramming, and oncogenes or complexes that can affect the tumor transcriptional and immune landscape, such as EZH2-PRC2 chromatin remodeling complex members and BAF/PBAF complex members. A number of significant genes from hypothesis-free CRISPR-Cas9 screens w ere incorporated for CD8+ T-Cell effector function, and in vivo screening of transplantable tumors in mice treated with immunotherapy. In addition, the Urea cycle (GQ:0000050) and Mismatch repair (G0:0006298) Gene Ontology' terms w ere included, a list of Human DNA repair genes, and the following entries from the Kyoto Encyclopedia of Genes and Genomes (KEGG): MAPK signaling pathway (hsa04010), PI3K-Akt signaling pathway (hsa04151), Wnt signaling pathway (hsa04310), JAK-STAT signaling pathway (hsa04630), Antigen processing and presentation (hsa04612). Finally, the list comprises more than 250 genes from signatures associated with response to immune checkpoint inhibition.

MPRA pool design

A pool of 19.220 150bp-long oligonucleotides (oligos) was synthesized commercially (Twist Biosciences, San Francisco, CA, USA). Each oligo contained the reference or alternative allele of the eQTL variant in the middle flanked by 50bp of reference transcriptomic sequence on either side. In the case of eQTLs residing in alternative transcript isoforms, oligos were synthesized for all possible transcripts. For eQTLs that were close to the end of the transcript a random sequence was added to bring up the length of the sequence to lOlbp. The random sequence was the same in all ohgos and was selected to lack any predicted 7 or 8bp- long RBP and miRNA-seed binding sites. An 8 bp-long barcode was added at the 3' end of the 101 bp-long sequence. Each allele was represented by 10 unique barcodes. All 8 bp barcodes that matched RBP or miRNA seed binding sites were removed. Finally, 20 bp sequences were added on either side of the oligo that matched the 5' and 3'-end sequences of A/?oI-/A'o/I-digested psicheck2 to allow cloning with the NEBuilder HiFi DNA Assembly kit (NEB, Ipswich, MA, USA). The pool also contained scrambled sequences with variants introduced in the middle as negative controls.

Pool amplification and cloning

Amplification of the pool, prior to cloning, was performed using 0.5 pM of each of the PCR lib fwd and PCR lib rev primer pair (Table 1, below) with the NEB Next High-Fidelity 2x PCR Master Mix (NEB, M0541L).

The following PCR conditions were used: 98 °C for 30 sec, 20 cycles (98 °C for 10 sec, 63 °C for 10 sec, 72 °C for 15 sec), 72 °C for 2 min. The amplified oligo pool was introduced into a V/iM-ZAG/l-digested psicheck2 vector using NEBuilder HiFi DNA Assembly kit (NEB, E2621S). as per the manufacturer’s protocol. The assembly reaction product was purified following a standard isopropanol precipitation protocol. The purified plasmid pool was transformed into Endura ElectroCompetent cells (Lucigen, Middleton, WI, USA) at 50 ng plasmid per 25 pl of bacteria ratio, following the provider’s protocol. A total of 8 transformation reactions were pooled together and plated onto large 15 cm LB Agar plates at 37 °C for 12 h. A large enough number of colonies to ensure at least 500 colonies/oligo representation was harvested directly from the LB Agar plates, and the plasmid pool was purified by performing at least 2 midipreps per 15 cm LB Agar plate, using the Qiagen Plasmid Plus Midi kit (Qiagen, #12943).

MPRA transfection and library prep

SNU719 and AGS cells were seeded in 15 cm plates to achieve 80% confluence the next day. Cells were transfected with 10 pg of the MPRA plasmid library using TransIT-X2 reagent (Mirus Bio, Madison, WI, USA) as per the manufacturer’s protocol, aiming for a transfection efficiency of 50-80%. Total RNA was collected 48 hr post-transfection using the miRNeasy mini kit (Qiagen, #217004). Genomic DNA was removed using the Turbo DNA- free kit (ThermoFisher Scientific) following the manufacturer’s ‘'Rigorous DNase treatment” protocol. Per replicate, 15 pg total RNA was reverse transcribed with SuperScript IV Reverse Transcriptase (ThermoFisher Scientific) using oligo-dT primers. Amplicon sequencing libraries from cDNA or plasmid pool DNA were constructed through two PCR reactions. In the first PCR round, 1: 10 diluted cDNA was amplified using 0.2 pM of the DT_barcodePE_Fv2 and 0.2 pM of an isomolar mix of the DT_barcodePE_Rv2 primers (0 to 6 random Ns, see Table 1). The PCR reaction was performed with the NEB Next High-Fidelity 2x PCR Master Mix (NEB. M0541L), and the following conditions: 98 °C for 30 sec, 10 cycles (98 °C for 10 sec, 63 °C for 10 sec, 72 °C for 15 sec), 72 °C for 2 min. Enough PCR reactions were run to ensure that all the cDNA from each replicate was amplified. In the second PCR reaction, 1 : 10 diluted PCR round 1 product was amplified using 0.5 pM of a unique pair of multiplexing Illumina primers (PE_i5 and PE index, see Table 1). The following PCR conditions were used: 98 °C for 30 sec, 10 cycles (98 °C for 10 sec, 62 °C for 10 sec, 72 °C for 15 sec), 72 °C for 2 min. For each replicate, the second round PCR product was purified through gel extraction using the Monarch Gel Extraction kit (NEB). The quality of each library was assessed by an Agilent Tapestation D1000 assay (Agilent, Santa Clara, CA, USA). An equimolar mix of all libraries was sent for single-end 150 bp sequencing on an Illumina sequencer, with 20% PhiX spike-in to increase library complexity. The mixed library was sequenced at a depth to ensure at least 10 M reads per replicate (>500 reads per oligo).

MPRA analysis

MPRA analysis was performed. Briefly, barcode counts were calculated from raw reads and then normalized per sample based on sequencing depth. For each sample, a barcode RNA to DNA ratio was calculated by dividing the barcode counts in each replicate to that in the plasmid pool library. The RNA to DNA ratios were then log-transformed and quantile normalized across samples. A two-sided Wilcoxon test was performed to compare barcode count ratios between reference and alternative allele oligos in each replicate. To combine replicate p-values, the Stouffer’s method was used, and false discovery rate (FDR) correction was applied. Each oligo was represented by 10 barcodes, so to obtain a per-oligo activity in each sample, the median activity was calculated. Fold-change was defined as the ratio of the alternative to the reference allele median activity.

Immune Checkpoint Inhibition cohort analysis

Pre-treatment tumor RNAseq data were retrieved from four Immune Checkpoint Inhibition (ICI) studies (anti-PDl or anti-CTLA4 treatment), of which three addressed melanoma patients (n=90), and one addressed gastric cancer patients (n=45). The combined cohort (n=135) included 55 responders and 80 non-responders to immunotherapy. Gene expression of pre-treatment tumor samples was quantified from RNAseq reads using Salmon v.0.91. To calculate differential expression of ADAR between responders and non-responders, a wilcoxon rank sum test with continuity correction was performed. The ICI cohort was also randomly split into a training (n=68), and a testing set (n=67), and the Variant Calling and ilQTL pipelines were run on the pre-treatment tumor RNAseq data. The 3 ' UTR ilQTL variants enriched in ICI responders in the training set (n=28, p- value < 0.05) were selected to comprise the Polygenic Risk Score. The score is calculated as the number of variants detected in the patient’s tumor sample, therefore ranging from 0 to 28. Table 2

Table 3: Genes prioritized for eQTL evaluation. Checkpoint and other immune-related cgenes were selected (cytokines, chemokines, antigen processing and presentation genes, etc), as well cancer metabolism genes, genes responsible for important pathways in cancer (JAK/STAT, AKT/MTOR, MAPK, etc), and genes prioritized from studies related to response to cancer immunotherapy or CRISPR/Cas9 screens for T Cell killing.

Table 4: 3’UTR eQTLs identified in immune-related genes

Table 5: MPRA results for 2 cell lines (SNU719, AGS)

Table 6

Table 7

Claims

CLAIMS What is claimed is:

1. A method of selecting a patient for treatment with an immunotherapy, comprising:

(a) obtaining a biological sample from the patient;

(b) assaying the biological sample for one or more non-coding somatic and/or germline mutations in 3 ' untranslated region (3 ' UTR) regions of the patient’s genome;

(c) classifying the patient as a likely responder or a likely non-responder to the immunotherapy based on the presence, absence or level of the one or more noncoding somatic and/or germline mutations in the 3 ' UTR regions; and

(d) selecting the patient for treatment with the immunotherapy if the patient is classified as a likely responder.

2. A method of predicting a patient response to treatment with an immunotherapy, comprising:

(a) obtaining a biological sample from the patient;

(b) assaying the biological sample for one or more non-coding somatic and/or germline mutations in 3 ' untranslated region (3' UTR) regions of the patient’s genome;

(c) classifying the patient as a likely responder or a likely non-responder to the immunotherapy based on the presence, absence or level of the one or more noncoding somatic and/or germline mutations in the 3 ' UTR regions.

3. A method of preventing or treating cancer in a patient, comprising:

(a) selecting the patient for treatment with an immunotherapy, the selection being based on classifying a patient as suitable for treatment with an immunotherapy based on assaying a biological sample from the patient for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; and

(b) administering an effective amount of the immunotherapy to the patient. A method of modulating an immune landscape phenotype in a patient, comprising:

(a) selecting the patient for treatment with an immunotherapy, the selection being based on classifying a patient as suitable for treatment with an immunotherapy based on assaying a biological sample from the patient for one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome; and

(b) administering an effective amount of the immunotherapy to the patient. The method of claim 4, wherein the modulation of the immune landscape phenotype comprises one or more of modulating immune cell infiltration, modulating T cell and B cell receptor diversity, and modulating antigen presentation or processing. The method of any one of claims 1-5, wherein the patient is afflicted with a cancer. The method of any one of claims 3-6, further comprising: obtaining a biological sample from the patient, sending the biological sample to a diagnostic laboratory, and receiving from the diagnostic laboratory a report that states whether the patient is a likely responder to an immunotherapy. The method of any one of claims 1-7, wherein the biological sample is selected from blood, plasma, serum, mucus, stool, sputum, saliva, nasal secretion, lavage fluid, respiratory fluid, and combinations thereof. The method of any one of claims 1-8, wherein the biological sample is selected from a tumor biopsy, tumor resection, frozen tumor tissue specimen, lymph node, bone marrow, circulating tumor cells, cultured cells, a formalin-fixed paraffin embedded tumor tissue specimen, bronchoalveolar lavage, skin, hair, urine, and combinations thereof. The method of claim 9, wherein the tumor biopsy is selected from a core biopsy, needle biopsy, surgical biopsy, and an excisional biopsy. The method of any one of claims 1 -10, wherein the assaying comprises one or more of polymerase chain reaction (PCR) amplification reaction, reverse-transcriptase PCR analysis, quantitative real-time PCR, single-strand conformation polymorphism analysis (SSCP), mismatch cleavage detection, heteroduplex analysis, deoxyribonucleic (DNA) acid sequencing, ribonucleic acid (RNA) sequencing, Northern blot analysis, in situ hybridization, array analysis, restriction fragment length polymorphism analysis, and combinations thereof. The method of any one of claims 1-11, wherein the method comprises detection with one or more next generation sequencing (NGS) methods. The method of any one of claims 1-12, wherein the method comprises ribonucleic acid (RNA) sequencing, whole genome sequencing, targeted (probe or amplification-based) genome sequencing, panel sequencing and/or whole exome sequencing. The method of any one of claims 1-13, wherein the mutations comprise one or more of single-nucleotide variants (SNVs) and/or short insertions-deletions (indels), optionally wherein the SNVs and/or indels is located within the 3’ UTR. The method of any one of claims 1-14, wherein the mutations or site of mutations are putative or experimentally-supported miRNA and/or RBP binding sites. The method of any one of claims 1-15, wherein the mutations are selected from one or more of the entries of Table 4, or are located in one or more of the entries of Table 5, or are located in one or more of the entries of Table 7, optionally one or more entries of FIG. 4B. The method of any one of claims 1-16, wherein the mutations are in one or more of B2M, HLA genes, CA NX, LDHA, PSMB2, HNRNPR, WARS, APOBEC3C, STAT1, and ADAR. The method of any one of claims 1-17, wherein the mutation is chrl : 154583325, T-to-C in ADAR. The method of any one of claims 1-18, wherein the classifying comprises one or more of gene expression quantitative trait loci (eQTL) and immune landscape QTL (ilQTL) analysis. The method of any one of claims 1-19, wherein the classifying comprises comparing the results of the assaying to one or more reference datasets that are indicative of patient response or non-response to the immunotherapy. The method of any one of claims 1-20, wherein the classifying comprises combining the results of the assaying into a polygenic risk score, signature, mathematical model, statistical model, or machine learning classifier. The method of any one of claims 1-19, wherein the classifying comprises combining the results of the assaying and one or more reference datasets that are indicative of patient response or non-response to the immunotherapy into a polygenic risk score, signature, mathematical model, statistical model, or machine learning classifier. The method of any one of claims 1-22, wherein the classifying comprises one or more steps of FIG. 1A. The method of any one of claims 1-23, further comprising assaying one or more non-coding mutations in genic loci or intergenic loci of the patient’s genome. The method of claim 24, wherein the genic loci are one or more of a 3’UTR, 5' UTR, intron, and promoters. The method of any one of claims 1-25, further comprising assaying one or more coding mutations of the patient’s genome. The method of any one of claims 1-26, wherein the method comprises merging and/or comparing (i) one or more non-coding somatic and/or germline mutations in 3' untranslated region (3' UTR) regions of the patient’s genome, with (ii) one or more noncoding mutations in genic loci or intergenic loci of the patient’s genome, and/or (iii) one or more coding mutations of the patient’s genome. The method of any one of claims 1-26, wherein the method comprises merging and/or comparing (i) one or more non-coding somatic and/or germline mutations in 3 ' untranslated region (3' UTR) regions of the patient’s genome, with (ii) one or more noncoding mutations in genic loci or intergenic loci of the patient’s genome, and/or (iii) one or more coding mutations of the patient’s genome with (iv) one or more reference datasets that are indicative of patient response or non-response to the immunotherapy. The method of any one of claims 1-28, further comprising assaying whether the biological sample is positive or negative for gene expression of one or more checkpoint marker. The method of claim 29, wherein the checkpoint marker is selected from programmed death 1 (PD-1), programmed death-ligand 1 (PD-L1), programmed death-ligand 2 (PD- L2), Cytotoxic T Lymphocyte Antigen-4 (CTLA-4), Lymphocyte Activation Gene-3 (LAG-3), T cell Immunoglobulin and Mucin protein-3 (TIM-3), B7-H3, B7-H4, inducible T cell co-stimulator (ICOS), Sialic acid-binding immunoglobulin-type lectin 7 (SIGLEC7), Sialic acid-binding immunoglobulin-type lectin 9 (SIGLEC9), and V-domain Ig suppressor of T cell activation (VISTA). The method of any one of claims 1-30, wherein the biological sample has been immunohistochemically stained for the expression of PD-L1, PD-L2 or both PD-L1 and PD-L2. The method of any one of claims 1-31, wherein the immunotherapy is an agent that modulates one or more checkpoint biomarker, selected from PD-1, PD-L1, PD-L2, CTLA- 4, LAG-3, TIM-3, B7-H3, B7-H4, ICOS, SIGLEC7, SIGLEC9, and VISTA. The method of any one of claims 1-32, wherein the immunotherapy is an agent that modulates one or more PD-1, PD-L1, PD-L2, and CTLA-4. The method of claim 33, wherein the agent that modulates:

PD-1 is an antibody or antibody format specific for PD-1;

PD-L1 is an antibody or antibody format specific for PD-L1;

PD-L2 is an antibody or antibody format specific for PD-L2; or CTLA-4 is an antibody or antibody format specific for CTLA-4. The method of claim 34, wherein the antibody or antibody format is selected from one or more of a monoclonal antibody, polyclonal antibody, antibody fragment, Fab, Fab', Fab'- SH, F(ab')2, Fv, single chain Fv, diabody, linear antibody, bispecific antibody, multispecific antibody, chimeric antibody, humanized antibody, human antibody, and a fusion protein comprising the antigen-binding portion of an antibody. The method of claim 34, wherein the antibody or antibody format specific for PD-1 is selected from nivolumab, pembrolizumab, and pidilizumab. The method of claim 34, wherein the antibody or antibody format specific for PD-L1 is selected from atezolizumab, avelumab, durvalumab, and BMS-936559. The method of claim 34, wherein the antibody or antibody format specific for CTLA-4 is selected from ipilimumab and tremelimumab. The method of any one of claims 1-38, wherein the method further comprises administering an agent that modulates one or more PD-1, PD-L1, PD-L2, and CTLA-4. The method of any one of claims 1-39, wherein the immunotherapy is a cellular immunotherapy treatment. The method of claim 40, wherein the administration is sequential or simultaneous. The method of any one of claims 1-41, wherein the administration is by intratumoral, intradermal, subcutaneous, intramuscular, intraperitoneal or intravenous injection, or direct injection into cancer tissue. The method of claim 42, wherein the administration is intratumoral. The method of any one of claims 1-43, wherein the treatment with the immunotherapy comprises reducing or eliminating immune evasion. The method of any one of claims 1-44, wherein the treatment with the immunotherapy comprises modulation to enable successful subsequent or concurrent treatment with immunotherapy. The method of any one of claims 3 and 6-45, wherein the cancer is selected from one or more of basal cell carcinoma, biliary tract cancer; bladder cancer; bone cancer; brain and central nervous system cancer; breast cancer; cancer of the peritoneum; cervical cancer; choriocarcinoma; colon and rectum cancer; connective tissue cancer; cancer of the digestive system; endometrial cancer; esophageal cancer; eye cancer; cancer of the head and neck; gastric cancer (including gastrointestinal cancer); glioblastoma; hepatic carcinoma; hepatoma; intra-epithelial neoplasm; kidney or renal cancer; larynx cancer; leukemia; liver cancer; lung cancer (e.g., small-cell lung cancer, non-small cell lung cancer, adenocarcinoma of the lung, and squamous carcinoma of the lung); melanoma; myeloma; neuroblastoma; oral cavity cancer (lip, tongue, mouth, and pharynx); ovarian cancer; pancreatic cancer; prostate cancer; retinoblastoma; rhabdomyosarcoma; rectal cancer; cancer of the respiratory system; salivary gland carcinoma; sarcoma; skin cancer; squamous cell cancer; stomach cancer; testicular cancer; thyroid cancer; uterine or endometrial cancer; cancer of the urinary system; vulval cancer; lymphoma including Hodgkin's and non-Hodgkin's lymphoma, as well as B-cell lymphoma (including low grade/follicular non-Hodgkin’s lymphoma (NHL); small lymphocytic (SL) NHL; intermediate grade/follicular NHL; intermediate grade diffuse NHL; high grade immunoblastic NHL; high grade lymphoblastic NHL; high grade small non-cleaved cell NHL; bulky disease NHL; mantle cell lymphoma; AIDS-related lymphoma; and Waldenstrom's Macroglobulinemia; chronic lymphocytic leukemia (CLL); acute lymphoblastic leukemia (ALL); Hairy cell leukemia; chronic myeloblastic leukemia; as well as other carcinomas and sarcomas; and post-transplant lymphoproliferative disorder (PTLD), as well as abnormal vascular proliferation associated with phakomatoses, edema (e.g. that associated with brain tumors), and Meigs’ syndrome. The method of any one of claims 3 and 6-46, wherein the cancer is melanoma. The method of any one of claims 3 and 6-46, wherein the cancer is a gastric cancer. The method of any one of claims 1-48 wherein the method identifies a patient for immunotherapy treatment who would not otherwise be identified for immunotherapy treatment using a IHC test for expression of a checkpoint biomarker, optionally selected from PD-1, PD-L1, PD-L2, CTLA-4, LAG-3, TIM-3, B7-H3, B7-H4, ICOS, SIGLEC7, SIGLEC9, and VISTA. A method of identifying an immune cell gene signature profile which is correlative with responsiveness or non-responsiveness to an immunotherapy, the method comprising one or more steps of FIG. 1A and/or FIG. 4A. A companion diagnostic, complementary diagnostic, or codiagnostic test kit, comprising:

(a) an array of nucleic acids suitable for detection of one or more non-coding somatic and/or germline mutations in 3' UTR regions of a patient’s genome, the one or more non-coding somatic and/or germline mutations in 3' UTR regions being indicative of a patient response or non-response to an immunotherapy; and

(b) instructions for use. A companion diagnostic, complementary diagnostic, or codiagnostic test kit, comprising reagents and instructions for use in one or more of claims 1-49.