WO2025054471A1

WO2025054471A1 - Methods and materials for assessing and treating cancers

Info

Publication number: WO2025054471A1
Application number: PCT/US2024/045616
Authority: WO
Inventors: Xiaosong Wang
Original assignee: University Of Pittsburgh - Of The Commonwealth System Of Higher Education
Priority date: 2023-09-08
Filing date: 2024-09-06
Publication date: 2025-03-13

Abstract

This document provides methods and materials involved in assessing and/or treating a mammal (e.g., a human) having cancer. For example, methods and materials that can be used to identify a cancer as being likely to respond to immune checkpoint blockade (ICB) (e.g., administration of one or more immune checkpoint inhibitors) are provided. For example, methods and materials that can be used to treat a mammal (e.g., a human) having cancer where the cancer treatment is selected based on whether or not the cancer is likely to be responsive to ICB (e.g., administration of one or more immune checkpoint inhibitors) are provided.

Description

METHODS AND MATERIALS FOR ASSESSING AND TREATING CANCERS

STATEMENT REGARDING FEDERAL FUNDING

This invention was made with government support under W81XWH-21-1 -1037, HT9425-24-1-0150, and W81XWH-22-1-1036 awarded by the Defense Health Agency, Medical Research and Development Branch, and CA097190, CA237964, CA181368, CA183976, and CA047904 awarded by the National Institutes of Health. The government has certain rights in the invention.

TECHNICAL FIELD

This document relates to methods and materials involved in assessing and/or treating a mammal (e.g., a human) having cancer. For example, methods and materials provided herein can be used to identify a cancer as being likely to respond to immune checkpoint blockade (ICB) (e.g., administration of one or more immune checkpoint inhibitors). For example, methods and materials provided herein can be used to treat a mammal (e.g., a human) having cancer where the cancer treatment is selected based on whether or not the cancer is likely to be responsive to ICB (e.g., administration of one or more immune checkpoint inhibitors).

BACKGROUND

One of the most significant breakthroughs in cancer treatment lies in the development of immune checkpoint blockade (ICB) therapy (Havel et al. Nat. Rev. Cancer, 19: 133-50 (2019); and Nishino et al., Nat. Rev. Clin. Oncol., 14:655-68 (2017)). However, a major limitation of ICB therapy is that only a limited number of patients receive benefits, whereas a significant subset of patients experiences severe autoimmune diseases such as endocrinopathies (Martins et al., Nat. Rev. Clin. Oncol., 16:563-80 (2019)). Furthermore, many patients (e g., about 4% to about 29%) not only cannot benefit from ICB therapy but rather experience dramatic acceleration of disease progression induced by ICB therapy, referred to as ‘hyperprogressive disease’ (HPD). SUMMARY

This document provides methods and materials for assessing and/or treating a mammal (e.g., a human) having cancer. For example, this document provides methods and materials for identifying a cancer as being likely to respond to ICB (e.g., administration of one or more immune checkpoint inhibitors). Tn some cases, this document provides methods and materials for using the intragenic rearrangement (IGR) burden of a cancer to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the IGR burden. For example, a sample containing cancer cells obtained from a mammal having cancer can be assessed to identify the mammal as having a cancer that is likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the IGR burden of the sample. In some cases, this document provides methods and materials for using the tumor associated antigen (TAA) burden of a cancer to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the TAA burden. For example, a sample containing cancer cells obtained from a mammal having cancer can be assessed to identify the mammal as having a cancer that is likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the TAA burden of the sample.

As demonstrated herein, a high IGR burden can indicate that the mammal (e.g., human) is likely to respond to one or more immune checkpoint inhibitors. For example, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal (e.g., a human) having cancer can be used to determine the IGR burden of the cancer, whereby a high IGR burden of the cancer indicates that the cancer is likely to respond to one or more immune checkpoint inhibitors.

Also as demonstrated herein, a high TAA burden can indicate that the mammal (e g., human) is likely to respond to one or more immune checkpoint inhibitors. For example, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal (e.g., a human) having cancer can be used to determine the TAA burden of the cancer, whereby a high TAA burden of the cancer indicates that the cancer is likely to respond to one or more immune checkpoint inhibitors.

Having the ability to identify a cancer as being likely to respond to ICB (e.g., administration of one or more immune checkpoint inhibitors) as described herein (e.g., based, at least in part, on the IGR burden and/or the TAA burden of the cancer) allows clinicians to assess cancer patients in a more accurate manner than current protocols. The ability to identify a cancer as being likely to respond to ICB (e.g., administration of one or more immune checkpoint inhibitors) as described herein (e.g., based, at least in part, on the IGR burden and/or the TAA burden of the cancer) also allows clinicians to provide a personalized approach in selected cancer treatments, thereby improving disease-free survival and/or overall survival for this identified patient population. In addition, the ability to identify a cancer as being likely to respond to ICB (e.g., administration of one or more immune checkpoint inhibitors) as described herein (e.g., based, at least in part, on the IGR burden and/or the TAA burden of the cancer) can minimize subjecting patients to ineffective treatments and/or avoid negative clinical outcomes such as hyperprogressive disease.

In general, one aspect of this document features methods for assessing a mammal having cancer where the methods can include, or consist essentially of, (a) determining that a sample obtained from the mammal and having cancer cells includes a high intergenic rearrangement (IGR) burden or a high tumor associated antigen (TAA) burden; and (b) classifying the cancer as being likely to respond to an immune checkpoint inhibitor based at least in part on the high IGR burden or the high TAA burden. The mammal can be a human. The immune checkpoint inhibitor can be an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti-CTL4A antibody, or an anti-LAG-3 antibody. The immune checkpoint inhibitor can be pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, or relatlimab. The immune checkpoint inhibitor can be BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK503, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, or DZNep. The cancer can include a solid tumor. The cancer can have been previously exposed to platinum. The cancer can lack a high tumor mutational burden (TMB). The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion. The method can include determining that the sample includes the high IGR burden. The method can include determining that the sample includes the high TAA burden. In another aspect, this document features methods for assessing a mammal having cancer where the methods can include, or consist essentially of, (a) determining that a sample obtained from the mammal and having cancer cells has the absence of a high IGR burden and the absence of a high TAA burden; and (b) classifying the cancer as not being likely to respond to an immune checkpoint inhibitor based at least in part on the absence of the high IGR burden and the absence of the high TAA burden. The mammal can be a human. The immune checkpoint inhibitor can be an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti- CTL4A antibody, or an anti-LAG-3 antibody. The immune checkpoint inhibitor can be pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, or relatlimab. The immune checkpoint inhibitor can be BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK503, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, or DZNep. The cancer can include a solid tumor. The cancer can have been previously exposed to platinum. The cancer can lack a high TMB. The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion.

In another aspect, this document features methods for selecting a treatment for a mammal having cancer where the methods can include, or consist essentially of, (a) determining that a sample obtained from the mammal and having cancer cells includes a high IGR burden or a high TAA burden; and (b) selecting an immune checkpoint inhibitor as a treatment for the cancer. The mammal can be a human. The immune checkpoint inhibitor can be an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti-CTL4A antibody, or an anti-LAG- 3 antibody. The immune checkpoint inhibitor can be pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, or relatlimab. The immune checkpoint inhibitor can be BMS-8, BMS-37, BMS-202, BMS- 230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK503, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN,

MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, or DZNep. The cancer can include a solid tumor. The cancer can have been previously exposed to platinum. The cancer can lack a high TMB. The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion. The method can include determining that the sample includes the high IGR burden. The method can include determining that the sample includes the high TAA burden.

In another aspect, this document features methods for selecting a treatment for a mammal having cancer where the methods can include, or consist essentially of, (a) determining that a sample obtained from the mammal and having cancer cells has the absence of a high IGR burden and the absence of a high TAA burden; and (b) selecting a cancer treatment other than an immune checkpoint inhibitor as a treatment for the cancer. The mammal can be a human. The cancer can include a solid tumor. The cancer treatment can include performing surgery. The cancer treatment can include radiation therapy. The cancer treatment can include administering, to the mammal, an anti -cancer agent selected from the group consisting of a chemotherapy and an angiogenesis inhibitor. The cancer can have been previously exposed to platinum. The cancer can lack a high TMB. The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion.

In another aspect, this document features methods for treating for a mammal having cancer where the methods can include, or consist essentially of, (a) determining that a sample obtained from the mammal and having cancer cells includes the presence of a high IGR burden or the presence of a high TAA burden; and (b) administering an immune checkpoint inhibitor to the mammal. The mammal can be a human. The immune checkpoint inhibitor can be an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti-CTL4A antibody, or an anti- LAG-3 antibody. The immune checkpoint inhibitor can be pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, or relatlimab. The immune checkpoint inhibitor can be BMS-8, BMS-37, BMS-202, BMS- 230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK503, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, or DZNep. The cancer can include a solid tumor. The cancer can have been previously exposed to platinum. The cancer can lack a high TMB. The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion. The method can include determining that the sample includes the high IGR burden. The method can include determining that the sample includes the high TAA burden.

In another aspect, this document features methods for treating cancer where the methods can include, or consist essentially of, administering an immune checkpoint inhibitor to a mammal identified as having cancer cells including the presence of a high IGR burden or the presence of a high TAA burden, thereby treating cancer within the mammal. The mammal can be a human. The immune checkpoint inhibitor can be an anti-PD-1 antibody, an anti-PD- L1 antibody, an anti-CTL4A antibody, or an anti -LAG-3 antibody. The immune checkpoint inhibitor can be pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, or relatlimab. The immune checkpoint inhibitor can be BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK5O3, panobinostat, ACY- 241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, or DZNep. The cancer can include a solid tumor. The cancer can have been previously exposed to platinum. The cancer can lack a high TMB. The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion. The method can include determining that the sample includes the high IGR burden. The method can include determining that the sample includes the high TAA burden.

In another aspect, this document features methods for treating a mammal having cancer where the methods can include, or consist essentially of, (a) determining that a sample obtained from the mammal and having cancer cells has the absence of a high IGR burden and the absence of a high TAA burden; and (b) administering a cancer treatment to the mammal, where the cancer treatment is not an immune checkpoint inhibitor. The mammal can be a human. The cancer can include a solid tumor. The cancer treatment can include performing surgery. The cancer treatment can include radiation therapy. The cancer treatment can include administering an anti-cancer agent selected from the group consisting of a chemotherapy and an angiogenesis inhibitor. The cancer can have been previously exposed to platinum. The cancer can lack a high TMB. The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion.

In another aspect, this document features methods for treating cancer where the methods can include, or consist essentially of, administering a cancer treatment that is not an immune checkpoint inhibitor to a mammal identified as having cancer cells having the absence of a high IGR burden and the absence of a TAA burden. The mammal can be a human. The cancer can include a solid tumor. The cancer treatment can include performing surgery. The cancer treatment can include radiation therapy. The cancer treatment can include administering an anti-cancer agent selected from the group consisting of a chemotherapy and an angiogenesis inhibitor. The cancer can have been previously exposed to platinum. The cancer can lack a high TMB. The high TMB can include greater than three mutations per million bases. A population of CD8⁺ T cells within a tumor microenvironment of the cancer can include a low incidence of T cell exhaustion.

Unless otherwise defined, 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 invention pertains. Although methods and materials similar or equivalent to those described herein can be used to practice the invention, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting.

The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

Figures 1 A - 1G. The distribution of IGR burden and its association with a T- inflamed signature in pan-cancer analysis. Figure 1 A depicts the landscape of IGR burden across all cancer types in International Cancer Genome Consortium (ICGC) Pan-Cancer Analysis of Whole Genomes (PCAWG). Figure IB shows a scatter plot of IGR burden, tumor mutation burden (TMB), and T-inflamed score (left), and a scatter plot of IGR burden and TMB (right panel). Tumors with high IGR burden or high TMB show high levels of T- inflamed signature (left panel). Tumors with high IGR burden tend to have low TMB levels, and vice versa (right panel). Figure 1C depicts a median of IGR burden and tumor mutation burden (TMB) in each cancer type. Cancer types that are more than one standard deviation above the median on the X-axis are IGR-driven cancers or on the Y-axis are TMB-driven cancers. Figures ID and IE depict T-inflamed signatures of four subgroups stratified based on IGR and TMB levels in TMB-driven cancers (Figure ID) and IGR-driven cancers (Figure IE). Sample sizes of each subgroup are labeled under each boxplot. P-values were calculated using a one-sided Wilcox rank-sum test. Figures IF and 1G depict the contributions of neoantigen markers to T-inflamed signature in TMB-dominated cancers (Figure IF) and in IGR-dominated cancers (Figure 1G). The left panels of Figures IF and 1G, p-values for each neoantigen marker based on a multivariate model containing all neoantigens when the confounding effects from other neoantigen variables were removed. The right panels of Figures IF and 1G comparing the composite models containing different neoantigen markers with the Y-axis referring to the transformed p-value of the F-test of each composite model. Different multivariable models were compared using ANOVA. **p<0.01.

Figures 2A - 2E. IGR burden correlates with T-inflamed signature in triple negative breast cancers (TNBCs). Figure 2A is a violin plot showing the distribution of IGR burden in TNBC and non-TNBC breast tumors. Figure 2B contains barplots illustrating fractions of patients with tumor infdtrating lymphocytes (TIL) subgroups (left) and mitotic scores (right). The p-values are calculated from the Chi-squared test of the contingency count table (<0.05*). Figure 2C contains jitter plots, with the medians shown in dotted horizontal lines, demonstrate that the differential distributions of T Cell CD8+, Macrophage Ml, Macrophage M2 and Memory CD4+ T Cells deconvoluted from CIBERSORT in IGRhigh and IGRiow tumors. Figure 2D depicts the correlations of neoantigen markers with T-inflamed signature in TNBC tumors that have matched WGS and RNAseq data (n=73). Left panel, the p-values for each neoantigen marker in the multivariate model containing all neoantigen markers. Right panel, comparing the composite models containing different neoantigen markers. **p<0.01. Figure 2E depicts pathway enrichment results from gene set enrichment analysis (GSEA). The pathways are sorted by the direction and logarithm of adjusted p-values from GSEA. Figures 3A - 3F. IGR burden correlates with T-inflamed signature in esophageal adenocarcinoma. Figure 3 A contains boxplots showing the distribution of T-inflamed signature and cell cycle signature in IGRhigh and IGRiow groups. Figure 3B contains dot plots with the medians shown in dotted horizontal lines, demonstrating the distributions of T Cell CD8+ and Macrophage Ml deconvoluted using CIBERSORT in IGRhigh and IGRiow tumors. Figure 3C contains violin plots of expression of genes relevant to anti-tumor immune response and immune checkpoints in TMBhigh and TMBiow samples. Figure 3D depicts pathway enrichment results from GSEA comparing IGRhigh vs IGRio tumors. The pathways are sorted by the direction and logarithm of adjusted p-values from GSEA. Figure 3E contains boxplot comparing the IGR burdens in relapse patients and non-relapse patients in the MEDI4736 clinical trial testing durvalumab in esophageal adenocarcinoma (ESAD) patients. P-value of one-sided Wilcox sum-rank test is shown. Figure 3F depicts the correlations of neoantigen markers with spatial TIL counts in TCGA breast cancer (BRCA) (top), TCGA uterine corpus endometrial carcinoma (UCEC) (middle), or high-grade serous carcinomas (HGSC) of the MSK dataset (bottom) that have matched WGS and spatial TIL count data. Left panels, the p-value for each neoantigen marker in the multivariate model containing all neoantigen markers. Right panels, comparing the composite models containing different neoantigen markers. *p<0.05.

Figures 4A - 4E. IGR burden predicts ICB therapy responses in metastatic urothelial carcinoma patients with low TMB, who received prior platinum therapy. Figure 4A contains dot plots, with the medians shown in dotted horizontal lines, showing the distribution of IGR burden and TMB in samples collected before platinum treatment (platinum-naive) or after platinum treatment (platinum-exposed). Figure 4B contains boxplots showing the IGR burden and TMB for different PD-L1 expression levels of immune cells (IC) based on immunohistochemistry (IHC) before and after platinum. The 3 groups for the IC classes were provided by the IMVigor210: ICO (< 1%), IC1 (>1% and <5%) and IC2+ (>5%). Figure 4C contains pairwise boxplots comparing the IGR burden in responders and non-responders, stratified by TMB levels. Figure 4D contains Kaplan Meier curves of TMB-low patients stratified by IGR levels whose samples are collected before platinum treatment (platinumnaive, left) or after platinum treatment (platinum-exposed, right). P-values of each biomarker using Cox-proportional hazard are shown in the bottom left for each panel. Figure 4E contains receiver operating characteristic (ROC) curves of TMB, IGR and composite biomarker scores in platinum-exposed tumors for determining patient response to immune checkpoint inhibition.

Figures 5A - 5C. Definition of intragenic rearrangements and the datasets and workflow used in this analysis. Figure 5A depicts the datasets and workflow used in the analysis. The variant calling results were retrieved from the ICGC PCAWG cohort for the IGR calculation. Normalized gene expressions were used for the estimation of tumor infiltrating immune subsets and T-inflamed signatures. The results were validated in five clinical datasets. Figure 5B contains a schematic showing the exon alterations resulting from intragenic rearrangements. Figure 5C contains a scatter plot that demonstrates that the number of missense variants almost linearly correlate with the number of all mutations (Pearson R=0.997, p<2.2e-16). Values on axis are normalized by square root.

Figures 6A - 6D. Correlations between IGR burden, TMB and frameshift burden in ICGC PCAWG dataset. Figure 6A is a scatter plot of frameshift burden (x-axis) and TMB (y- axis), colored by T-inflamed signature. Figure 6B is a scatter plot of IGR burden (x-axis) and frameshift burden (y-axis). Figure 6C depicts median TMB vs frameshift burden for each cancer type. Figure 6D depicts median IGR burden vs frameshift burden for each cancer type.

Figures 7A - 7D. Correlations between IGR burden, TMB, somatic copy number aberration (SCNA) and Fusions in the ICGC PCAWG dataset. Scatter plots of ICGC tumors: SCNA vs TMB (Figure 7A), SCNA vs IGR (Figure 7B), Fusion vs IGR (Figure 7C), and Fusion vs SCNA (Figure 7D), colored by T-inflamed signature are shown. Pearson’s R Figures 7A - 7D are 0.023, 0.22, 0.298 and 0.852 respectively.

Figures 8A - 8F. IGR burden correlates with tumor-infiltrating lymphocytes and pro- inflammatory pathways in WGS560 TNBC samples. Figure 8A depicts the distributions of IGR burden in TNBC subtypes. Figure 8B is a scatter plot revealing the association between homologous recombination deficiency (HRD) score (X-axis) and inflame signature (Y-axis), colored according to IGR burden. Figure 8C contains boxplots showing the distribution of inflaming signature, total mitoses and HRD score in IGRhigh and IGRio tumors (samples outside the 10%-90% range are considered outliers). Figure 8D is a scatter plot showing the correlation between SCNA and HRD score (Pearson R=0.263, p=0.0008), and between SCNA and T inflamed signature (Pearson R=-0.037, p=0.76). Figure 8E depicts the p-value for each marker in the multivariate model containing all markers including neoantigen markers and HRD against T inflamed signature when the confounding effects from other variables were removed. Figure 8F contains enrichment plots for the pathways of interest revealed by GSEA analysis.

Figures 9A - 9B. Survival plots of patients stratified based on TMB levels of platinum naive or platinum exposed tumors in the IMVigor210 dataset. Figure 9A contains Kaplan Meier curves of patients stratified based on TMB levels in platinum naive tumors. Figure 9B contains Kaplan Meier curves of patients stratified based on TMB levels in platinum treated tumors. All patients are treated with platinum, but the tumor samples are collected either before platinum treatment (platinum naive) or after platinum treatment (platinum treated). P-values of log-rank tests are shown in the bottom left for each panel.

Figure 10. IGR burden was not predictive of ICB benefit in a dataset for advanced melanoma treated with nivolumab (Riaz et al., Cell, 171 :934-49 (2017)). Treatment naive tumor samples show no significant difference between responders and non-responders in terms of their IGR burden levels.

Figures 11 A-l IB. A relationship between IGR burden and clinical outcomes in patients with triple-negative breast cancer (TNBC), highlighting a differential impact based on treatment modality. Specifically, a higher IGR burden is associated with improved clinical outcomes in TNBC patients receiving chemoimmunotherapy, but conversely, it correlates with poorer overall survival in those not treated with immunotherapy. Figure 11 A) A Kaplan- Meier survival curve depicting the overall survival of TNBC patients enrolled in a prospective phase II clinical trial (n=19) evaluating the efficacy of a chemoimmunotherapy regimen combining Carboplatin, nab-paclitaxel, and pembrolizumab in metastatic TNBC (mTNBC). Figure 1 IB) A Kaplan-Meier survival curve for TNBC patients from the UPMC retrospective cohort (n=46), where 83% of the patients were treated with chemotherapy.

Figures 12A-12D. Development of TAA burden model based on known CTAs and putative TAAs and its correlation with patient response to PD-L1 blockade in metastatic urothelial carcinoma patients of the IMvigor210 clinical trial. Figure 12A) Schematic showing the representative heterogenous expression profiles of canonical TAAs in GTEx normal tissues and TCGA tumor tissues (see Methods). Figure 12B) Heterogenous expression profile analysis (HEP A) markedly enriched known CTAs from the human genome. Figure 12C) CTA burden (based on known CTAs) and tumor-associated antigen (TAA) burden (based on both known CTAs and putative TAAs identified by HEP A) were significantly higher in responders to PD-L1 blockade than non-responders in metastatic urothelial carcinoma (two-tailed Welch’s T test). CR, complete response, PR, partial response, SD, stable disease, PD, progressive disease. Figure 12D) The distribution of TAA burden in different Lund2 and TCGA molecular subtypes of urothelial carcinomas of the IMVigor210 dataset. PD-L1 immune cell (IC) levels are depicted as different gray-scales. The boxplots display the median and interquartile range (IQR), with the whiskers extending up to 1.5 times the IQR from the quartiles. P-values are based on two-tailed Welch’s T test. **, P<0.01, ***, P<0.001, ****, P<0.0001.

Figures 13A-13E. PD-L1 immune cell levels modify the predictive association of TAA burden with patient response and overall survival following immune checkpoint inhibition. Figure 13 A) Interaction effect between TAA burden and potential confounding clinicopathological variables on patient response to PD-L1 blockade, x-axis shows the loglO p-value for interaction effect from logistic regression (to binary response). Multiple logistic regression models were generated for pair-wise interactions of the TAA burden with each of the clinical variables. The models with or without the interaction term for each of the potential interacting variables were compared via Chi-Square test. Figure 13B) TAA burden in responders (R) and non-responders (NR) stratified by PD-L1 immune cell levels. P-values were based on two-tailed Welch’s T test. The boxplots display the median and interquartile range (IQR), with the whiskers extending up to 1.5 times the IQR from the quartiles. Figure 13C) Kaplan-Meier survival curves showing the association of TAA burden with overall survival benefit in patients stratified by PD-L1 immune cell (IC) levels. High TAA burden was determined based on mean + MAD, and P-values were based on log-rank tests. Figure 13D) The association of TAA burden with known biomarkers for predicting ICB response in the IMvigor210 dataset. Pearson correlations between TAA burden and known biomarkers are shown as heatmap. Figure 13E) TAA burden was the most influential covariate of binary response in the ICO patient cohort. Multivariate logistic regression was utilized to model binary response, while considering the presence of other covariates. The bar chart shows the p-value for each marker in the multivariate model, containing all markers, when the confounding effects from other variables were removed.

Figures 14A-14E. Tumor associated antigen burden associates with immune check inhibition response in urothelial carcinomas of non-exhausted tumor immune context. Figure 14 A) Identification of cell state signature as surrogate marker of IC level via correlating deconvoluted cell type signatures with IC levels. 71 cell states were deconvoluted using Ecotyper based on RNAseq data of the whole IMvigor210 patient cohort of all IC levels. Figure 14B) Expression signature of exhausted CD8 cells (Ecotyper CD8 S3) in metastatic urothelial carcinomas of different IC and TC levels. The levels of exhausted CD8 T cells showed better correlation with IC levels than TC levels. Figure 14C) Johnson-Neyman plot showing the correlation of the slope of TAB indicating the predictive value of TAB with Ecotyper CD8 S3 state (left), or T cell exhaustion signature (right) in all patients of the IMVigor210 cohort. The slope of TAB was calculated based on its slope of correlation with binary responses (CR/PR vs SD/PD). Figure 14D) TAA burdens in responders (CR/PR) or non-responders (SD/PD) from tumors of different T cell exhaustion states. The T cell exhaustion states were determined based on Ecotyper CD8 T cell state S3 (left), or T cell exhaustion signature (right). Figure 14E) The receiver operating characteristic curves of TAB, CTB, TMB, or the combination of TAB with TMB for predicting binary responses in the urothelial carcinomas with low exhausted CD8 T cells (Ecotyper S3) stratified based on median TMB levels. The AUROCs for each biomarker or combination are indicated in the figure. P values of the box plots were based on two-tailed Welch’s T test. The boxplots of B and D display the median and IQR, with the whiskers extending up to 1.5 times the IQR from the quartiles.

Figures 15A-15C. The association of tumor associated antigen burden with treatment outcome of anti-PDl or anti-PD-Ll in a separate cohort of metastatic TCC patients. Figure 15 A) Tumor associated antigen burden associated with ICB responses in TCC tumors with unelevated T cell exhaustion signature scores but not in T cell exhausted tumors. Patients were treated with anti-PDl or anti-PDLl (Rose dataset; Rose etal., Br. J. Cancer, 125(9): 1251-60 (2021)). Histology pure TCC patients (n=61) were included in the analysis. TCC patients were stratified based on Ecotyper CD8 T cell state S3 (left), or T cell exhaustion signature (right). The boxplots display the median and IQR, with the whiskers extending up to 1.5 times the IQR from the quartiles. P-values were based on one-tailed Welch’s T test. Figure 15B) The receiver operating characteristic curves of TAB, CTB, TMB, or TAB + TMB for predicting the ICB response of TCC tumors with unelevated Ecotyper CD8 (S3) scores. The patients (n=41) were stratified into TMB unelevated (left) or TMB high (right) groups. The TMB high classification was provided with the Rose dataset. Figure 15C) Progression-free survival (left) and overall survival (right) of TCC patients with unelevated exhausted CD8 (S3) scores (n=41). TAB high tumors were determined based on the cutoff of mean plus median absolute deviation of all TCC tumors. The hazard ratios (HR) and p-values are shown in the figure. The p-values were based on log-rank tests.

Figuresl6A-16C. Tumor associated antigen burden associated with patient response to immune checkpoint blockade in head and neck squamous carcinoma. Figure 16A) The median TAA burden and tumor mutation burden of major solid tumor types of the TCGA Pan-cancer cohort. Figure 16B) TAA burdens (left) and CTA burdens (right) in frozen tumor samples collected at pretreatment and posttreatment timepoints in the Obradovic dataset (Obradovic et al., Clin. Cancer Res., 28(10):2094-109 (2022)). Patients were treated with 1 month of 240 mg nivolumab every 2 weeks for two doses and were stratified based on responders and non-responders provided by the authors. P-values were based on one-tailed Welch’s T test. Figure 16C) The receiver operating characteristic curves of TAA burden in HNSC patients with unelevated exhausted CD8 T cell state. Left panel shows a neoadjuvant trial of advanced-stage HNSC patients treated with the anti-PD-1 antibody nivolumab (Obradovic dataset). Right panel shows patients recruited for a neoadjuvant pembrolizumab trial (Uppaluri dataset; Uppaluri Ret al., Clin. Cancer Res., 26(19):5140-52 (2020)).

Figures 17A-17B. TAA burden decreases with increased CD8 T cell state S3 in Pancancer. Figure 17A) The correlation of CTA burden, TAA burden, and TMB with CD8 T cell state S3 in the TCGA Pan-cancer patient cohort. The density of the respective antigen burden is shown on the left and the Ecotyper dendric cell state S3 is shown as gray gradient. The median level of the respective antigen burden is shown as dashed line. Figure 17B) TAA burden levels in different cancer entities of the TCGA patient cohort stratified by tertiles of CD8 S3 levels. Cancer types with more than 100 tumor samples were included. The boxplot displays the median and IQR, with the whiskers extending up to 1.5 times the IQR from the quartiles. P-values were based on one-tailed Welch’s T test. *, p<0.05, **, p<0.01, ***, p<0.001, ****, p<0.0001.

Figure 18. The process of immunoediting, which involves initial anti-tumor response mediated by TAAs and the development of TAA tolerance during the tumor evasion phase. In the absence of activated immunity within tumors, TAAs are not be targeted for editing, allowing for activation of TAA-reactive immunity either through co-stimulation or ICB treatment. In contrast, a pre-existing T cell response can diminish TAA burden in tumors at the equilibrium stage via immunoediting. Tumor progression can induce TAA tolerance by limiting TAA-reactive T cell repertoire and/or loss of HLA-I, which can lead to a TAA- tolerance state within the tumors. This suggests three prerequisites for mounting TAA- reactive immune response in established tumors: 1) a high-TAA burden in tumor, 2) a nonexhausted tumor immune context, 3) the presence of costimulatory signal induced by immune-stimulating molecular patterns.

Figure 19. The pathological response rates of patients from different IC groups in the IMVigor210 trial.

Figure 20. Survival ROC curve for TAA burden and Foundation One TMB for predicting patient alive two year after ICB treatment. Kaplan-Meier estimator was used to calculate the sensitivity and specificity for each marker in predicting patient survival at month 24 following ICB treatment based on overall survival data. The R package “survivalROC” is used for this analysis.

Figures 21A-21B. The p-values for TAB, CTB, TMB, molecular subtypes, and clinical variables in the multivariate models containing all variables when the effects from other variables were removed. Figure 21 A) Patient subjects of the ICO group in the IMVigor210 cohort. Figure 2 IB) Patient subjects of the IC 1-2+ group in the IMVigor210 cohort. The p-value of the coefficient of each variable was calculated by ANOVA. The left panel shows multivariate logistic regression to model binary response, adjusted for the presence of other covariates. The right panel shows multivariate Cox proportional hazard regression to model overall survival. Patient subjects stratified by ICO or IC1-2+ groups in the IMVigor210 cohort were included in the analyses. A type-II ANOVA was applied to the fitted logistic regression or Cox proportional hazard regression models to evaluate the significance of each covariate independently, ensuring that the unique contribution of each marker was assessed in the presence of others. The analysis highlighted the relative importance of each marker, with p-values indicating their individual contributions to predicting binary response, after accounting for confounding effects from the other variables.

Figure 22. HEPA based TAB better prioritized responders in the ICO group of the IMVigor210 urothelial carcinoma cohort than known CTA burden (left) and FMOne tumor mutation burden (right).

Figure 23. The distribution of Ecotyper CDS T cell S3, epithelial cell S4, and dendritic cell S3 scores in different Lund2 and TCGA molecular subtypes of urothelial carcinomas of the IMVigor210 dataset. TAA burdens are shown as gray gradient. P-values were based on two-tailed Welch’s T test.*, P<0.05, **, P<0.01, ***, P<0.001, ****, PO.OOOl.

Figures 24A-24B. Two-dimensional density plots showing the correlation between Ecotyper CDS T cell S3 state with T cell exhaustion (Figure 24A) and T effector signatures (Figure 24B).

Figures 25A-25B. Three-way interactions between TMB, Ecotyper CDS S3 signature, and the association of TAB with ICB benefit in urothelial carcinoma patients of the IMVigor210 cohort. Figure 25A) JohnsonNeyman plot showing the correlation of the slope of TAB with Ecotyper signature of exhausted CDS T cell (S3) in patient subjects stratified based on median TMB levels. Figure 25B) Johnson-Neyman plot showing the correlation of the slope of TAB with foundation one TMB in patient subjects stratified based on Ecotyper CDS S3 signature (high vs unelevated). The y-axis shows the slope of TAB, which was calculated based on its slope of correlation with binary responses (CR/PR vs SD/PD).

Figure 26. Correlation between TMB (FoundationOne) and Ecotyper CDS S3 signature in urothelial carcinoma of the IMVigor210 patient cohort.

Figures 27A-27B. Sensitivities and positive predictive values of TAA burden at different cut points in TCC tumors with unelevated CDS S3 state in the IMvigor210 (Figure 27A) and UNC datasets (Figure 27B).

Figures 28A-28B. Associations of T AA and CTA burdens with progression-free survival (PFS) and overall survival (OS) in the TCGA bladder cancer dataset. Figure 28A) Association of TAA burden with patient survival in the TCGA bladder cancer dataset. Figure 28B) Association of CTA burden with patient survival in the TCGA bladder cancer dataset. Patients were stratified based on Ecotyper CD8 S3 scores (median+ median absolute deviation). The p-values were calculated based on log rank tests.

Figure 29. TAA burden in responders and non-responders of nivolumab treatment in HNSC patients stratified based on Ecotyper CDS S3 scores. Figure shows the patients from a neoadjuvant trial of advanced-stage HNSC patients treated with the anti-PD-I antibody nivolumab (Obradovic dataset). P-values were based on one-tailed Welch's T test.

Figure 30. The receiver operating characteristic curves of TAA burden and CT A burden for predicting ICB benefits in two HNSC ICB clinical trial datasets. Left panel shows a neoadjuvant trial of advanced-stage HNSC patients treated with the anti-PD-I antibody nivolumab (Obradovic dataset). Right panel shows patients recruited for a neoadjuvant pembrolizumab trial (Uppaluri dataset). All patient subjects in the two trials were included in the analyses.

Figure 31 A- 3 IB. The levels of CTA, TAA, and tumor mutation burdens in different cohorts of transitional cell carcinoma patients treated with immune checkpoint inhibitors divided by tertiles of the CDS T cell S3 scores. Figure 31 A) The levels of CT A, T AA, and tumor mutation burdens in different patient groups of the IMVigor210 cohort divided by tertiles of the CD8 T cell S3 scores. The PD-L 1 immune cell levels are illustrated as grayscale dots. R, responder (CR/PR), NR, non-responder (PD/SD). Figure 3 IB) The levels of CTA, TAA, and tumor mutation burdens in different groups of transitional cell carcinoma patients treated at UNC (the Rose dataset) divided by tertiles of the CD8 T cell S3 scores. PD-L1 TC or 1C levels were not available in this dataset.

Figures 32A-32B. Receiver operating characteristic curves of TAA and CTA burdens and TMB for predicting ICB benefits in a retrospective patient cohort treated with anti-PDl immune checkpoint inhibition (Liu et al., Nat. Med., 25(\2y.\9 6-21 (2019)). Figure 32A) The predictive value of TAB, CTB, and TMB in the overall cohort stratified by CD8 S3 and TMB levels. Figure 32B) The predictive value of TAB, CTB, and TMB in the patient cohort that did not receive steroids, stratified by CD8 S3 and TMB levels.

DETAILED DESCRIPTION

This document provides methods and materials for assessing and/or treating a mammal (e.g., a human) having cancer. For example, this document provides methods and materials for identifying a cancer as being likely to respond to ICB (e.g., administration of one or more immune checkpoint inhibitors), and, optionally, treating the mammal. In some cases, the methods and materials described herein can be used to predict responsiveness to one or more immune checkpoint inhibitors. For example, a sample (e.g., a sample containing one or more cancer cells) from a mammal having cancer can be assessed to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the IGR burden of the sample. In another example, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal having cancer can be assessed to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the TAA burden of the sample.

In some cases, a mammal (e.g., a human) having cancer can be assessed to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the IGR burden of the cancer. As used herein, the IGR burden of a cancer refers to the number of IGRs (e.g., cryptic IGRs) present in the genome of a cancer cell. In some cases, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal having cancer can be assessed to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the IGR burden of the cancer. For example, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal having cancer can be assessed for the number of IGRs (e.g., cryptic IGRs) present in the genome of one or more cancer cells within the sample.

An IGR can be any appropriate type of IGR. Examples of IGRs that can be included in an IGR burden and can be used to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein include, without limitation, duplications (e.g., exon duplications), deletions, and inversions.

An IGR can be any size IGR (e.g., can include any appropriate length of nucleotides). For example, an IGR that can be included in an IGR burden and can be used to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein can include from about 0.05 kilobases (kb) to about 5000 kb (e.g., from about 0.05 kb to about 4500 kb, from about 0.05 kb to about 4000 kb, from about 0.05 kb to about 3500 kb, from about 0.05 kb to about 3000 kb, from about 0.05 kb to about 2500 kb, from about 0.05 kb to about 2000 kb, from about 0.05 kb to about 1500 kb, from about 0.05 kb to about 1000 kb, from about 0.05 kb to about 500 kb, from about 0.05 kb to about 250 kb, from about 0.05 kb to about 100 kb, from about 0.05 kb to about 50 kb, from about 0.05 kb to about 10 kb, from about 0.05 kb to about 1 kb, from about 0.05 kb to about 0.5 kb, from about 0.5 kb to about 5000 kb, from about 1 kb to about 5000 kb, from about 50 kb to about 5000 kb, from about 100 kb to about 5000 kb, from about 500 kb to about 5000 kb, from about 1000 kb to about 5000 kb, from about 1500 kb to about 5000 kb, from about 2000 kb to about 5000 kb, from about 2500 kb to about 5000 kb, from about 3000 kb to about 5000 kb, from about 3500 kb to about 5000 kb, from about 4000 kb to about 5000 kb, from about 4500 kb to about 5000 kb, from about 0.5 kb to about 4500 kb, from about 1 kb to about 4000 kb, from about 500 kb to about 3500 kb, from about 1000 kb to about 3000 kb, from about 1500 kb to about 2500 kb, from about 0.5 kb to about 500 kb, from about 1 kb to about 1000 kb, from about 500 kb to about 1500 kb, from about 1000 kb to about 2000 kb, from about 2000 kb to about 3000 kb, from about 2500 kb to about 3500 kb, from about 3000 kb to about 4000 kb, or from about 3500 kb to about 4500 kb).

An IGR can be located at any appropriate location within a genome. In some cases, an IGR that can be included in an IGR burden and can be used to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein can include one or more exons. In some cases, an IGR that can be included in an IGR burden and can be used to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein can include one or more introns.

An IGR burden can include any appropriate number of IGRs. For example, an IGR burden can include from about 1 IGR to about 1,000,000 IGRs. In some cases, a high IGR burden refers to an IGR burden that is greater than about 150 IGRs. For example, a high IGR burden refers to an IGR burden that is from about 150 IGRs to about 1,000,000 IGRs.

In some cases, a high IGR burden can be based on the median IGR burden for a statistically significant data set. For example, for a data set for a cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a breast cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a breast cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for an ovarian cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for an ovarian cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a uterine cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a uterine cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for an esophageal cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for an esophageal cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a bladder cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a bladder cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a head and neck cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a head and neck cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a lung cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a lung cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a liver cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a liver cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a sarcoma, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a sarcoma, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. For example, for a data set for a uveal cancer, any IGR burden value greater than the median IGR burden value for that data set can be considered a high IGR burden value. In some cases, for a data set for a uveal cancer, any IGR burden value greater than one, two, or three standard deviations above the median IGR burden value for that data set can be considered a high IGR burden value. When comparing a result from a sample to the median of a statistically significant data set, it is understood that comparable sequencing approaches and techniques are used.

Any appropriate method can be used to determine the IGR burden of a cancer. In some cases, a LINX algorithm can be used to determine the IGR burden of a cancer. In some cases, the IGR burden of a cancer can be determined as described in Example 1.

In some cases, a mammal (e.g., a human) having cancer can be assessed to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the TAA burden of the cancer. As used herein, the TAA burden of a cancer refers to the number of TAAs present in (e.g., expressed by) a cancer cell. In some cases, a TAA can be an oncofetal polypeptide. In some cases, a TAA can be a cancer placenta antigen. In some cases, a TAA can be a cancer germline antigen. In some cases, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal having cancer can be assessed to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors based, at least in part, on the TAA burden of the cancer. For example, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal having cancer can be assessed for the number of TAAs present in (e.g., expressed by) one or more cancer cells within the sample.

A TAA can be any appropriate TAA. In some cases, a TAA that can be included in a TAA burden and can be used to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein can be a cancer-testis antigen (CTA). Examples of TAAs that can be included in a TAA burden and can be used to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein include, without limitation, MAGEA polypeptides, BAGE polypeptides, MAGEB polypeptides, GAGE polypeptides, SSX polypeptides, NY-ESO-1 polypeptides, MAGECI polypeptides, SYCP1 polypeptides, BRDT polypeptides, MAGEC2 polypeptides, SPANX polypeptides, XAGE polypeptides, HAGE polypeptides, SAGE polypeptides, ADAM2 polypeptides, PAGE-5 polypeptides, LIPI polypeptides, polypeptides encoded by a NA88A pseudogene, IL13RA polypeptides, TSP50 polypeptides, CTAGE-1 polypeptides, SPA17 polypeptides, ACRBP polypeptides, CSAGE polypeptides, MMA1 polypeptides, CAGE polypeptides, BORIS polypeptides, HOM-TES-85 polypeptides, AF15q 14 polypeptides, HCA661 polypeptides, JARID1B polypeptides, LDHC polypeptides, MORC polypeptides, SGY-1 polypeptides, SPO11 polypeptides, TPX1 polypeptides, NY-SAR-35 polypeptides, FTHL17 polypeptides, NXF2 polypeptides, TAF7L polypeptides, TDRD1 polypeptides, TDRD polypeptides, TEX15 polypeptides, FATE polypeptides, TPTE polypeptides, CT45 polypeptides, HORMAD polypeptides, CT47 polypeptides, SLC06A1 polypeptides, TAG polypeptides, LEMD1 polypeptides, HSPB9 polypeptides, CCDC110 polypeptides, ZNF165 polypeptides, SPACA3 polypeptides, CXorf48 polypeptides, THEG polypeptides, ACTL8 polypeptides, NLRP4 polypeptides, COX6B2 polypeptides, LOC348120 polypeptides, CCDC33 polypeptides, LOCI 96993 polypeptides, PASD1 polypeptides, LOC647107 polypeptides, TULP2 polypeptides, CT66 polypeptides, PRSS54 polypeptides, RBM46 polypeptides, CT69 polypeptides, CT70 polypeptides, SPINLW1 polypeptides, TSSK6 polypeptides, ADAM29 polypeptides, CCDC36 polypeptides, LOC440934 polypeptides, SYCE1 polypeptides, CPXCR1 polypeptides, TSPY1 polypeptides, TSGA10 polypeptides, PIWIL polypeptides, ARMC3 polypeptides, AKAP3 polypeptides, Cxorf61 polypeptides, PBK polypeptides, C21orf99 polypeptides, OIP5 polypeptides, CEP290 polypeptides, CAB YR polypeptides, SPAG9 polypeptides, MPH0SPH1 polypeptides, R0PN1 polypeptides, PLAC1 polypeptides, CALR3 polypeptides, PRM polypeptides, CT96 polypeptides, LY6K polypeptides, IMP-3 polypeptides, AKAP4 polypeptides, DPPA2 polypeptides, KIAA0100/MLAA-22 polypeptides, DCAF12 polypeptides, SEMG1 polypeptides, POTE polypeptides, G0LGAGL2 FA polypeptides, NUF2/CDCA1 polypeptides, RHOXF2/PEPP2 polypeptides, OTOA polypeptides, CCDC62 polypeptides, GPATCH2 polypeptides, CEP55 polypeptides, FAM46D polypeptides, TEX14 polypeptides, CTNNA2 polypeptides, FAM 133 A polypeptides, LYPD6B polypeptides, ANKRD45 polypeptides, EL0VL4 polypeptides, IGSF11 polypeptides, TMEFF polypeptides, ARX polypeptides, SPEF2 polypeptides, GPAT2 polypeptides, TMEM108 polypeptides, N0L4 polypeptides, PTPN20 A polypeptides, SPAG4 polypeptides, MAEL polypeptides, RQCD1 polypeptides, PRAME polypeptides, TEX101 polypeptides, SPATAI 9 polypeptides, 0DF1 polypeptides, 0DF2 polypeptides, 0DF3 polypeptides, 0DF4 polypeptides, ATAD2 polypeptides, ZNF645 polypeptides, KIF2C polypeptides, SPAG1 polypeptides, SPAG6 polypeptides, SPAG8 polypeptides, SPAG17 polypeptides, FBXO39 polypeptides, RGS22 polypeptides, cylin A polypeptides, KP-OVA52 polypeptides, CCDC83 polypeptides, TEKT polypeptides, NR6A1 polypeptides, TMPRSS12 polypeptides, TPPP2 polypeptides, PRSS55 polypeptides, DMRT1 polypeptides, HEMGN polypeptides, DNAJB8 and polypeptides. In some cases, a TAA that can be included in a TAA burden and can be used to identify a cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein can be as described in Example 3.

In some cases, a panel of tumor associated antigen polypeptides can be assessed to determine a TAA burden across that panel of polypeptides for a particular cancer, and that determined TAA burden can be used to identify that cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein such as when the cancer is PD-L1 immune cell or tumor cell negative, or when the tumor has a low T cell exhaustion signature. In some cases, such a panel of polypeptides can include those polypeptides listed in Example 8. In some cases, such a panel of polypeptides can include 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, or 100 percent of the polypeptides listed in Example 8. For example, a panel of polypeptides that includes any 45 or more of the polypeptides listed in Example 8 (e.g., the first 45 polypeptides listed in Example 8) can be assessed to determine a TAA burden across that panel of polypeptides for a particular cancer, and that determined TAA burden can be used to identify that cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein.

In some cases, a panel of putative tumor associated antigens that includes the putative

TAAs predicted by heterogeneous expression profile analysis (HEPA) listed in Example 8 can be assessed to determine a TAA burden across that panel of TAAs for a particular cancer, and that determined TAA burden can be used to identify that cancer as being likely to respond to one or more immune checkpoint inhibitors as described herein, such as when the cancer is PD-L1 immune cell or tumor cell negative, or when the tumor has a T cell exhaustion signature.

In some cases, a high TAA burden can be based on the median TAA burden for a statistically significant data set. For example, for a data set for a cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a breast cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a breast cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for an ovarian cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for an ovarian cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a uterine cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a uterine cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for an esophageal cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for an esophageal cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a bladder cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a bladder cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a head and neck cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a head and neck cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a lung cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a lung cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a liver cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a liver cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a sarcoma, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a sarcoma, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. For example, for a data set for a uveal cancer, any TAA burden value greater than the median TAA burden value for that data set can be considered a high TAA burden value. In some cases, for a data set for a uveal cancer, any TAA burden value greater than one, two, or three standard deviations above the median TAA burden value for that data set can be considered a high TAA burden value. When comparing a result from a sample to the median of a statistically significant data set, it is understood that comparable sequencing approaches and techniques are used.

Any appropriate method can be used to determine the TAA burden of a cancer. In some cases, a HEPA algorithm can be used to determine the TAA burden of a cancer. In some cases, the TAA burden of a cancer can be determined as described in Example 2.

In some cases, a mammal (e.g., a human) having cancer can be assessed to determine the TMB of the cancer. As used herein, the TMB of a cancer refers to the number of mutations (e.g., non-inherited mutations) present in the genome of a cancer cell. In some cases, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal having cancer can be assessed to determine the TMB of the cancer. For example, a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal having cancer can be assessed for the number of mutations present in the genome of one or more cancer cells within the sample.

In some cases, a high TMB can be any level that is greater than three mutations per million bases (mb). In some cases, a high TMB can be any level that is greater than five mutations per mb. In some cases, a high TMB can be any level that is greater than ten mutations per mb.

Any appropriate method can be used to determine the TMB of a cancer. In some cases, the TMB of a cancer can be determined as described in Example 1 or Example 2.

In some cases, a mammal (e.g., a human) having cancer can be assessed to determine the incidence of T cell exhaustion of a cancer within the mammal. As used herein, the T cell exhaustion of a cancer refers to a state in which T cells lose at least some their ability to kill certain cells (e.g., cancer cells).

Any appropriate type of T cell can be assessed to determine the incidence of T cell exhaustion of a cancer within a mammal (e.g., a human) having cancer. For example, CD8⁺ T cells can be assessed to determine the incidence of T cell exhaustion within a mammal (e.g., a human) having cancer.

An exhausted T cell can express one or more T cell exhaustion markers. Examples of T cell exhaustion markers include, without limitation, PDCD1 polypeptides, CTLA4 polypeptides, HAVCR2 polypeptides, LAG3 polypeptides, CD 160 polypeptides, CD244 polypeptides, TIGIT polypeptides, ENTPD1 polypeptides, and BTLA polypeptides.

Any appropriate method can be used to determine the incidence of T cell exhaustion of a cancer within the mammal. In some cases, the incidence of T cell exhaustion of a cancer can be determined as described in Example 2.

Any appropriate mammal having cancer can be assessed and/or treated as described herein. Examples of mammals that can have cancer and can be assessed and/or treated as described herein include, without limitation, humans, non-human primates (e.g., monkeys), dogs, cats, horses, cows, pigs, sheep, mice, and rats. In some cases, a human having cancer can be assessed and/or treated as described herein. When assessing a mammal (e.g., a human) having cancer as described herein and/or treating a mammal (e.g., a human) having cancer as described herein, the cancer can be any type of cancer. For example, a cancer assessed and/or treated as described herein can include one or more solid tumors. In some cases, a cancer assessed and/or treated as described herein can be a primary cancer. In some cases, a cancer assessed and/or treated as described herein can be a metastatic cancer. In some cases, a cancer assessed and/or treated as described herein can be a refractory cancer. In some cases, a cancer assessed and/or treated as described herein can be a relapsed cancer. In some cases, a cancer assessed and/or treated as described herein can be a cancer that was previously treated with one or more platinum-based cancer treatments. In some cases, a cancer assessed and/or treated as described herein can be a cancer that has a low TMB. In some cases, a cancer assessed and/or treated as described herein can be a cancer that has low T cell exhaustion signature. In some cases, a cancer assessed and/or treated as described herein can be a cancer that has a low neoantigen burden. Examples of cancers that can be assessed and/or treated as described herein include, without limitation, breast cancers (e.g., TNBCs), ovarian cancers, uterine cancers (e.g., endometrial cancers such as uterine corpus endometrial cancers), cervical cancers, esophageal cancers (e.g., esophageal adenocarcinomas), bladder cancers (e.g., urothelial cancers), lung cancers (e.g., lung adenocarcinomas), head and neck cancers (e.g., head and neck squamous carcinomas), liver cancers (e.g., liver hepatocellular carcinomas), uveal cancers (e.g., uveal melanomas), and sarcomas. In some cases, a mammal (e.g., a human) having cancer and being assessed and/or treated as described herein, can have cancer present in multiple (e.g., two or more) locations. For example, a human having cancer and being assessed and/or treated as described herein can have a cancer that has metastasized to multiple different locations.

In some cases, the methods described herein can include identifying a mammal (e.g., a human) as having cancer. Any appropriate method can be used to identify a mammal as having cancer. For example, imaging techniques and biopsy techniques can be used to identify mammals (e.g., humans) as having cancer.

Any appropriate sample from a mammal (e.g., a human) having cancer can be assessed as described herein (e.g., for a high IGR burden and/or a high TAA burden of the cancer). In some cases, a sample can be a biological sample. In some cases, a sample can contain one or more cancer cells. In some cases, a sample can contain one or more biological molecules (e.g., polypeptides and nucleic acids such as DNA and RNA). Examples of samples that can be assessed as described herein include, without limitation, tissue samples such surgical tumor samples and cancer biopsies. A sample can be a fresh sample or a fixed sample (e.g., a formaldehyde-fixed sample or a formalin-fixed sample). In some cases, one or more biological molecules can be isolated from a sample (e.g., from one or more cancer cells within the sample). For example, nucleic acid can be isolated from a sample and can be assessed as described herein. In another example, polypeptides can be isolated from a sample and can be assessed as described herein.

Any appropriate method can be used to obtain sequence data for determining a burden described herein (e.g., a high IGR burden, a high TAA burden). For example, RNA and/or polypeptide detection methods can be used to identify the presence or absence of TAAs. In some cases, RNA detection methods including, without limitation, RT-PCR, qRT- PCR, Nanostring, expression microarrays, targeted mRNA sequencing, and whole transcriptome sequencing can be used to determine the presence or absence of TAAs. For example, RT-PCR, qRT-PCR, Nanostring, expression microarrays, targeted mRNA sequencing can be used to assess a sample (e.g., a sample containing cancer cells) for TAAs by examining the list of genes set forth in Example 4 or 8. In some cases, next generation sequencing (NGS)-based detection techniques such as RNAseq can be used to identify the presence or absence of TAAs. For example, RNAseq can be used to assess a sample (e.g., a sample containing cancer cells) for TAAs by examining the list of genes set forth in Example 4 or 8.

As described herein, the IGR burden and/or the TAA burden of a cancer can be used to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors. In some cases, the presence of a high IGR burden (IGRhigh) in a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal (e.g., a human) having cancer can be used to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors. For example, a mammal having a cancer and identified as having a high IGR burden in a sample (e.g., a sample containing one or more cancer cells) obtained from the mammal can be identified as being likely to respond to one or more immune checkpoint inhibitors. In some cases, the presence of a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal (e.g., a human) having cancer can be used to identify the cancer as being likely to respond to one or more immune checkpoint inhibitors. For example, a mammal having a cancer and identified as having a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from the mammal can be identified as being likely to respond to one or more immune checkpoint inhibitors.

In some cases, the IGR burden and the TAA burden of a cancer can be used to identify a cancer as not being likely to respond to one or more immune checkpoint inhibitors. For example, the absence of a high IGR burden and the absence of a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal (e.g., a human) having cancer can be used to identify the cancer as being unlikely to respond to one or more immune checkpoint inhibitors. In some cases, a mammal (e.g., a human) having a cancer that is identified as lacking a high IGR burden and as lacking a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from the mammal can be identified as not being likely to respond to one or more immune checkpoint inhibitors.

In some cases, a mammal (e g., a human) having a cancer that is identified as being likely to respond to one or more immune checkpoint inhibitors as described herein (e.g., based, at least in part, on a high IGR burden and/or a high TAA burden of the cancer) can be selected to receive one or more immune checkpoint inhibitors to treat the cancer. For example, a mammal having a cancer and identified as having a high IGR burden and/or a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from the mammal can be selected to receive one or more immune checkpoint inhibitors.

In some cases, a mammal (e.g., a human) having a cancer that is identified as not being likely to respond to one or more immune checkpoint inhibitors as described herein (e.g., based, at least in part, on the absence of a high IGR burden and the absence of a high TAA burden of the cancer) can be selected to receive an alternative cancer treatment (e.g., one or more cancer treatments that do not include an immune checkpoint inhibitor) to treat the cancer. For example, a mammal having a cancer that is identified as lacking a high IGR burden and as lacking a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal (e.g., a human) having cancer can be selected to receive an alternative cancer treatment (e.g., one or more cancer treatments that do not include any immune checkpoint inhibitors).

This document also provides methods and materials for treating a mammal (e.g., a human) having cancer. In some cases, a mammal (e.g., a human) having cancer and assessed as described herein (e.g., to determine whether or not the cancer is likely to respond to one or more immune checkpoint inhibitors based, at least in part, on (a) the presence of a high IGR burden, (b) the presence of a high TAA burden, or (c) the absence of both a high IGR burden and a high TAA burden of the cancer) can be administered or instructed to self-administer one or more (e g., one, two, three, four, five, or more) cancer treatments, where the one or more cancer treatments are effective to treat the cancer within the mammal. For example, a mammal having cancer can be administered or instructed to self-administer one or more cancer treatments selected based, at least in part, on whether or not the cancer is likely to respond to one or more immune checkpoint inhibitors (e.g., based, at least in part, on (a) the presence of a high IGR burden, (b) the presence of a high TAA burden, or (c) the absence of both a high IGR burden and a high TAA burden of the cancer).

When treating a mammal (e.g., a human) having a cancer that is identified as being likely to respond to ICB (e.g., likely to respond to one or more immune checkpoint inhibitors) as described herein (e.g., based, at least in part, on the presence of a high IGR burden and/or a high TAA burden of the cancer), the mammal can be administered or instructed to self-administer an ICB (e g., one or more immune checkpoint inhibitors). For example, a mammal having a cancer identified as having a high IGR burden and/or a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from the mammal can be administered or instructed to self-administer one or more immune checkpoint inhibitors. An immune checkpoint inhibitor that can be administered to a mammal (e.g., a human) having cancer and identified has being likely to respond to one or more immune checkpoint inhibitors as described herein can be any appropriate immune checkpoint inhibitor. An immune checkpoint inhibitor can inhibit one or more polypeptides involved in an immune checkpoint pathway. Examples of immune checkpoint pathways include, without limitation, PD-1/PD-L2 pathways, and CTLA-4 pathways. An immune checkpoint inhibitor can inhibit any polypeptide involved in an immune checkpoint pathway.

Examples of polypeptides involved in an immune checkpoint pathway that can be inhibited by an immune checkpoint inhibitor as described herein include, without limitation, PD-1 polypeptides, PD-L1 polypeptides, CTLA4 polypeptides, and LAG-3 polypeptides.

An immune checkpoint inhibitor can inhibit polypeptide activity of a polypeptide involved in an immune checkpoint pathway or can inhibit polypeptide expression of a polypeptide involved in an immune checkpoint pathway. Examples of compounds that can inhibit polypeptide activity of a polypeptide involved in an immune checkpoint pathway include, without limitation, antibodies (e.g., neutralizing antibodies) that target (e.g., target and bind) to a polypeptide involved in an immune checkpoint pathway and small molecules that target (e.g., target and bind) to a polypeptide involved in an immune checkpoint pathway. Examples of compounds that can inhibit polypeptide expression of a polypeptide involved in an immune checkpoint pathway include, without limitation, nucleic acid molecules designed to induce RNA interference of polypeptide expression of a polypeptide involved in an immune checkpoint pathway (e.g., a siRNA molecule or a shRNA molecule), antisense molecules that can target (e.g., are complementary to) nucleic acid encoding a polypeptide involved in an immune checkpoint pathway, and miRNAs that can target (e.g., are complementary to) nucleic acid encoding a polypeptide involved in an immune checkpoint pathway. Examples of immune checkpoint inhibitors that can be administered to mammal (e.g., a human) having cancer and identified as being likely to respond to one or more immune checkpoint inhibitors as described herein include, without limitation, anti-PD- 1 antibodies, anti-PD-Ll antibodies, anti-CTL4A antibodies, and anti -LAG-3 antibodies. In some cases, one or more immune checkpoint inhibitors selected from the group consisting of ipilimumab, pembrolizumab, nivolumab, cemiplimab, avelumab, and durvalumab can be administered to mammal (e g., a human) having cancer and identified as being likely to respond to an immune checkpoint inhibitor as described herein. In some cases, an immune checkpoint inhibitor that can be administered to mammal (e.g., a human) having cancer and identified as being likely to respond to one or more immune checkpoint inhibitors as described herein can be as shown in Table 1.

Table 1. Exemplary immune checkpoint inhibitors.

DZNep

PubChem Compound CID: 73087

In some cases, an immune checkpoint inhibitor can be as described elsewhere (see, e.g., Smith et al., Am. J. Transl. Res., 11(2):529-541 (2019) at, for example, Table 1; Terranova-Barberio et al., Immunotherapy, 8(6):705-719 (2016) at, for example, Table 1; and Marin-Acevedo et al., Hematol. Oncol., 14(1):45 (2021) at, for example, Table 1).

When treating a mammal (e.g., a human) having a cancer that is identified as not being likely to respond to an immune checkpoint inhibitor as described herein (e.g., based, at least in part, on the absence of a high IGR burden and the absence of a high TAA burden of the cancer), the mammal can be administered or instructed to self-administer one or more (e.g., one, two, three, four, five, or more) alternative cancer treatments (e.g., one or more cancer treatments that do not include any immune checkpoint inhibitor). For example, a mammal having a cancer identified as lacking a high IGR burden and as lacking a high TAA burden in a sample (e.g., a sample containing one or more cancer cells) obtained from a mammal (e.g., a human) having cancer can be administered or instructed to self-administer one or more alternative cancer treatments that do not include any immune checkpoint inhibitor.

In some cases, one or more (e.g., one, two, three, four, five, or more) alternative cancer treatments (e.g., one or more cancer treatments that do not include any immune checkpoint inhibitor) can include administering to the mammal one or more (e.g., one, two, three, or more) alternative anti-cancer agents used to treat cancer and/or performing one or more (e.g., one, two, three, or more) therapies used to treat cancer. For example, an alternative anti-cancer agent that can be used to treat a mammal (e.g., a human) having cancer and identified as not being likely to respond to an immune checkpoint inhibitor as described herein can be a chemotherapeutic agent. For example, an alternative anti-cancer agent that can be used to treat a mammal (e.g., a human) having cancer and identified as not being likely to respond to an immune checkpoint inhibitor as described herein can be a cytotoxic agent. For example, an alternative anti -cancer agent that can be used to treat a mammal (e.g., a human) having cancer and identified as not being likely to respond to an immune checkpoint inhibitor as described herein can be an angiogenesis inhibitor. Examples of anti-cancer agents that can be administered to a mammal (e.g., a human) having cancer and identified as not being likely to respond to an immune checkpoint inhibitor as described herein to treat the mammal include, without limitation, sorafenib, regorafenib, ramucirumab, and any combinations thereof. Examples of therapies that can be used to treat a mammal (e.g., a human) having cancer and identified as not being likely to respond to an immune checkpoint inhibitor as described herein include, without limitation, radiation therapies and/or surgeries.

In some cases, when treating a mammal (e.g., a human) having cancer as described herein, the treatment can be effective to treat the cancer. For example, the number of cancer cells present within a mammal can be reduced using the methods and materials described herein. In some cases, the size (e.g., volume) of one or more tumors present within a mammal can be reduced using the methods and materials described herein. For example, the methods and materials described herein can be used to reduce the size of one or more tumors present within a mammal having cancer by, for example, 10, 20, 30, 40, 50, 60, 70, 80, 90, 95, or more percent. In some cases, the methods and materials described herein can be used to treat cancer in a manner such that the size (e.g., volume) of one or more tumors present within a mammal does not increase.

In some cases, when treating a mammal (e.g., a human) having cancer as described herein, the treatment can be effective to improve survival of the mammal. For example, the methods and materials described herein can be used to improve disease-free survival (e.g., relapse-free survival). For example, the methods and materials described herein can be used to improve progression-free survival. For example, the methods and materials described herein can be used to improve the survival of a mammal having cancer by, for example, 10, 20, 30, 40, 50, 60, 70, 80, 90, 95, or more percent. For example, the methods and materials described herein can be used to improve the survival of a mammal having cancer by, for example, at least 6 months (e g., about 6 months, about 8 months, about 10 months, about 1 year, about 1.5 years, about 2 years, about 2.5 years, or about 3 years).

One or more immune checkpoint inhibitors can be administered to a mammal (e.g., a human) having cancer in any appropriate amount (e.g., any appropriate dose). In some cases, an effective dose of one or more immune checkpoint inhibitors can be a flat dose. In some cases, as effective dose of one or more immune checkpoint inhibitors can be based on the body of a mammal (e.g., a human) to be treated as described herein. An effective amount of one or more immune checkpoint inhibitors can be any amount that can treat a mammal having cancer without producing significant toxicity to the mammal. The effective amount of one or more immune checkpoint inhibitors can remain constant or can be adjusted as a sliding scale or variable dose depending on the mammal’s response to treatment. Various factors can influence the actual effective amount used for a particular application. For example, the frequency of administration, duration of treatment, use of multiple treatment agents, route of administration, and/or severity of the cancer in the mammal being treated may require an increase or decrease in the actual effective amount administered.

One or more immune checkpoint inhibitors can be administered to a mammal (e.g., a human) having cancer at any appropriate frequency. The frequency of administration can be any frequency that can treat a mammal having cancer without producing significant toxicity to the mammal. For example, the frequency of administration can be from about twice a day to about one every other day, from about once a day to about once a week, from about once a day to about once a month, from about once a week to about once a month, or from about twice a month to about once a month. The frequency of administration can remain constant or can be variable during the duration of treatment. As with the effective amount, various factors can influence the actual frequency of administration used for a particular application. For example, the effective amount, duration of treatment, use of multiple treatment agents, and/or route of administration may require an increase or decrease in administration frequency.

One or more immune checkpoint inhibitors can be administered to a mammal (e.g., a human) having cancer for any appropriate duration. An effective duration can be any duration that can treat a mammal having cancer without producing significant toxicity to the mammal. For example, the effective duration can vary from several weeks to several months, from several months to several years, or from several years to a lifetime. Multiple factors can influence the actual effective duration used for a particular treatment. For example, an effective duration can vary with the frequency of administration, effective amount, use of multiple treatment agents, and/or route of administration.

The invention will be further described in the following examples, which do not limit the scope of the invention described in the claims. EXAMPLES

Example 1: Associations ofIGR burden with immune cell infiltration and response to ICB in cancer

This Example describes the identification of IGR burden increases with prior platinum treatment and can be used to predict patient response to ICB (e.g., treatment with one or more ICB agents). IGRs are typically at tens of kb-level distance, and result in exon duplications or deletions (Figures 5A and 5B). In some cases, the number of cryptic IGRs present within the tumor genome can be used to predict patient response to ICB.

Materials and Methods Data and preprocessing

The data processing workflow and datasets used in this study are described in Figure 5A and Table 2 respectively. TMB was estimated by the total number of missense mutations; somatic SCNA was assessed as described elsewhere (Goldman etal., Nat. Commun., 11 :3400 (2020)). As shown in Figure 5C, the total number of missense mutations almost linearly correlated with the total number of single nucleotide variants (Pearson’s R=0.997, /r<2.2e- l 6, Figure 5D). Biomarkers were square rooted to keep the outlier distribution within a smaller range.

Attorney Docket No. 48881-0054WO1 / 06200

Table 2. Datasets used.

Estimation of intragenic rearrangement burden

IGR can be estimated from variant calling format files from whole genome sequencing (WGS) data and chimeric junction files from RNAseq data. For the ICGC WGS data, the structural variant (SV) calling files from dRanger-snowman and svfix were used. For RNAseq data, chimeric junction files were generated using star aligner vl.8.1. SV junctions were mapped to the exon annotation files for genome build GRCh37 (WGS data) or GRCh38 (RNAseq data) to identify IGR exon junctions. The IGR burden was calculated as the square root of the total number of intragenic rearrangements. Among the datasets analyzed, it was found that the size factor, an indicator of coverage, shows strong correlation on IGR burden in the datasets with read length less than 75 bp. This may be due to the fact that to detect IGR, reads that span the IGR junction are required to identify the precise IGR junctions. When the read length is short, the coverage has a more significant effect on detecting the junction reads. Thus, for the datasets with read length less than 75 bp, the IGR burden was adjusted by dividing the scaled size of the number of sequencing reads. For stratification of IGR or TMB high/low levels, the median was used as the binary cutoff if skewness was below 3 (a skewness greater than 3 indicates absolute non-normal distribution). In the ICGC Pan-cancer dataset, median plus median absolute deviation (MAD) was used as the binary cutoff for the pan-cancer dataset as the distributions of IGR and TMB values become highly right-skewed (the skewness were 9.69 and 4.99 respectively) when different cancers were pooled together.

Expression analysis

CIBERSORT was used on un-logarithm TPM expression data to estimate quantifications of infiltrating immune cell types. For the estimation of T-inflamed signature, the expressions of inflammatory genes were first retrieved (Table 3). Then R package singscore/ 1.14.0 was used to analyze the single sample signature score by rank-based statistics. The cell cycle signature was calculated as the mean of the cell cycle gene set (Table 3). Differentially expressed genes were calculated using R package limma/v3.50.0 with default parameters. The significant genes (p<0.05) were used for pathway analysis. The GSEA pathway analysis using hallmark database was conducted using R packages msigdbr/v7.4.1 an fgsea/vl.21.0. Hallmark pathways with significantly adjusted p-value were shown in the figures.

Table 3.

Statistical analysis

To perform survival analysis on metastatic urothelial carcinoma data, the Kaplan- Meier plot in R package survival/v3.2. 7 as applied. TMB and IGR were utilized in the multivariable Cox -proportional hazard models. In the analysis for TMB, log-rank test was employed to compare the survival distributions between two groups (Figure 9). One-sided Wilcox rank-sum test was employed to calculate the p value of group comparison. Benjamini-Hochberg method was used for false discovery rate adjustment in pathway enrichment analysis. A linear combination of normalized IGR burden and TMB was built for the composite biomarker. Receiver operating characteristic (ROC) curve was used to evaluate the performances of biomarker in response prediction.

Data availability

Genomic and clinical data used in this study can be retrieved through the links provided in Table 2. The IGR burden quantification tool and the scripts used in this study are available through Github. The dataset, gene list and IGR burden used in this study are provided in the Tables 2-4.

Results

High IGR burden defines a group of TMB-low cancer entities.

To assess IGR burden, the total number of intragenic rearrangements were calculated from the structural variant calls of the pan-cancer WGS dataset provided by ICGC (n = 1033). Assessing the distributions of IGR burden across distinct cancer types revealed that low grade glioma (LGG) exhibited the lowest median IGR burden, while breast cancer (BRCA) and Ovarian Cancer (OV) showed the highest median IGR burden (Figure 1 A). Next, the associations between high TMB and IGR burdens in the Pan-cancer dataset were examined (Figure IB). It was found that tumors show either high TMB or high IGR burden (A=0.09,/2=0.004), suggesting two classes of cancer entities driven by either point mutations or intragenic rearrangements. Moreover, it was also observed moderate correlation between TMB and frameshift burden (A=0.59, Figure 6A) but mutual exclusivity in the relationship between IGR and frameshift indels ( ?=0.05, Figure 6B). IGR and SCNA showed weak correlation between each other: IGR high tumors tend to have modest level of SCNA, and SCNA high tumors show low IGR burdens (7?=0.22, Figure 7B). Furthermore, gene fusions resulting from intergenic rearrangements were quantitated and compared to IGR and SCNA. Gene fusion burden showed strong correlation with SCNA, but weak correlation with IGR burden (Figures 7C and 7D).

The distribution of TMB and IGR burden was examined in different tumor entities based on their median levels (Figure 1 C). Based on one standard deviation above the median of the whole population, selected cancer entities were grouped into IGR-driven cancers including breast cancer (BRCA), ovarian cancer (OV), uterine corpus endometrial carcinoma (UCEC), and esophageal adenocarcinoma (ESAD), or TMB-driven cancers including skin cutaneous melanoma (SKCM), lung adenocarcinoma (LU AD), colon adenocarcinoma (COAD), lymphoma (DLBC), and bladder cancer (BLCA). While TMB served as a predictive marker of ICB response in most TMB-driven cancers, its predictive value was limited in IGR-driven cancers. The distribution of frameshift burden versus IGR or TMB in different cancer entities were shown in Figures 6C and 6D.

To investigate the contributions of TMB and IGR burden to T-inflamed signature, the T-inflamed signature was compared in TMB-dominated and IGR-dominated cancers stratified into four groups according to their IGR and TMB levels (Figure ID, E). In TMB- dominated cancers, the T-inflamed signatures of TMBhigh tumors were significantly higher than those of TMBiow tumors in IGRiow tumors (p=0.0002, Figure ID). In IGR-dominated cancers, samples with high IGR burden show significantly elevated T-inflamed signature compared to those with low IGR burden in TMBiow tumors (p=0.0002, Figure IE). Next, it was sought to explore the contributions of multiple types of neoantigen burdens to T cell inflammation levels estimated using the T-inflamed signature. In addition to IGR and TMB, Indels, SCNA, as well as gene fusions resulting from intergenic rearrangements were also quantitated. In multivariate linear regression analyses that included TMB, Indels, SCNA, fusions, TMB was the most important regressor of T cell inflammation in simple mutation dominated cancers (p=0.002) (Figure IF, left panel), whereas IGR burden was the most influential regressor of T cell inflammation in IGR dominated cancer types (p=0.0098) (Figure 1G, left panel). Inclusion of IGR burden in the composite model containing TMB, Indels, SCNA, and fusions resulted in a significant increase of performance in predicting T cell inflammation in IGR-dominated cancers but not in TMB-dominated cancers (Figures 1F- 1G, right panel).

IGR burden correlates with tumor-infiltrating lymphocytes in triple-negative breast cancer To validate these findings in IGR-driven cancers, an independent breast cancer cohort, WGS560, was analyzed in which triple-negative breast cancer (TNBC) exhibited a much higher overall IGR burden than non-TNBC breast cancer (Figure 2A). Among TNBC subtypes, BL1, IM, and M subtypes exhibited higher IGR levels than LAR and BL2 subtypes (Figure 8A). TNBC tumors stratified by their IGR burden exhibited significant differences in lymphocyte infiltrations (p=0.029) and mitotic scores (p=0.039, Figure 2B) based on histopathology data, with higher proportions of severe or moderate lymphocyte infiltration and mitotic score 3 in IGRhigh samples.

The infiltrated immune cells in tumor microenvironment deconvoluted were then examined using CIBERSORT. Compared to tumors with low IGR burden, TNBC tumors bearing high IGR burden exhibited higher levels of T Cell CD8+, Macrophage Ml and CD4 memory-activated cells, but lower level of Macrophage M2, suggesting a type-1 anti-tumor immunity (Figure 2C). In addition, IGRhigh tumors also exhibited higher T-inflamed signature, total mitoses, and homologous recombination deficiency (HRD) scores compared to IGRiow tumors (Figure 8C). This suggested that homologous recombination deficiency and increased mitosis may contribute to the increased IGR burden in TNBC tumors. While IGR burden had a positive correlation with both T-inflamed signature (7?=0.20) and HRD score (7?=0.58), HRD score exhibited only modest correlation with T-inflamed signature (7?=0.11 , Figure 8B). On the other hand, SCNA exhibited a positive correlation with HRD (7?=0.26), but not with T-inflamed signature (R=0.037, Figure 8D). Multivariable linear regression against T-inflamed signature that included TMB, Indel, IGR, Fusion, and SCNA showed that IGR burden was the only significant contributor with the p-value of 0.008 within the model, when the confounding effects from other neoantigen variables were removed (Figure 2D, left panel). On the other hand, TMB and SCNA showed minimal contribution to T-inflamed signature. Next, it was sought to test if adding HRD into the multivariate model would diminish the predictive effect of IGR burden, and results showed that IGR burden was still the only significant predictive variable in the model (p=0.014, Figure 8E). This suggested that while HRD may contribute to both increased SCNA and IGR burden, IGR burden appeared to be the most important regressor of T-inflammation in TNBC. Furthermore, inclusion of the IGR burden in the composite model containing TMB, Indel, SCNA, and fusion resulted in a significant improvement in fitting to the T-inflamed signature (Figure 2D, right panel).

Gene set enrichment analysis comparing IGR high and low tumors revealed upregulation in pro-inflammatory pathways, such as inflammatory response, interferon-gamma response and TNFa signaling via NF-KB pathways, and down-regulation in metabolism pathways such as fatty acid and xenobiotic metabolism (Figure 2E and Figure 8F).

IGR burden correlates with immune response in esophageal adenocarcinoma

Next, the association of IGR burden with immune microenvironment was examined in another ESAD dataset from the ICGC ESAD-UK project (N=100). To examine how TMB level confounds IGR burden in the association with T-inflamed signature and other gene expression signatures, the whole population was divided into four groups according to their values of TMB and IGR burden. A trend of increased T-inflamed signature was observed in IGRhigh tumors compared to IGRiow tumors in both TMB levels (Figure 3A), and IGRhigh tumors have a significant higher cell cycle signature than IGRiow tumors in the TMBiow group. In addition, IGRhigh tumors have higher Macrophage Ml and T Cell CD8+ compared to IGRiow tumors (Figure 3B). Next, the expression of genes relevant to anti-tumor immune response was assessed in IGRhigh vs IGRiow tumors (Figure 3C). Generally, the expression of the TIL markers and key immune checkpoints were significantly upregulated in tumors with high IGR burdens in both TMB levels; however, the difference of expression was more obvious when the mutation burden is low. Further, pathway enrichment analysis revealed upregulation of proliferation pathways such as KRAS signaling, MYC targets VI, G2M checkpoint, and immune response pathways, such as IL2 signaling, inflammatory response, and interferon y response in the tumors with high IGR burden (Figure 3D). Taken together, these results demonstrated that increased IGR burden is associated with type 1 immune response in ESAD. To investigate if IGR burden correlates with ICB response in ESAD, a phase II MEDI4736 trial was accessed in which durvalumab was given to ESAD patients who received prior trimodality chemoradiation therapy (Mamdani et al., Front. Oncol., 11 :736620 (2021)). The results suggested that lower IGR burden was significantly associated with disease relapse, suggesting the correlation of IGR burden with ICB treatment outcome in ESAD (Figure 3E).

IGR burden is a pivotal contributor to spatial abundance of TILs among neoantigen markers in IGR-dominated cancer types.

To examine the correlation of IGR burden with spatial abundance of tumorinfiltrating lymphocytes (TILs), the spatial TIL count data for TCGA tumors was compiled from deep learning of histopathological images (Saltz et al., Cell Rep., 23(1): 181-93 e7 (2018)). Among the TCGA tumors that have WGS data and spatial TIL counts, there were 90 breast carcinomas and 51 endometrial carcinomas. Multivariate linear regression using all neoantigen markers against spatial TIL counts revealed that IGR burden was one of the most influential regressors of spatial TIL counts among all neoantigen markers in breast and endometrial cancers. Inclusion of IGRs to the composite model containing TMB, Indel, SCNA, and fusions significantly increased the predictive value of the model (Figure 3F, upper and middle panels).

Next, it was sought to examine the association of IGR burden with spatial TIL counts in ovarian cancer based on a WGS dataset for high-grade serous carcinoma (HGSC) matched with spatial TIL counts from the MSK-IMPACT cohort. Similar to the above results, IGR burden was the most significant predictor of spatial TIL abundance among all new antigen markers and inclusion of IGR burden to the composite model containing TMB, Indel, SCNA, and fusions significantly increased the predictive value of the model (Figure 3F, lower panel). TMB and indels showed a more significant contribution to the spatial TIL counts in UCEC, consistent with the better predictive value of simple mutation burden in UCEC.

Association of IGR burden with ICB benefit in platinum-exposed metastatic urothelial carcinoma

To further examine whether IGR burden was associated with patient response to ICB treatment, a large clinical trial dataset was accessed in which patients with metastatic urothelial carcinoma received atezolizumab and prior platinum treatment. Urothelial carcinoma was classified as TMB-driven cancer (Figure 1C), however previous research has demonstrated that platinum-based induction therapy leads to extensive DNA damage and activation of the DNA damage repair pathway. Therefore, it was examined whether IGR burden increases with platinum treatment and correlates with T-inflamed signature in the platinum-exposed patients. The distributions of IGR burden and TMB were compared before and after platinum treatment. IGR burden after platinum was significantly higher than before the treatment, whereas no difference in TMB was observed before and after the platinum treatment (Figure 4A). This finding implied that the DNA damaging effect of platinum treatment could lead to increased IGR burden. IGR burden was then correlated with PD-L1 expression as measured by immunohistochemistry (IHC). In patients whose samples were collected before platinum, there was no difference across the PD-L1 expression levels for both IGR and TMB (Figure 4B). After platinum treatment, IGR burden was significantly increased in PD-L1 IC2+ expression, whereas TMB was not (Figure 4B).

The patients with progressive disease were categorized under ICB treatment as nonresponders, and the rest as responders. For platinum-exposed tumors, responders exhibited a significant higher IGR burden compared to non-responders in TMBiow tumors (Figure 4C). The performance of IGR and TMB also was evaluated in survival analysis and ROC. Multivariate cox-proportional hazard models were used to calculate the hazard ratio of each biomarker. TMB was predictive of survival for the platinum-naive tumors, but not for platinum-exposed tumors (Figure 9). However, for the platinum-exposed tumors, IGR was predictive of ICB response in TMB low tumors (Figure 4D). There was no correlation between IGR burden and TMB in both groups (A=0.057 in platinum-naive group; -0.109 in platinum-exposed group). ROC curves revealed that TMB alone has modest predictive value in the TMBiow platinum-exposed tumors with an AUROC of 0.582, whereas IGR burden alone resulted in an AUROC of 0.736. Linear combination of IGR with TMB resulted in an AUROC of 0.757 (Figure 4E). This suggested that IGR burden effectively predicted ICB response in platinum-exposed urothelial tumors when TMB was low. Furthermore, a clinical dataset for advanced melanoma treated with nivolumab (Riaz et al., Cell, 171 : 934-49 (2017)) also was analyzed, and the results showed that IGR burden was not predictive of ICB benefit in melanoma that have the highest TMB level among all cancer entities (Figure 10).

Association of IGR burden with ICB benefit in triple-negative breast cancer (TNBC).

To examine the association of IGR burden with ICB benefit in TNBC, a prospective phase II clinical trial dataset (n=19) that evaluated the efficacy of a chemoimmunotherapy regimen combining carboplatin, nab-paclitaxel, and pembrolizumab in patients with metastatic TNBC (mTNBC) was analyzed (Wilkerson AD, et al. Clin Cancer Res 2024 Jan 5;30(l):82-93. PMID: 37882661). Kaplan-Meier survival analysis revealed that a higher IGR burden was significantly associated with improved overall survival in patients treated with this chemoimmunotherapy regimen (Figure 11 A). The hazard ratio (HR) for overall survival in patients with high IGR burden compared to those with lower IGR burden was 0.20, indicating a potential predictive effect of higher IGR levels in this context (p=0.086). To further explore the prognostic role of IGR burden, data from a retrospective cohort of TNBC patients treated at UPMC (n=46), where 83% of patients received chemotherapy were analyzed. All patients in the UPMC cohort did not receive immunotherapy. Kaplan-Meier survival analysis (Figure 1 IB) demonstrated a contrasting effect, where higher IGR burden correlated with poorer overall survival in TNBC patients who did not receive immunotherapy. This adverse association suggests that while high IGR burden may predict favorable outcomes in the context of chemoimmunotherapy, it might be a negative prognostic factor in the absence of immunotherapy (HR=4.29, p=0.032).

Together, these results demonstrate that IGR can be used to predict ICB response and/or overall survival in mammals (e.g., humans) having cancer (e.g., an esophageal cancer, a triple-negative breast cancer, or a bladder cancer such as a bladder cancer previously treated with one or more platinum-based cancer treatments).

Example 2: TAA burden predicts ICB benefit in tumor entities with low T cell exhaustion and mutation burden.

This Example describes the development of a TAA-based scoring method that can be used to identify a patient as being likely to respond to ICB treatment (e.g., likely to respond to one or more immune checkpoint inhibitors).

Shared tumor-associated antigens (TAAs), such as cancer-testis antigens (CTAs), have been studied for more than two decades and constitute a major class of targets for cancer vaccination and adoptive T-cell therapy (Coulie et al., Nat. Rev. Cancer, 14(2): 135-46 (2014); and Wang et al., Nucleic Acids Res., 34(Database issue):D607-12 (2006)). While robust antitumor responses against CTAs have been observed in murine models, cancer vaccine trials against CTAs in unselected melanoma patients have been largely disappointing (Gjerstorff et al., Oncotarget, 6(18)45772-87 (2015); and Bethune et al., Curr. Opin. Biotechnol., 48: 142-52 (2017)), suggesting a hallmark difference between human and murine models. Since 2011, immune checkpoint blockade (ICB), which targets suppressive receptors on T cells, has achieved major breakthroughs in cancer treatment leading to durable responses (Pardoll, Nat. Rev. Cancer, 12(4):252-64 (2012)). A major limitation of ICB therapy, however, is that only a small proportion of patients receive benefits, whereas a substantial proportion experience severe immune-related adverse effects (irAEs) or treatment-induced hyper-progression (Martins et al., Nat. Rev. Clin. Oncol., 16(9):563-80 (2019); and Adashek etal., Trends Cancer, 6(3): 181-91 (2020)). For example, most urothelial carcinoma (UC) patients do not derive benefits from ICB treatment despite the high costs of treatment (Walia et al., Cancers (Basel), 2021 ; 14(1)), and only 15-25% responders show durable response (Lavoie et al., J. Urol., 202(l):49-56 (2019)). On the other hand, 28.8% of UC patients experienced severe or even lethal irAEs (Sanda et al., Oncologist, 28(12): 1072-8 (2023)).

In this example, a TAA burden (TAB) algorithm was developed based on known CTAs and putative TAAs that attenuates the influence of batch effects between unmatched tumor and normal tissue datasets. Furthermore, a link was unveiled between TAB and the effectiveness of ICB therapies within certain cancer types in the context of a non-exhausted TIME and low TMB. To unify the analysis, a gene expression signature was pinpointed for quantitating CD8⁺ T-cell exhaustion state and applied consistent stratification criteria in the analyses of multiple independent datasets. The results suggest that TAB associates with ICB benefits in UC and Head-Neck Squamous cell Carcinoma (HNSC) with low T-cell exhaustion state. In most tumor entities, it was observed that tumors exhibiting a higher T- cell exhaustion state tended to have a lower level of TAAs, suggesting immunoediting of TAAs in tumors with pre-existing immunity. This study challenges the prevailing belief on the lack of association of TAAs with ICB response in cancer.

MATERIALS AND METHODS

Genomic datasets and clinical trial datasets

TCGA Pan-cancer gene expression and non-synonymous somatic mutation data (PanCancer GDC version) were retrieved from UCSC Xena browser (xenabrowser.net). The Genotype-Tissue Expression (GTEx) bulk normal somatic tissue expression data were retrieved from GTEx portal (gtexportal.org/). The clinical trial datasets were retrieved from Gene Expression Omnibus (GEO, ncbi.nlm.nih.gov/geo), the European Genome-phenome Archive (EGA, ega-archive.org), or database of Genotypes and Phenotypes (dbGaP, dbgap.ncbi.nlm.nih.gov/), and the details of these datasets were summarized in Table 4 (Example 3).

The CT A burden algorithm

To calculate CTA burden, a known CTA gene set was compiled from the CT antigen database (Almeida et al., Nucleic Acids Res., 37(Database issue):D816-9 (2009)), a database that documents experimentally validated and curated CTAs. The list of known CTA genes used in this study is provided in Table 5 (Example 4). To calculate CTA burden, the tumor expression data from the clinical trial datasets was compared with the GTEx expression data for normal somatic tissues (n=13280). Immunoprivileged tissues such as brain, ovary, and testis as well as cell lines were excluded from the analysis. The expression data in the tumor dataset and GTEx dataset were normalized using quantile normalization. To mitigate batch effects between the clinical trial dataset (Table 4), comprising tumor samples, and the GTEX somatic normal tissue dataset, which gathers normal tissue data from healthy individuals, the “ComBaf ’ R package was employed. Although it was not feasible to match the outcome variable across these datasets, the application of ComBat addressed the baseline variations of TAAs that showed outlier expressions in both tumor and normal samples.

An initial binary algorithm was designed and used to estimate CTA burden by counting the number of overexpressed CTA as set forth in U.S. Provisional Patent Application Serial No. 63/537,304. Given a set of cancer-testis antigens, for a patient i, an overexpression event of CTA j was counted if its relative expression quantile R was greater than that of a normal tissue counterpart: CTA count i = (Rgi > Rgn )B. All the overexpression events were summed up to represent the CTA counts. The normal tissue expression profde was extracted from TCGA normal samples. A threshold 99% was set in practice to allow for some rare outlier overexpression of TAAs in normal tissues, where the overexpression event was counted when the Rgi was greater than 99% of all the Rgn.

A more robust algorithm was designed and used to calculate the cumulative overexpression pattern of CTAs in tumors as set forth in U.S. Provisional Patent Application Serial No. 63/654,743 and as described herein. Instead of simply counting the overexpressed CTAs, this algorithm was designed to transform the overexpression levels using logistic sigmoid function and quantitate CTA cumulative overexpression patterns with minor consideration of the overexpression levels. This, to some degree, attenuated the batch effects of clinical trial datasets with unmatched normal tissues. In this method, the differential overexpression levels (Jex >) between a selected tumor from a clinical trial dataset and nonimmunoprivileged normal somatic tissues profded by GTEx were estimated. For this calculation, the higher of two thresholds: 95% percentile of expressions in normal somatic tissues, or 98% percentile of the normal tissue counterpart where the tumor originates from, was used to compute level of overexpression for a CTA.

Then CTA burden was calculated using the following formula:

This algorithm allows quantitation of the cumulative overexpression pattern of CTAs with consideration of their overexpression levels. This method to some degree incorporated the level of Aexp in the model via logistic sigmoid transformation and the A in the formula determines the slope of the sigmoid curve. The higher the A, the deeper was the Hill Slope, and the output was closer to counting the number of overexpressed CTAs.

Heterogenous expression profile (HEP A) analysis and calculation ofTAA burden

HEPA analysis was performed using the TCGA Pan-cancer dataset, and the GTEx Pan-normal tissue dataset. For a given gene i in a given cancer type k, an outlier expression score was calculated based on an adjusted upper quantile mean of its expression in TCGA cancer type k, which is defined as the mean level of the 75th percentile to 95th percentile expression values of gene i. To evaluate the expression profile of candidate antigens in normal tissues, for the same gene i, its average expression in normal somatic tissue types was calculated based on trimmed mean using the GTEx dataset, which was used to calculated a normal expression depreciation penalty for gene i as described elsewhere (Xu etal., Cancer Res., 72(24): 6351-61 (2012)). The immuneprivileged organs such as brain, testis, and ovary were excluded. Then the HEPA score was calculated by subtracting the log2 transformed outlier expression score in cancer type k and the normal expression depreciation penalty. Genes were then ranked by their max HEPA scores across different cancer types, known protein coding genes with and a max HEPA cut off of 0.2 was considered as putative TAA encoding genes based on the optimal enrichment and detection of known CTAs. Genes encoding cytokines, or genes detected in normal blood (based on Human Protein Atlas database: proteinatlas.org) were filtered out. The final TAA gene set was generated by combining the CTA gene set with the putative TAA gene set, and TAA burden was calculated based on the same algorithm as the CTA burden detailed above. Calculation of immune signatures and cell states

To calculate immune-related scores, cancer immune-related gene sets as described by Trujillo et al. (Cancer Immunol Res., 6(9):990-1000 (2018)) were collected, and a T-cell exhaustion signature that compared genes upregulated in exhausted CD8⁺ T cells compared to effector CD8⁺ T cells during virus infection was accessed (Wherry et al., Immunity, 27(4):670-84 (2007)). Singscore (Foroutan etal., BMC Bioinformatics, 19(l):404 (2018)) was then used to calculate gene set scores for each tumor sample. To systematically quantitate immune cell states based on RNAseq data, a computational tool called Ecotyper (Luca et al., Cell, 184(21):5482-96 (2021)) that utilizes transcriptome sequencing to characterize cell states and ecosystems in tumors was leveraged. High Ecotyper CD8 S3 state was defined based on the cutoff of median + median absolute deviation based on the outlier expression profile of the score across tumors.

Statistical analysis

For interaction analysis, multiple logistic regression models were generated for pair- wise interactions of the TAA burden with each of the potential interacting factors and the multivariate regression models with or without the interaction term for each of the potential interacting variables were compared via Chi-Square test implemented in the ‘anova’ function, as described elsewhere (Wang et al., Nat. Commun., 13 (1 ) :2936 (2022)). Two way and three-way interactions between TAB and confounding variables were analyzed by Johnson-Neyman plot using R package “interactions”. E-values of the box or dot plots were analyzed based on Welch’s T test. Two-tailed tests were used during the hypothesis generating process using the IMvigor210 dataset, and one-tailed tests were used when directional hypothesis was made to provide better statistical power on clinical trial datasets of small sample sizes. The correlation analyses of various ICB markers were performed based on Pearson’s statistics. The composite model for TMB and TAB were generated based on logistic regression modeling binary response. The cutoffs were determined based on mean or median plus median absolute deviation in accordance with the normality of the data. For data that showed strong outliers, such as Ecotyper CD8 S3 state, median plus median absolute deviation (MAD, default constant) was used as cutoff, for data that show normality, such as TAB or CTB, mean plus MAD was used as cutoff. For survival analysis, the P-values were calculated based on log-rank tests. Receiver operating characteristic (ROC) curves were used for analyzing the predictive values of biomarkers on binary responses. Multivariate logistic regression or Cox proportional hazard regressions were utilized to model binary response or overall survival respectively, while considering the presence of other covariates. For models consisting solely of continuous variables, p values of z statistics (Pr(>|z|)) were computed. In cases involving categorical covariates, a type-II ANOVA was employed on fitted logistic regression or Cox proportional hazard regression models to assess the independent significance of each covariate.

Data availability

The TCGA Pan-cancer gene expression datasets can be retrieved from xenabrowser.net, and GTEx normal tissue expression data can be retrieved from gtexportal.org/. The clinical trial datasets can be retrieved from GEO (ncbi.nlm.nih.gov/geo), dbGaP (dbgap.ncbi.nlm.nih.gov/), or EGA (ega-archive.org). The accession numbers of these clinical trial datasets are summarized in Table 4. The CTA and TAA burdens, CD8 S03 and T cell exhaustion scores for all datasets, putative TAAs predicted by HEP A, and immune gene sets used in this example are as described in Wang (Wang et al., Cancer Immunol Res., (2024) doi.org/10.1158/2326-6066.CIR-23-0932). The R package for calculating CTA and TAA burdens is available on GitHub at: github.com/wangxlab/TAA-burden.

RESULTS

Development of a TAA burden algorithm based on recognized CTAs and putative TAAs.

In this example, an algorithm was developed to calculate the cumulative overexpression pattern of TAAs in tumors. Instead of simply counting the overexpressed TAAs, this algorithm transformed the overexpression levels using logistic sigmoid function and quantitates TAA cumulative overexpression pattern with minor consideration of the overexpression levels. This, to some degree, attenuated the batch effects of clinical trial datasets with unmatched normal tissues. To systematically study the cumulative expression pattern of TAAs, a known CTA gene list (n=243) was first compiled from the CTD database, a database that documents experimentally validated cancer-testis (CT) antigens identified thus far. In addition, it was sought to leverage the HEPA algorithm, to identify putative TAAs based on the Pan-cancer RNAseq datasets from TCGA, and normal somatic tissue datasets from the GTEx Project. HEPA analysis utilized specific algorithms designed to account for the unique expression profiles of TAAs: 1) prototype TAAs demonstrate remarkable overexpression in a limited subset of tumors, and 2) their expression in somatic normal tissues is restricted to immunologically privileged locations, such as germ cells (Fig. 12A). The efficacy of the HEPA algorithm has been validated through studies of a large panel of patient blood from multiple tumor entities, ranking the human genes by their HEPA scores substantially enriched the known CTAs from the genome (Fig. 12B). Finally, a CTA burden was calculated that represents the overexpression pattern of known CTAs, and a TAA burden that represents the overexpression pattern of both known CTAs and putative TAAs.

Cumulative overexpression pattern of TAAs is associated with clinical response to PD-L1 blockade in mUC.

To examine the effect of TAB on clinical outcomes, one of the largest transcriptomic datasets of ICB clinical trials for non-melanoma cancer patients was used as the discovery set. This trial profiled a large cohort of metastatic UC (mUC) patients treated with atezolizumab in the IMvigor210 clinical trial by transcriptome sequencing (Mariathasan el al., Nature, 554(7693 ): 544-8 (2018)). UC, also known as transitional cell carcinoma (TCC), originates from the urothelial cells of the urinary system, and is one of the highest TAA expressers among non-melanoma cancer entities. In the IMvigor210 trial, histology pure or predominant TCC patients were enrolled, and all patients were treated with 1,200 mg i.v. every 3 weeks until loss of clinical benefit (Balar etal., Lancet, 389(10064):67-76 (2017); and Rosenberg et al., Lancet, 387(10031): 1909-20 (2016)). In this study, patients with complete or partial responses (CR or PR) were defined as responders, and those with stable or progressive disease (SD or PD) as non-responders. When applying the model to the RNAseq data of pre-treatment mUC samples, a significant increase in both CTA and TAA burdens was observed in responders compared with non-responders. In particular, TAB exhibited a more pronounced and statistically significant association with treatment response (Fig. 12C). Next, the levels of CTA and TAA burdens across various molecular subtypes of bladder cancer within the IMVigor210 cohort were investigated. The basal/SCC-like subtype of Lund2, and the TCGA III/IV subtypes (recognized as basal subtypes), exhibited significantly elevated CTA and TAA burdens (Fig. 12D).

Clinicopathological variables that modify the association ofCTAs with ICB treatment.

To investigate how clinicopathological variables influence the relationship between TAB and the effectiveness of ICB therapy, a comparative analysis was conducted of multivariate logistic regression models by incorporating versus excluding an interaction term for each clinicopathological variable. This revealed that PD-L1 staining on immune cells displayed the most pronounced interaction effect with the association of TAB and ICB benefits (Fig. 13 A). PD-L1 expression in immune cells was a better predictive marker for ICB response than PD-L1 expression in tumor cells in certain cancer types such as UC and head and neck cancer (Kim el al., Sci. Rep., 6:36956 (2016); and Liu el al., Dis. Markers, 2020:8375348 (2020)). PD-L1 expression levels in immune cells (referred to as IC levels) can be classified into ICO, IC1, or IC2+ based on PD-L1 expression in <1%, >1% and <5%, or >5% tumor-infiltrated immune cells. In the IMvigor210 Trial, a favorable response to ICB treatment was observed in 16% (13/83) of ICO patients (Fig. 19). In the analysis, a significant difference in TAB was observed between responders and non-responders within the ICO group, but not in the IC1 or IC2+ groups (Fig. 13B). The differences of TAB associated with ICB response within IC level groups were more pronounced compared to stratification by tumor cell PD-L1 levels. Stratified survival analysis based on IC levels indicated a significant survival difference between tumors categorized as TAB-high and TAB-low within the ICO group (Fig. 12C, Fig. 20). This finding suggests a potential variation in the immunogenicity of TAAs in accordance with TIME.

TAA burden is the most influential covariate of ICB benefit in the PD-L1 ICO group.

TAB was then compared with other known predictive markers of ICB response, including TMB, intragenic rearrangement (IGR) burden (Zhang et al., Cancer Immunol. Res., 2024:OF1-OF9 (2024)), CD274 (PD-L1), T effector signature (Herbst et al. , Nature , 515(7528):563-7 (2014)), CXCL13 (Hsieh et al., Cancers (Basel), 14(2) (2022)), and panfibroblast TGF-P signature (Mariathasan et al., Nature, 554(7693):544-8 (2018)). TAB showed minimal to no correlation with these biomarkers (Fig. 13D). In a multivariate logistic regression model including all these markers, TAB was the only significant predictor of binary response in the ICO group (Fig. 13E). To examine if this association could be driven by molecular subtypes or other clinical variables, multivariate logistic regression as performed modeling binary responses and Cox proportional hazard regression analyses modeling overall survival. Both models incorporated biomarkers such as TAB, CTB, TMB, molecular subtypes such as Lund2 and TCGA subtypes, and clinicopathological variables such as metastasis status, Bacillus Calmette-Guerin (BCG) vaccine administration, and platinum exposure. These analyses revealed that TAA and CTA burdens held the greatest influence on treatment outcomes in the ICO group (Fig. 21A), whereas TMB was the most influential antigen covariate on treatment outcome in the IC1-2+ groups (Fig. 21B). In addition, TAA burden demonstrated a higher predictive value than CTA burden in this analysis, underscoring the advantage of including putative TAAs in the predictive model for better estimation of TAA burden (Fig. 21 A and Fig. 22).

PD-L1 IC level is associated with a cell ecosystem that reflects pre-existing antitumor immunity.

The impact of PD-L1 IC levels on the functional states of tumor-infiltrated immune cells and their potential contribution to tolerance towards TAAs was investigated. A computational tool called Ecotyper that utilize transcriptome sequencing to characterize the states of a variety of tumor, stroma, and immune cell types in tumors was leveraged. The resulting 71 cell states were correlated with PD-L1 IC levels across the entire IMVigor210 patient cohort. This analysis identified CD8 T cell S3, epithelial cell (EC) S4, and dendritic cell (DC) S03 as the most correlated cell states (Fig. 14A). Among the bladder cancer molecular subtypes, this cell ecosystem was more enriched in basal-like tumors (Fig. 23), which aligns with the enrichment of PD-L1⁺ immune cells in this subtype (Lopez-Beltran et al., Cancers (Basel), 13(1) (2021); doi 10.3390/cancersl3010131). EC S4 and DC S03 represent pro-inflammatory epithelial cells and mature inflammatory DC cells respectively. CD8 T cell S03 state expressed both markers of effector cells (i.e., IFNG, GZMB) and exhausted cells (i.e., LAG3). Among different PD-L1 levels, the CD8 S3 state displayed a stronger correlation with immune-cell PD-L1 levels compared to tumor cell levels (Fig. 14B). Next, the correlation between CD8 S3 state and a T-cell exhaustion signature, and an effector T-cell signature was analyzed. CD8 S3 state positively correlated with the T-cell exhaustion signature but negatively correlated with the effector T-cell signature (Fig. 24). These findings suggested that PD-L1 IC stains were indicative of a pro-inflammatory cell environment compromised by T-cell exhaustion, potentially resulting from pre-existing antitumor immunity. To ensure consistency in the analyses, CD8 T cell S3 state was employed as a stratification factor, which incorporated markers for both CD8⁺ T cells and the T cell exhaustion state. In contrast, T-cell exhaustion signatures were derived from differentially expressed genes by comparing exhausted CD8⁺ T cells with effector CD8⁺ T cells. Practically, unlike the limited availability of PD-L1 data, CD8 T cell S3 state can be determined through deconvolution analysis of RNAseq data, offering a more accessible stratification method.

TA A burden correlates with ICB benefit in urothelial tumors of low CD8 T cell S3 state.

It was speculated that the relationship between TAB and the benefits of ICB therapy may be weakened in an exhausted tumor immune context, as supported by the diminished predictive value of TAB observed with higher CD8 S3 state and higher T-cell exhaustion signature (Fig. 14C). In the subgroup analyses based on CD8 T cell S3 state or T-cell exhaustion signature, significantly higher TAB were observed in responders within the patient group exhibiting a non-elevated CD8 T cell S3 state or T-cell exhaustion signature (Fig. 14D). In the IMVigor210 cohort, a total of 10. 1% of mUC patients exhibited high TAB levels and a non-elevated CD8 T cell S3 state. Within the TCGA primary bladder cancer cohort, the incidence of tumors with high TAB and low CD8 T cell S3 state increased as tumor stages progressed: 9.2% in stage II (n=130), 14.8% in stage III (n=135), and 15.2% in stage IV (n=132). These data suggested a growing proportion of patients with high TAB and low CD8 T cell S3 state as the tumor stage advances.

Further stratified interaction analysis revealed that in tumors with low TMB, TAB exhibited a strong predictive association only when CD8 S3 state was diminished to a certain level (Fig. 25A). Conversely, in tumors with a non-elevated CD8 S3 state, the predictive value of TAB was evident at lower levels of TMB (Fig. 25B). This suggested that TMB served as a distinct modifying factor, which aligns with the lack of correlation between TMB and PD-L1 levels (Fig. 26) (Jardim et al., Cancer Cell, 39(2): 154-73 (2021)). The patients with unelevated CD8 S3 state were thus stratified into TMB high and low groups based on median TMB levels. Among the TMB low group, the AUROC of TAB, CTB, and TMB for predicting ICB benefit were 0.87, 0.78, and 0.57 respectively, and combining TAB with TMB did not result in better predictive values (Fig. 14E, left), while in TMB high tumors, the AUROCs for TAB and CTB decreased to 0.70 and 0.62 respectively (Fig. 14E, right).

Association of TAB with clinical benefit of UC patients treated with various anti-PDl PD-Ll agents.

To test if this finding could be generalizable to various anti-PDl/PD-Ll agents, an independent dataset for mUC was examined as a validation set; in this dataset, patients were treated with either anti-PDl such as pembrolizumab or nivolumab, or anti-PD-Ll, such as atezolizumab, avelumab, or durvalumab. Clinical benefits were defined as complete or partial responses following ICB and histology pure TCC were included in this analysis to match the pathological characteristics of the IMVigor210 cohort. TAB was only significantly higher in the responders compared to non-responders within the patient group with unelevated Ecotyper CD8 S3 state or T-cell exhaustion signature (Fig. 15 A). The effect was diminished in T-cell exhausted tumors. The AUROC for TAB and CTB were 1.0 and 0.9 in the TMB low group, and 0.86 and 0.8 in the TMB high group (although such high AUROC could be biased due to the small sample size). TMB did not add additional predictive effect when combined with TAB (Fig. 15B). Among the tumors with unelevated exhausted CD8⁺ T-cell state, TAB high tumors (7/41) showed significantly better progression-free survival (PFS) and overall survival (OS) compared to the rest of the tumors (Fig. 15C). In fact, only 2 out of 7 patients with high TAB progressed, whereas all patients with unelevated TAB progressed. About 86% (23/34) patients with high TAB survived at the end of follow-up, compared to 14% (1/7) patients with unelevated TAB.

Taken together, the data suggest the association of TAB with ICB benefits in UC with unelevated T-cell exhaustion signature: in the IMvigor210 cohort, TAB showed a positive predictive value of 44% at a sensitivity of 84% for predicting ICB benefits (binary responses); in the UNC cohort, TAB showed a positive predictive value of 60% at a sensitivity of 86% (Fig. 27). Next, it was sought to examine if such associations were predictive rather than prognostic. As neither the IMvigor210 cohort nor the UNC cohort have a non-lCB treated control arm, it was sought to test the prognostic effect of TAA and CTA burdens in the TCGA patient cohort and examined their associations with PFS and OS in the bladder cancer patients stratified based on T-cell exhaustion state. The results showed that neither high TAA burden nor high CTA burden were significantly associated with PFS and OS in tumors from the TCGA bladder cancer patient cohorts similarly stratified by CD8⁺ T- cell exhaustion states (Fig. 28). These data suggested that the associations of TAB with patient outcomes were more likely predictive of ICB benefit rather than prognostic.

TAA burden associates with clinical benefit in head and neck cancer patients treated with anti-PDl.

Next, it was sought to explore whether this finding could be generalized to other cancer types with overall high CTA burden but moderate to low TMB. To identify suitable cancer types for this analysis, the median CTA burden was plotted against median TMB levels for major solid tumors profiled by TCGA. Processed gene expression and mutation data from TCGA were used for this analysis. Among the solid tumor types, head and neck squamous carcinoma (HNSC), liver hepatocellular carcinoma (LIHC), uterine corpus endometrial carcinoma (UCEC), and adenoid cystic carcinoma (ACC), were the cancer types that showed highest TAB levels but moderate to low TMB (Fig. 16A). Among these cancer types, HNSC was observed to have relatively lower TMB level than bladder cancer, and the TAB levels were higher in Human Papillomavirus (HPV) negative tumors than positive tumors. In both datasets, clinical or pathologic response data following ICB treatment were available, but TMB or patient survival data were not provided.

A neoadjuvant trial of advanced HNSC patients treated with 240 mg nivolumab every 2 weeks for two doses prior to surgery, in which most of the patients were HPV-negative was assessed (Obradovic et al., Clin. Cancer Res., 28(10):2094-109 (2022)). Tadalafil was administered in half of the patients in addition to nivolumab, which did not result in difference in response rates. Thus, patients from both treatment arms were included in this analysis. In pretreatment samples, TAA burden was significantly higher in responders than non-responders (p=0.045), the difference was not significant for CTA burden (p=0.16) (Fig. 16B). In post-treatment samples, both TAA burden and CTA burden showed more obvious reduction following ICB treatment in responders compared to non-responders. When stratified by exhausted CD8⁺ T-cell state, HNSC tumors with unelevated CD8 S3 signature showed a trend of more significant difference in TAB levels comparing responders to non- responders (Fig. 29). Among the HNSC tumors with unelevated exhausted CD8⁺ T-cell state, the AUROCs of TAB and CTB for predicting ICB response were 0.71 and 0.6 respectively (Fig. 5C, left panel), and among all subjects, the AUROC of TAB and CTB were 0.68 and 0.66 respectively (Fig. 30, left panel).

To further verify this result, another phase II trial dataset for neoadjuvant pembrolizumab in advanced HPV-unrelated HNSC patients (Uppaluri et al., Clin. Cancer Res., 26(19): 5140-52 (2020)) was assessed. In baseline tumors, the AUROC of TAB and CTB for pathologic responses were 0.68 and 0.60 in all patient subjects (Fig. 30, right panel), and 0.85 and 0.88 respectively in patients with unelevated CD8 T cell S3 signature (Fig. 16C, right panel). These data suggested the association of TAB with ICB benefit in HNSC patients, particularly in tumors with low levels of exhausted CD8⁺ T cells.

TAA burden may be reduced with pre-existing T-cell immunity in cancer.

To assess the influence of pre-established T-cell immunity on TAB, the association of TAB with IC levels in the IMVigor210 cohort was evaluated. This revealed a significantly higher TAB in the ICB responders characterized by an ICO status compared to those in the IC1-2+ group (Fig. 13B). To validate these findings, patients with UC were stratified into three distinct groups according to the CD8 T cell S3 state tertiles across the IMVigor210 and Rose datasets. It was consistently found that both CTA and TAA burdens were elevated in ICB responders with low levels of CD8 T cell S3 signature compared to intermediate or high levels, while TMB persistently exhibited higher levels in responders across all IC categories (Fig. 31). This pattern suggested that TAAs overexpressed in tumors without pre-existing antitumor immunity, may act as primary antigenic targets following an ICB-induced immune response.

To expand the investigation into the association of CTA, TAA, and TMB with the CD8 T cell S3 state, TCGA pan-cancer data was utilized for a broader perspective. This analysis reinforced the above findings: with increases in CD8 S3 state, both CTA and TAA demonstrated a substantial reduction, in contrast to TMB, which maintained high levels in tumors with increasing CD8 S3 state (Fig. 17A). Moreover, in most tumor types, heightened CD8 S3 states correlated inversely with TAB (Fig. 17B). These data collectively suggested that TAB may be diminished in tumors with pre-existing T-cell immunity, possibly resulting from immunoediting processes. In addition, it was observed that melanomas with high T-cell exhaustion displayed a greater TAB, suggesting a deeper state of TAA tolerance in melanoma compared to other cancer types (Fig. 17B).

Assessing the influence ofCD8 T cell S3 state on the association of TAB with ICB response in melanoma

Considering that previous studies indicated TAAs lack predictive value for ICB effectiveness in melanoma (Havel et al., Nat. Rev. Cancer, 19(3): 133-50 (2019)), it was assessed whether stratification by CD8 T cell S3 state might identify predictive patterns. To this end, a large-scale whole-transcriptomic dataset was necessary. A single suitable dataset was identified that included transcriptomic profiles from over 100 melanoma patients who had undergone ICB treatment, along with corresponding TMB data (Liu et al., Nat. Med., 25(12): 1916-27 (2019)). This stratification revealed that in CD8 S3 high melanoma, TMB was a significant predictor of ICB response (AUC=0.71). Whereas among melanomas with lower CD8 S3 scores, CTB and TAB provided moderate predictive utility for ICB success in cases where TMB exceeded the median level (Fig. 32), differing from the findings in TCC and HNSC. More interestingly, discounting patients who received steroids from the analysis enhanced the predictive values of TMB, CTB, and TAB across all groups.

Together, these results demonstrate a correlation between TAA burden and ICB response in certain tumors such as those characterized by lower T-cell exhaustion and mutational burden. For example, these results support an immune-editing model of TAAs in cancer progression (Fig. 18). In tumors devoid of activated immunity, TAAs may remain unedited by the immune system, therefore TAA-reactive immunity can be mounted following ICB treatment. The existence of this TAA-responsive state is corroborated by the demonstration of a correlation of TAA burden with ICB response in tumors that lack an established T-cell response, which exhibit a higher TAA burden than T-cell exhausted tumors. In contrast, a pre-existing T-cell response might diminish TAA burden in tumors at the equilibrium stage via immunoediting, while tumor progression could eventually induce TAA tolerance by restricting the TAA-reactive T-cell repertoire and/or loss of HLA-I, referred to herein as a TAA-tolerance state. The presence of a TAA-tolerant state could be inferred from the elevated TAA burden in PD-L1 IC2+ or high CD8 S3 groups in advanced or metastatic UC cancers. Tumors with different levels of T-cell exhaustion may serve as snapshots of various stages of this immunoediting and tolerance process against TAAs (Fig. 18). This contrasts with mutations, which due to genomic instability, can be perpetually produced during tumor progression, thereby preventing immune tolerance. Even if the tumors are edited by the immune system, a plethora of new neoantigen epitopes are continuously generated, ensuring that ICB can elicit an effective T-cell response against neoantigens. Furthermore, unlike neoantigens that can directly stimulate immune response, the activation of TAA-reactive immune responses may require additional costimulatory signals to mount an effective immunity against TAAs.

Taken together, these results support the association of TAA burden with ICB response in TMB-low and PD-L 1 -negative patients that otherwise would have a low costeffectiveness and low response to adverse effect ratio. These results suggest three possible prerequisites for mounting TAA-reactive immune response in established tumors: 1) a high- TAA burden in tumor, 2) a non-exhausted tumor immune context, 3) the presence of costimulatory signal induced by immune-stimulating molecular patterns. These patterns may include double-strand breaks due to homologous recombination deficiency or platinum-based therapies, somatic mutations, and the administration of BCG, etc. These results also demonstrate that TAA burden can be used as a biomarker to predict whether a mammal (e.g., a human) having cancer (e.g., urothelial carcinoma and head and neck squamous carcinoma with low T cell exhaustion state) is likely to respond to one or more immune checkpoint inhibitors.

Example 3: Datasets used in Example 2.

Table 3.

Example 4: List of CT A genes used in Example 2.

Table 4.

Example 5: Assessing Cancer for Responsiveness to ICB

A tissue sample containing one or more cancer cells is obtained from a human having cancer. The obtained sample is examined for the IGR burden and/or the TAA burden.

If the sample includes a high IGR burden, then the cancer is classified as being responsive to ICB (e.g., administration of one or more immune checkpoint inhibitors).

If the sample lacks a high IGR burden and lacks a high TAA burden, then the cancer is classified as not being responsive to ICB (e.g., administration of one or more immune checkpoint inhibitors).

Example 6: Treating cancer

If the sample includes a high IGR burden and/or a high TAA burden, then the human is administered or instructed to self-administer one or more immune checkpoint inhibitors. Once administered to the human, the one or more immune checkpoint inhibitors can reduce the number of cancer cells present in the human.

Example 7: Treating cancer

If the sample lacks a high IGR burden and lacks a high TAA burden, then the human is administered one or more (e.g., one, two, three, four, five, or more) alternative cancer treatments (e.g., one or more cancer treatments that do not include any immune checkpoint inhibitors). The alternative cancer treatment can reduce the number of cancer cells present in the human.

Example 8: Panel of tumor associated antigens predicted by HEPA analysis. Known CT antigens are annotated in the table.

Example 9: Immune gene sets used in the TAA burden study

GSE9650 EFFECTOR VS EXHAUSTED CD8 T CELL UP:

LAMT0R5, MRPL34, SEMA4A, CDKN2D, SLC1A5, LYSMD1, ELAVL1, RSU1, SRP68, FAM1 17A, SRPK1, SEPTIN6, MLX, TSPAN31, RPP25L, CDC37, GRAMD2B, TIAM1, FAM89B, PALD1, ETS1, RNPS1, SSNA1, ANXA6, TWF2, BRAP, ICAM2, PSMD13, CCR2, EIF2B5, CCNDBP1, DAP, AP1M1, ITGAX, LDAH, EIF3B, UBQLN1, KLRC1, PRPSAP1, IL18RAP, HSD11B1, KLRK1, DBI, PLAC8, LGALS9B, MKNK2, HCLS1, LSM1, FCGR2B, GLIPR2, SLC66A2, ANAPC16, ACP5, PRIM2, LEF1, NSMCE1, RABGGTA, CORO1B, TIMM44, CAB39L, B4GALT1, ZFYVE19, ITGAL, ITGB7, DCPS, PSMB3, PIM2, HMCES, PHTF1, RORA, ISYNA1, ZIK1, COMMD7, WTAP, PKP3, ENTPD4, PRKCH, UBE2Z, PSMB2, STB, TUBA3C, YIPF3, WARSI, REPSI, KCNJ8, ANXA1, JKAMP, PTPN6, LRWD1, ATP5PO, ARFRP1, RAD 17, SWAP70, MFNG, IL17RA, HIKESHI, SELENOH, RPN1, PPIF, CTSD, USP22, DHRS1, HNRNPAB, ATP6V0C, GCAT, NDUFB6, KLRD1, SMIM20, BLVRA, PRKAG1, GIT1, HIPK1, ATP6V0B, SATB1, C8orfi3, RACGAP1, LY6H, TBCB, RNF14, DEAF1, KLRG1, TMEM208, TMEM223, HADHB, SETD6, LAMT0R4, GDAP2, TMEM147, CTSA, ABCA2, ORC5, MIEN1, GSTT2, AK3, RNF167, GPC1, TXNL4A, DGKA, PDIA6, IFITM10, RPS6KA4, XPNPEP1, BCKDK, CALU, ARL4C, ANAPC5, PNPO, DNAJB1, AURKAIP1, EIF6, PIK3CD, CIB1, RMCI, TMEM45A, EIF2S1, DPEP1, EFTUD2, GOLM1, UBE2H, EIF4A3, DUS1L, EBNA1BP2, SCP2, CSRP3, STK38, CHFR, DNM1, FGFR3, CMAS, BNIP3L, MTSS2, DDX41, FHL2, DYM, USP5, DCTN5, KLK8, BSCL2, EIF3L, PCGF5, FEZ2, PTTG1, TMC6, PPIB, SMPD1, HERPUD1, PLD3, MTCH1, AXIN1, EIF2B4, KIAA2013, MACROH2A1, VAV1, SNX1, DPM2, CPT2, CMTM7, LSM4, MTMR1, and MBP.

GSE9650 EFFECTOR VS EXHAUSTED CD8 TCELL DOWN:

DUSP6, MAGEL2, F2RL1, AFP, MY06, PTPN12, CRISP2, EFNB3, LCLAT1, CFH, ANXA3, EGR2, DDIT4, RGS16, TNFRSF9, DOCK7, PLEKHA1, ZFP28, GCM2, HAO2, CKMT2, NKIRAS1, GPLD1, MRPS2, IMMT, CD244, ZNF821, CLDN11, RIN2, RGS10, SLX9, EPCAM, OVGP1, SCANT) 1, SCN7A, TLR7, C16orf72, NRK, GSTO1, COCH, NSDHL, MITF, CXCL13, PAWR, MTRF1L, SCAMPI, CD22, SYT1, TERFI, SCRG1, RFLNB, EOMES, ILIA, H3C7, YAP1, ACADVL, NQO1, CADM1, P2RX4, INCAI, HMGA2, PHLDB2, SLC30A1, GMCL1, PNRC1, PTPRJ, WFS1, LIN9, SMIM4, ADAM7, GDNF, MAP2, PER2, SPOCK2, POLR2C, MDN1, IRF6, CXCL14, CELF4, ATP5MF, ZNF239, AOPEP, ZNRF1, CCRL2, NDUFA13, CANX, RPRD1B, SIX1, KIAA1217, NELFE, TMEM150A, SLC12A2, NOTCH4, PLA2G10, EFS, RBM15, APP, SUB1, PHLDA3, GDAP1, HINFP, PAX1, KANSL2, WLS, TUG1, ABCG1, PBDC1, MAP1S, AGAP1, GATA2, VAMP7, HLA-DMA, ENPP2, SPRED2, TRPC1, CYP2A6, EVI5, EIF2AK2, POU2F1, CYP4V2, ACSL1, NCAN, COL19A1, SCN1A, HTRA2, SPOUT1, POLR1B, H19, TM4SF1, TM2D3, STRA6, NFIB, MSX1, NR4A2, GNAO1, DPP7, TGM2, CHL1, PRXL2A, GSTM3, TM2D1, GCSAM, METAP2, HOXC6, SPP1, KCTD12, GABRR2, TAPBP, FRK, CARMI, LHCGR, PIK3C2G, DFFA, IFIH1, CFHR2, ZRANB1, NEFH, MRPL48, PTK6, IRS1, ZNF35, BET1, ATP2A2, SSBP2, FGF6, FHL1, Hl-4, CELA1, PTGER2, CP A3, ADGRG1, TNFRSF4, VCAM1, CLCA1, SMAD1, KCNAB1, TBX15, NKX2-2, TWSG1, RHAG, TCF4, GTF3C4, AARD, NEURODI, PCLO, GPM6B, AUH, REXO5, CPSF2, SLC7A11, NAP1L2, COPRS, SLC6A4, IGF1R, AHR, ERCC5, RXYLT1, EXOSC8, CSF1, and MCAM.

T effector signature:

GZMA, GZMB, PRF1, EOMES, IFNG, TNF, CXCL9, CXCL10, CD8A, CD4, FOXP3, ICOS, and CTLA4 Pan-Fibroblast-TBRS:

ACTA2, ACTG2, ADAM12, ADAM19, CNN1, COL4A1, CCN2, CTPS1, RFLNB, FSTL3, HSPB1, IGFBP3, PXDC1, SEMA7A, SH3PXD2A, TAGLN, TGFBI, TNS1, and TPM1.

OTHER EMBODIMENTS

It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

WHAT IS CLAIMED IS:

1. A method for assessing a mammal having cancer, wherein said method comprises:

(a) determining that a sample obtained from said mammal and comprising cancer cells comprises a high intergenic rearrangement (IGR) burden or a high tumor associated antigen (TAA) burden; and

(b) classifying said cancer as being likely to respond to an immune checkpoint inhibitor based at least in part on said high IGR burden or said high TAA burden.

2. The method of claim 1, wherein said mammal is a human.

3. The method of any one of claims 1-2, wherein said immune checkpoint inhibitor is selected from the group consisting of an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti- CTL4A antibody, and an anti-LAG-3 antibody.

4. The method of any one of claims 1-2 wherein said immune checkpoint inhibitor is selected from the group consisting of pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, and relatlimab.

5. The method of any one of claims 1-2, wherein said immune checkpoint inhibitor is selected from the group consisting of BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK5O3, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, and DZNep.

6. The method of any one of claims 1-5, wherein said cancer comprises a solid tumor.

7. The method of any one of claims 1-6, wherein said cancer was previously exposed to platinum.

8. The method of any one of claims 1-7, wherein said cancer lacks a high tumor mutational burden (TMB).

9. The method of claim 8, wherein said high TMB comprises greater than three mutations per million bases.

10. The method of any one of claims 1-6, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

11. The method of any one of claims 1-6, wherein said method comprises determining that said sample comprises said high IGR burden.

12. The method of any one of claims 1-6, wherein said method comprises determining that said sample comprises said high TAA burden.

13. A method for assessing a mammal having cancer, wherein said method comprises:

(a) determining that a sample obtained from said mammal and comprising cancer cells comprises the absence of a high IGR burden and the absence of a high TAA burden; and

(b) classifying said cancer as not being likely to respond to an immune checkpoint inhibitor based at least in part on said absence of said high IGR burden and said absence of said high TAA burden.

14. The method of claim 13, wherein said mammal is a human.

15. The method of any one of claims 13-14, wherein said immune checkpoint inhibitor is selected from the group consisting of an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti- CTL4A antibody, and an anti-LAG-3 antibody.

16. The method of any one of claims 13-14, wherein said immune checkpoint inhibitor is selected from the group consisting of pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, and relatlimab.

17. The method of any one of claims 13-14, wherein said immune checkpoint inhibitor is selected from the group consisting of BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK5O3, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, and DZNep.

18. The method of any one of claims 13-17, wherein said cancer comprises a solid tumor.

19. The method of any one of claims 13-17, wherein said cancer was previously exposed to platinum.

20. The method of any one of claims 13-19, wherein said cancer lacks a high TMB.

21. The method of claim 20, wherein said high TMB comprises greater than three mutations per million bases.

22. The method of any one of claims 13-21, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

23. A method for selecting a treatment for a mammal having cancer, wherein said method comprises:

(a) determining that a sample obtained from said mammal and comprising cancer cells comprises a high IGR burden or a high TAA burden; and

(b) selecting an immune checkpoint inhibitor as a treatment for said cancer.

24. The method of claim 23, wherein said mammal is a human.

25. The method of any one of claims 23-24, wherein said immune checkpoint inhibitor is selected from the group consisting of an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti- CTL4A antibody, and an anti-LAG-3 antibody.

26. The method of any one of claims 23-24 wherein said immune checkpoint inhibitor is selected from the group consisting of pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, and relatlimab.

27. The method of any one of claims 23-24, wherein said immune checkpoint inhibitor is selected from the group consisting of BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK5O3, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, and DZNep.

28. The method of any one of claims 23-27, wherein said cancer comprises a solid tumor.

29. The method of any one of claims 23-28, wherein said cancer was previously exposed to platinum.

30. The method of any one of claims 23-29, wherein said cancer lacks a high TMB.

31. The method of claim 30, wherein said high TMB comprises greater than three mutations per million bases.

32. The method of any one of claims 23-29, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

33. The method of any one of claims 23-29, wherein said method comprises determining that said sample comprises said high IGR burden.

34. The method of any one of claims 23-29, wherein said method comprises determining that said sample comprises said high TAA burden.

35. A method for selecting a treatment for a mammal having cancer, wherein said method comprises:

(b) selecting a cancer treatment other than an immune checkpoint inhibitor as a treatment for said cancer.

36. The method of claim 35, wherein said mammal is a human.

37. The method of any one of claims 35-36, wherein said cancer comprises a solid tumor.

38. The method of any one of claims 35-37, wherein said cancer treatment comprises performing surgery.

39. The method of any one of claims 35-37, wherein said cancer treatment comprises radiation therapy.

40. The method of any one of claims 35-37, wherein said cancer treatment comprises administering, to said mammal, an anti-cancer agent selected from the group consisting of a chemotherapy and an angiogenesis inhibitor.

41. The method of any one of claims 35-40, wherein said cancer was previously exposed to platinum.

42. The method of any one of claims 35-41, wherein said cancer lacks a high TMB.

43. The method of claim 42, wherein said high TMB comprises greater than three mutations per million bases.

44. The method of any one of claims 35-41, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

45. A method for treating for a mammal having cancer, wherein said method comprises:

(a) determining that a sample obtained from said mammal and comprising cancer cells comprises the presence of a high IGR burden or the presence of a high TAA burden; and

(b) administering an immune checkpoint inhibitor to said mammal.

46. The method of claim 45, wherein said mammal is a human.

47. The method of any one of claims 45-46, wherein said immune checkpoint inhibitor is selected from the group consisting of an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti- CTL4A antibody, and an anti-LAG-3 antibody.

48. The method of any one of claims 45-46, wherein said immune checkpoint inhibitor is selected from the group consisting of pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, and relatlimab.

49. The method of any one of claims 45-46, wherein said immune checkpoint inhibitor is selected from the group consisting of BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK5O3, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, and DZNep.

50. The method of any one of claims 45-49, wherein said cancer comprises a solid tumor.

51. The method of any one of claims 45-50, wherein said cancer was previously exposed to platinum.

52. The method of any one of claims 45-51, wherein said cancer lacks a high TMB.

53. The method of claim 52, wherein said high TMB comprises greater than three mutations per million bases.

54. The method of any one of claims 45-51, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

55. The method of any one of claims 45-51, wherein said method comprises determining that said sample comprises said high IGR burden.

56. The method of any one of claims 45-51, wherein said method comprises determining that said sample comprises said high TAA burden.

57. A method for treating cancer, wherein said method comprises administering an immune checkpoint inhibitor to a mammal identified as having cancer cells comprising the presence of a high IGR burden or the presence of a high TAA burden, thereby treating cancer within said mammal.

58. The method of claim 57, wherein said mammal is a human.

59. The method of any one of claims 57-58, wherein said immune checkpoint inhibitor is selected from the group consisting of an anti-PD-1 antibody, an anti-PD-Ll antibody, an anti- CTL4A antibody, and an anti-LAG-3 antibody.

60. The method of any one of claims 57-58, wherein said immune checkpoint inhibitor is selected from the group consisting of pembrolizumab, nivolumab, cemiplimab, ipilimumab, tremelimumab, durvalumab, dostarlimab, avelumab, atezolizumab, and relatlimab.

61. The method of any one of claims 57-58, wherein said immune checkpoint inhibitor is selected from the group consisting of BMS-8, BMS-37, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, SB415286, vorinostat, decitabine, entitostat, JQ1, BET151, GSK5O3, panobinostat, ACY-241, azacytidine, DB36, DB71, DB15, CVN, MGCD0103, SNDX-275, IMP32, BMS986016, TSR-022, Sym023, ATIK2a, and DZNep.

62. The method of any one of claims 57-61, wherein said cancer comprises a solid tumor.

63. The method of any one of claims 57-62, wherein said cancer was previously exposed to platinum.

64. The method of any one of claims 57-63, wherein said cancer lacks a high TMB.

65. The method of claim 64, wherein said high TMB comprises greater than three mutations per million bases.

66. The method of any one of claims 57-63, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

67. The method of any one of claims 57-63, wherein said method comprises determining that said sample comprises said high IGR burden.

68. The method of any one of claims 57-63, wherein said method comprises determining that said sample comprises said high TAA burden.

69. A method for treating a mammal having cancer, wherein said method comprises:

(b) administering a cancer treatment to said mammal, wherein said cancer treatment is not an immune checkpoint inhibitor.

70. The method of claim 69, wherein said mammal is a human.

71. The method of any one of claims 69-70, wherein said cancer comprises a solid tumor.

72. The method of any one of claims 69-71, wherein said cancer treatment comprises performing surgery.

73. The method of any one of claims 69-71, wherein said cancer treatment comprises radiation therapy.

74. The method of any one of claims 69-71, wherein said cancer treatment comprises administering an anti-cancer agent selected from the group consisting of a chemotherapy and an angiogenesis inhibitor.

75. The method of any one of claims 69-74, wherein said cancer was previously exposed to platinum.

76. The method of any one of claims 69-75, wherein said cancer lacks a high TMB.

77. The method of claim 76, wherein said high TMB comprises greater than three mutations per million bases.

78. The method of any one of claims 69-75, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

79. A method for treating cancer, wherein said method comprises administering a cancer treatment that is not an immune checkpoint inhibitor to a mammal identified as having cancer cells comprising the absence of a high IGR burden and the absence of a high TAA burden.

80. The method of claim 79, wherein said mammal is a human.

81. The method of any one of claims 79-80, wherein said cancer comprises a solid tumor.

82. The method of any one of claims 79-81, wherein said cancer treatment comprises performing surgery.

83. The method of any one of claims 79-81, wherein said cancer treatment comprises radiation therapy.

84. The method of any one of claims 79-81, wherein said cancer treatment comprises administering an anti-cancer agent selected from the group consisting of a chemotherapy and an angiogenesis inhibitor.

85. The method of any one of claims 79-84, wherein said cancer was previously exposed to platinum.

86. The method of any one of claims 79-85, wherein said cancer lacks a high TMB.

87. The method of claim 86, wherein said high TMB comprises greater than three mutations per million bases.

88. The method of any one of claims 79-85, wherein a population of CD8⁺ T cells within a tumor microenvironment of said cancer comprises a low incidence of T cell exhaustion.

89. The method of any one of claims 1-88, wherein said high TAA burden is assessed across at least 50 percent of the genes listed in Example 8.

90. The method of any one of claims 1-88, wherein said high TAA burden is assessed across at least 75 percent of the genes listed in Example 8.

91. The method of any one of claims 1-88, wherein said high TAA burden is assessed across at least 80 percent of the genes listed in Example 8.

92. The method of any one of claims 1-88, wherein said high TAA burden is assessed across at least 90 percent of the genes listed in Example 8.

93. The method of any one of claims 1-88, wherein said high TAA burden is assessed across at least 100 percent of the genes listed in Example 8.