Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Immune biomarkers in breast cancer pathology specimens : characterization and clinical implications Burugu, Samantha 2019

You are currently on our download blacklist and unable to view media. You will be unbanned within an hour.
To un-ban yourself please visit the following link and solve the reCAPTCHA, we will then redirect you back here.

Item Metadata


24-ubc_2019_september_burugu_samantha.pdf [ 4.97MB ]
JSON: 24-1.0379406.json
JSON-LD: 24-1.0379406-ld.json
RDF/XML (Pretty): 24-1.0379406-rdf.xml
RDF/JSON: 24-1.0379406-rdf.json
Turtle: 24-1.0379406-turtle.txt
N-Triples: 24-1.0379406-rdf-ntriples.txt
Original Record: 24-1.0379406-source.json
Full Text

Full Text

IMMUNE BIOMARKERS IN BREAST CANCER PATHOLOGY SPECIMENS: CHARACTERIZATION AND CLINICAL IMPLICATIONS by  Samantha Burugu   B.Sc., University of Ottawa, 2012 M.Sc., McGill University, 2015  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Pathology and Laboratory Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  June 2019  © Samantha Burugu, 2019 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Immune biomarkers in breast cancer pathology specimens: Characterization and clinical implications  submitted by Samantha Burugu in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Pathology and Laboratory Medicine   Examining Committee: Torsten O. Nielsen Supervisor  John J. Spinelli Supervisory Committee Member  Peter C.V. Black University Examiner Kenneth W. Harder University Examiner  Additional Supervisory Committee Members: Kevin L. Bennewith Supervisory Committee Member  Nickolas Myles Supervisory Committee Member  Brad H. Nelson Supervisory Committee Member  iii  Abstract Immunotherapy is dramatically changing the landscape of cancer treatment and is becoming incorporated into the standard of care for some tumor types. Until recently, breast cancer has not been generally considered particularly immunogenic. However, breast cancer is a heterogeneous disease and increasing evidence suggests that patients with basal-like breast cancer, an aggressive subtype lacking targeted therapy options, may be amenable to immunotherapy.  My research goals have included the investigation and clinical characterization of two emerging targetable immune checkpoint biomarkers: lymphocyte-activation gene 3 (LAG-3) and the T-cell Immunoglobulin and Mucin domain-containing molecule 3 (TIM-3), by applying immunohistochemistry to a well-annotated tissue microarray cohort of 3,992 breast cancers. As an additional research goal, I evaluated a novel in situ multiplex biomarker assessment method (Nanostring-based digital spatial profiling-DSP) for its compatibility with breast cancer tissue microarrays, to generate immune profiles from patient surgical specimens.  I report that the expression of LAG-3 or TIM-3 on intra-epithelial tumor-infiltrating lymphocytes (iTILs) was observed in a minority of cases (11%) in the whole cohort, but was significantly enriched in basal-like breast cancers (33% and 28% of basal-like breast cancers being infiltrated with LAG-3+ and TIM-3+iTILs, respectively). Furthermore, I found that LAG-3+iTILs and TIM-3+iTILs were present in breast cancers co-infiltrated with established immunotherapy targets (program cell death-1/PD-1 and its ligand, PD-L1). In multivariate analyses, LAG-3+iTILs or TIM-3+iTILs were independent favorable prognostic factors in breast cancer patients. In the last part of the thesis, I profiled the tumor immune microenvironment of two basal-like-enriched breast cancer cohorts, quantifying the expression of 31 immuno-oncology biomarkers using DSP. I then validated the digital counts for CD8 and PD1 by iv  comparing with immunohistochemistry, and CD45 digital counts by comparing with hematoxylin & eosin-stained stromal TILs counts. Lastly, I identified a 4-biomarker signature indicative of a pre-existing immunity in breast cancer patients.   The body of work presented here may help guide the selection of breast cancer patients for clinical trial evaluation of emerging immunotherapy agents. Furthermore, I show that digital spatial profiling technology can efficiently and quantitatively profile immune expression on breast cancer patient specimens using only a tiny fraction from precious tumor samples.   v  Lay Summary Immune checkpoint inhibitors are drugs that restore patients’ own immune system to fight cancer and have revolutionized cancer treatment. However, in most cancers (including breast cancer), the success of current immunotherapy targets has been limited to few patients for reasons that are still unknown, so new drugs for new targets are being investigated. Using breast tumor samples from more than 3,000 patients, I found that two emerging targets for immunotherapy (LAG-3 and TIM-3) were particularly enriched in patients with an aggressive breast cancer subtype called basal-like, providing the largest evidence yet about the presence of LAG-3 and TIM-3 in breast cancer. Furthermore, using a novel technology that identifies multiple immune populations in a patient’s tumor, I identified 4 key immune populations that might help to select patients amenable to immune-modulating therapies. The work presented here can inform the prioritization and design of clinical trials for immune therapies in breast cancer.        vi  Preface Portions of Chapter 1 are modified versions based on two publications:   Sections 1.2 (Immune system in cancer and all subsections), (CTLA-4) and (PD-1/PD-L1) are based on a published review: Burugu S, Asleh-Aburaya K, Nielsen TO. Immune infiltrates in the breast cancer microenvironment: detection, characterization and clinical implication. Breast Cancer. 2017 Jan;24(1):3-15. doi: 10.1007/s12282-016-0698-z. Epub 2016 May 2. Review. PubMed PMID: 27138387.  Co-author Asleh-Aburaya K and I were responsible for writing the manuscript under the supervision of Dr. Nielsen. The content in the sections described in this thesis chapter was originally written by me, with the exceptions of the sections describing the prognostic value of immune biomarkers (which are modified versions of sections initially written by Asleh-Aburaya K).  Section (LAG-3) and section (TIM-3) are based on a published review:  Burugu S, Dancsok AR, Nielsen TO. Emerging targets in cancer immunotherapy. Semin Cancer Biol. 2018 Oct;52(Pt 2):39-52. doi: 10.1016/j.semcancer.2017.10.001. Epub 2017 Oct 5. Review. PubMed PMID: 28987965.  Co-author Dancsok AR and I equally contributed to the entire manuscript writeup under the supervision of Dr. Nielsen. Section was written by me and Section this thesis is a modified section originally drafted by co-author Dancsok AR.   Versions of Chapter 2 and Chapter 3 have been published:  vii  Burugu S, Gao D, Leung S, Chia SK, Nielsen TO. LAG-3+ tumor infiltrating lymphocytes in breast cancer: clinical correlates and association with PD-1/PD-L1+ tumors. Ann Oncol. 2017 Dec 1;28(12):2977-2984. doi: 10.1093/annonc/mdx557. PubMed PMID: 29045526.  Burugu S, Gao D, Leung S, Chia SK, Nielsen TO. TIM-3 expression in breast cancer. Oncoimmunology. 2018 Aug 23;7(11):e1502128. doi:10.1080/2162402X.2018.1502128. eCollection 2018. PubMed PMID: 30377566  For both publications, I was responsible for drafting the manuscript, preparing the figures and tables and addressing comments from reviewers. I optimized LAG-3 and TIM-3 antibody immunohistochemistry staining conditions and performed the initial statistical analyses. Scoring of the immunohistochemically-stained breast cancer tissue microarrays was performed by Dr. Dongxia Gao. I coordinated the scoring system for the immune biomarkers. Samuel Leung conducted the final statistical analyses and helped in addressing comments from reviewers. Dr. Stephen Chia was involved in the manuscript writeup and addressing comments from reviewers. The experimental design for both studies, manuscript writeup and addressing comments from reviewers were supervised by Dr. Torsten Nielsen.    Chapter 4 is unpublished material and part of a manuscript in preparation: Staining of the two breast cancer tissue microarrays by the digital spatial profiling was performed by Yang Liang and JingJing Gong at the Nanostring headquarter in Seattle, US after the Nielsen lab was selected for a technology access program grant. I coordinated the fluorescent markers included for visualization and for the selection of the regions of interest to be analyzed by digital spatial profiling. I prepared the data to send to a biostatistician (Xing Ren) at Nanostring who conducted the unsupervised hierarchical cluster analyses. All additional analyses viii  were conducted by me. Early results were selected for a poster presentation at the British Columbia Cancer Summit held in Vancouver on November 23, 2018.      Chapter 5 is unpublished material written by myself   All human research studies performed in the thesis were conducted under the human ethics approval of the UBC Research Ethics Board: Certificate number: H17-01385 Immune biomarkers in breast cancer.        ix  Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ......................................................................................................................... ix List of Tables .............................................................................................................................. xiv List of Figures ............................................................................................................................. xvi List of Abbreviations ............................................................................................................... xviii Acknowledgements ......................................................................................................................xx Dedication ................................................................................................................................... xxi Chapter 1: Introduction ................................................................................................................1 1.1 Breast cancer ................................................................................................................... 1 1.1.1 Heterogeneity of breast cancers .................................................................................. 1 1.1.2 Evolution of breast cancer treatments ......................................................................... 3 Chemotherapy ..................................................................................................... 3 Targeted therapy ................................................................................................. 4 1.2 Immune system and cancer ............................................................................................. 4 1.2.1 Innate immunity .......................................................................................................... 6 Tumor-associated neutrophils ............................................................................. 6 Macrophages ....................................................................................................... 7 1.2.2 Adaptive immunity ..................................................................................................... 8 Tumor-infiltrating lymphocytes in breast cancer ................................................ 8 x B cells................................................................................................................ 10 T cells ................................................................................................................ 11 CD8+ TILs .................................................................................................. 11 CD4+ T cells ............................................................................................... 13 1.3 Cancer immunotherapy: A new strategy in breast cancer treatment ............................ 15 1.3.1 Immune augmentation .............................................................................................. 15 Direct effects ..................................................................................................... 15 Indirect effects .................................................................................................. 16 1.3.2 Immune restoration via immune checkpoint receptor inhibition .............................. 17 CTLA-4 ............................................................................................................. 18 PD-1/PD-L1 ...................................................................................................... 19 1.3.3 Beyond CTLA-4 and PD-1/PD-L1 ........................................................................... 21 LAG-3 ............................................................................................................... 21 TIM-3 ................................................................................................................ 22 1.4 Rationale for thesis and research objectives ................................................................. 23 Chapter 2: LAG-3+ tumor infiltrating lymphocytes in breast cancer: clinical correlates and association with PD-1/PD-L1+ tumors .......................................................................................32 2.1 Introduction ................................................................................................................... 33 2.2 Materials and methods .................................................................................................. 34 2.2.1 Study cohorts ............................................................................................................ 34 2.2.2 Immunohistochemistry ............................................................................................. 34 2.2.3 LAG-3, PD-1 and PD-L1 scoring ............................................................................. 35 2.2.4 Statistical Analysis .................................................................................................... 36 xi  2.3 Results ........................................................................................................................... 37 2.3.1 Selection of LAG-3+TILs cutoffs on the training set ............................................... 37 2.3.2 LAG-3+iTILs are enriched in ER negative subtypes and associated with improved survival .................................................................................................................................. 37 2.3.3 LAG-3+iTILs are strongly associated with PD-L1/PD-1+ tumors .......................... 38 2.3.4 Exploratory survival analyses of concurrent tumor infiltration with LAG-3+ and CD8+T cells .......................................................................................................................... 39 2.4 Discussion ..................................................................................................................... 39 Chapter 3: TIM-3 expression in breast cancer .........................................................................56 3.1 Introduction ................................................................................................................... 56 3.2 Material and methods .................................................................................................... 58 3.2.1 Study cohorts ............................................................................................................ 58 3.2.2 Immunohistochemistry and scoring .......................................................................... 59 3.2.3 Statistical analysis ..................................................................................................... 60 3.3 Results ........................................................................................................................... 61 3.3.1 Distribution of TIM-3+TILs in breast cancers ......................................................... 61 3.3.2 The presence of TIM-3+iTILs in breast cancer is associated with unfavorable clinico-pathological factors ................................................................................................... 62 3.3.3 TIM-3+iTILs correlate with the presence of other immune checkpoint markers (LAG-3, PD-1, PD-L1) and overall H&E sTILs .................................................................. 63 3.3.4 TIM-3+iTILs are associated with good prognosis in early breast cancer ................. 63 3.4 Discussion ..................................................................................................................... 64 xii  Chapter 4: Quantitative in situ multiplex immune profiling of breast cancer patients using digital spatial profiling technology .............................................................................................87 4.1 Introduction ................................................................................................................... 88 4.2 Material and methods .................................................................................................... 89 4.2.1 Study cohort .............................................................................................................. 89 4.2.2 Digital spatial profiling ............................................................................................. 90 4.2.3 Immunohistochemistry and scoring .......................................................................... 91 4.2.4 Statistical analysis ..................................................................................................... 91 4.3 Results ........................................................................................................................... 92 4.3.1 Overview of DSP digital counts of the 31 biomarkers ............................................. 92 4.3.2 Validation of CD8 and PD1 DSP counts by immunohistochemistry and correlations with tumor infiltrating lymphocyte counts ........................................................................... 93 4.3.3 Distinct breast cancer immune expression profiles are illustrated by DSP .............. 94 4.3.4 Identification of a set of biomarkers most associated with immune-enriched profiles in both cohorts....................................................................................................................... 95 4.4 Discussion ..................................................................................................................... 97 4.5 Proposal to apply the Nanostring digital spatial profiling technology to clinical trials material ................................................................................................................................... 102 4.5.1 Hypothesis: ............................................................................................................. 103 4.5.2 Materials and methods: ........................................................................................... 103 4.5.3 Statistical design: .................................................................................................... 104 4.5.4 Significance of research: ......................................................................................... 106 Chapter 5: Overall conclusions and future directions............................................................121 xiii  5.1 Summary of findings................................................................................................... 121 5.1.1 Perspectives for LAG-3-targeted agents ................................................................. 122 5.1.2 Perspectives for TIM-3-targeted agents .................................................................. 123 5.1.3 Digital Spatial profiling .......................................................................................... 124 5.2 Significance of the research and perspectives for the future of breast cancer immunotherapy ....................................................................................................................... 125 Bibliography ...............................................................................................................................127 Appendix A .................................................................................................................................152 Appendix B .................................................................................................................................154 Appendix C .................................................................................................................................155 Appendix D .................................................................................................................................156  xiv  List of Tables Table 1.1 Description of breast cancer clinical intrinsic subtypes................................................ 30 Table 1.2 Breast cancer immunotherapy clinical trials ................................................................. 31 Table 2.1 Association of LAG-3+intra-epithelial and stromal tumor-infiltrating lymphocytes with clinicopathological parameters in the training set ........................................................................ 50 Table 2.2 Association of LAG-3+ intra-epithelial tumor-infiltrating lymphocytes and clinicopathological parameters. .................................................................................................... 51 Table 2.3 Differences in survival by LAG-3 expression and breast cancer subtypes, as assessed by interaction test .......................................................................................................................... 52 Table 2.4 Multivariate analyses of LAG-3+iTILs in the whole cohort and among ER- patients for breast cancer-specific survival. ..................................................................................................... 53 Table 2.5 Associations among immune response biomarkers in breast cancer patients .............. 54 Table 2.6 Association of hematoxylin and eosin-stained (H&E) stromal TILs with PD-1+, LAG-3+, and PD-1+/LAG-3+ concurrent infiltration. ........................................................................... 55 Table 3.1 Association of TIM-3+ iTILs with clinico-pathological parameters on the initial training cohort. .............................................................................................................................. 78 Table 3.2 TIM-3+ intra-epithelial tumor-infiltrating lymphocytes association with clinico-pathological parameters in breast cancer ...................................................................................... 79 Table 3.3 Association of TIM-3+ stromal tumor-infiltrating lymphocytes with clinico-pathological parameters. ............................................................................................................... 80 Table 3.4 Association of TIM-3+ intra-epithelial tumor-infiltrating lymphocytes with other immune biomarkers in breast cancer ............................................................................................ 81 xv  Table 3.5 Multivariate analyses of TIM-3+ intra-epithelial tumor-infiltrating lymphocytes in the whole cohort, among estrogen receptor negative and in basal-like patients for breast cancer-specific survival including H&E sTILs as a covariate.................................................................. 82 Table 3.6 Multivariate analyses of TIM-3+stromal tumor-infiltrating lymphocytes ................... 84 Table 3.7 Multivariate analyses of TIM-3/PD-1/LAG-3+iTILs among ER negative breast cancer patients for breast cancer-specific survival. .................................................................................. 86 Table 4.1 Cohort description....................................................................................................... 115 Table 4.2 Cohort A DSP immune expression profile associations with clinicopathological parameters* ................................................................................................................................. 116 Table 4.3 Cohort B DSP immune expression profile associations with clinicopathological parameters ................................................................................................................................... 117 Table 4.4 Identification of key biomarkers in breast tumors stratified by H&E sTILs levels from Cohort A and B. .......................................................................................................................... 118 Table 4.5 Assessment of the 4-biomarker signature for detecting immune-enriched breast tumors..................................................................................................................................................... 119 Table 4.6 List of biomarkers in the commercially available DSP immuno-oncology panel that will be used for the study on MA.5 clinical trial ........................................................................ 120  xvi  List of Figures Figure 1.1 Cancer-immune interaction ......................................................................................... 26 Figure 1.2 Assessment of tumor-infiltrating lymphocytes in the breast cancer microenvironment........................................................................................................................................................ 27 Figure 1.3 Immune regulation in the breast cancer microenvironment ........................................ 28 Figure 2.1 Representative immunohistochemical staining of LAG-3, PD-1, and PD-L1 in breast tumor tissue microarray cores ....................................................................................................... 44 Figure 2.2 LAG-3+iTILs association with breast cancer-specific survival. ................................. 45 Figure 2.3 Association with relapse-free survival for LAG-3+ intra-epithelial tumor-infiltrating lymphocytes. ................................................................................................................................. 46 Figure 2.4 Correlation between LAG-3, PD-1, PD-L1 and CD8 scores. ..................................... 47 Figure 2.5 Association between CD8+iTILs and the prognostic value of immune checkpoint markers. ......................................................................................................................................... 48 Figure 2.6 Association with breast cancer-specific survival for LAG-3+/PD-1+ intra-epithelial tumor-infiltrating lymphocytes in the presence or absence of CD8+ intra-epithelial tumor-infiltrating lymphocytes among ER negative patients. ................................................................. 49 Figure 3.1TIM-3 staining in breast cancer patients ...................................................................... 68 Figure 3.2 Distribution of TIM-3+ tumor-infiltrating lymphocytes in the whole (validation) cohort ............................................................................................................................................ 69 Figure 3.3 H&E sTIL correlation with immune biomarkers. ....................................................... 70 Figure 3.4 TIM-3+intra-epithelial tumor-infiltrating lymphocytes association with breast cancer-specific survival in the whole (validation) cohort and by breast cancer subtype. ........................ 71 xvii  Figure 3.5 Overall survival for TIM-3+intra-epithelial tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes. .................................................. 72 Figure 3.6 Relapse-free survival for TIM-3+ intra-epithelial tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes. ............................................ 73 Figure 3.7 TIM-3+ stromal tumor-infiltrating lymphocytes association with breast cancer-specific survival in the whole cohort and by breast cancer subtype. ............................................ 74 Figure 3.8 Overall survival for TIM-3+ stromal tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes. ............................................................. 75 Figure 3.9 Relapse-free survival for TIM-3+ stromal tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes. ............................................................. 76 Figure 3.10 Prognostic value of TIM-3, PD-1 and LAG-3+iTILs co-infiltration among ER negative breast cancer patients...................................................................................................... 77 Figure 4.1 Schematic of DSP analysis process on Cohort A and B TMAs ................................ 107 Figure 4.2 Core-to-core agreement on DSP measurements for biomarker expression counts ... 108 Figure 4.3 Distribution of DSP digital counts. ........................................................................... 109 Figure 4.4 Comparison of background variance and sample cellularity between Cohort A and Cohort B. ..................................................................................................................................... 110 Figure 4.5 Validation of digital spatial profiling counts by immunohistochemistry and H&E staining. ....................................................................................................................................... 111 Figure 4.6 Unsupervised hierarchical clustering of patients in Cohort A and B based on biomarker expression counts analyzed by DSP. ......................................................................... 113 Figure 4.7 Prognostic value of DSP immune profiles in breast cancer patients from Cohort A. 114  xviii  List of Abbreviations BCSS Breast cancer-specific survival BRCA Breast cancer gene CDK4/6 Cyclin-dependent kinase 4/6 CEACAM-1 Carcinoembryonic antigen-related cell adhesion molecule 1 CEF Cyclophosphamide-Epirubicin-5’Fluorouracil CK5/6 Cytokeratin 5/6 CMF Cyclophosphamide-Methotrexate-5’Fluorouracil CTLA4 Cytotoxic T lymphocyte-associated antigen 4 DAB  3,3’-diaminobenzidine EGFR Epidermal growth factor receptor 1 ER Estrogen receptor FOXP3 Forkhead box P3 GZMB Granzyme B H&E sTILs Hematoxylin and eosin-stained stromal tumor-infiltrating lymphocytes HER2 Human epidermal growth factor receptor 2 IFNγ Interferon gamma IHC Immunohistochemistry iTILs Intra-epithelial tumor-infiltrating lymphocytes LAG3 Lymphocyte-activation gene 3 LPBC Lymphocyte-predominant breast cancer MDSC Myeloid-derived suppressor cells xix  MHC Major histocompatibility complex PAM50 Prediction analysis of microarray 50 pCR Pathologic complete response PD1 Programmed cell-death 1 PDL1 Programmed cell-death ligand 1 PR Progesterone receptor PTEN Phosphatase and tensin homolog RECIST Response evaluation criteria in solid tumors ROI Region of interest sTILs Stromal tumor-infiltrating lymphocytes TAM Tumor-associated macrophages TIM3 T-cell Immunoglobulin and Mucin domain-containing molecule 3 TNBC Triple negative breast cancer TNFα Tumor necrosis factor alpha VISTA V-domain Ig suppressor of T cell activation  xx  Acknowledgements I would like to express my utmost gratitude to an incredible supervisor, Dr. Torsten Nielsen, who introduced me to the immuno-oncology field. For your willingness to actively listen to my ideas and share your knowledge, for your encouragement and insightful advices throughout my PhD, un très grand merci!  I would like to thank my supervisory committee (Drs. David Huntsman, Brad Nelson, John Spinelli, Nickolas Myles and Kevin Bennewith) for their guidance and helpful comments.   To my GPEC family (Christine Chow, Angela Cheng, Karama Asleh and Jenny Wang), I can’t thank you enough for all the support inside and outside the lab and to my fellow Nielsen lab mates (Dongxia Gao, Amanda Dancsok, Samuel Leung, Jamie Yu, Angela Goytain, Alvin Qiu and Neal Poulin), thank you for creating such as great work environment. In particular, Doris for your significant contributions in my thesis and great baby advice; Sam for helping me learn statistics in an easy way and always being available to answer to my not-so-quick questions. This thesis could not have been completed without the generous funding support from the Fonds de Recherche du Quebec en Santé and the UBC Faculty of Medicine Graduate award.   Special and heartfelt thanks to my husband, Emmanuel, for always believe in me, encouraging me to pursue a PhD and for giving me the gift of becoming a mother. To my mother, family and friends, thank you for your unconditional support throughout my degree.     xxi  Dedication  To Eugenie and Rose   1  Chapter 1: Introduction  1.1 Breast cancer  Breast cancer is the most common cancer diagnosed among Canadian women. According to the Canadian Cancer Society, breast cancer accounted for 25% of cancer newly diagnosed among women in 2017. Campaigns for early breast cancer detection and evolution of treatments led to high survival rates for breast cancer patients. 87% of Canadian women diagnosed with early breast cancer survive at least 5 years (Canadian Cancer Society). However, the survival rates dramatically decrease in patients who present or progress to metastatic breast cancer to 22% of chance for survival after 5 years. In the past 2 decades, molecular profile studies revealed breast cancer as not a single disease but an agglomeration of multiple diseases.   1.1.1 Heterogeneity of breast cancers  Different molecular features have been used to classify breast cancers including gene or protein expression, DNA aberrations, and miRNA profiles1-6. The seminal molecular study that profiled >8,000 genes of breast cancer patients using cDNA microarrays identified 4 distinct intrinsic breast cancer subtypes based on gene expression patterns1. These were: Luminal-like, basal-like, Erb-B2+ and normal-like.  Gene expression patterns that distinguished the intrinsic subtypes included expression of transcription factors such as the estrogen receptor for luminal-like breast cancers, expression of genes associated with basal epithelial cells for basal-like breast cancers and over-expression of ERBB2 gene (coding for human epidermal growth factor receptor 2 protein or HER2) for Erb-B2+ breast cancers 1. Breast cancer intrinsic subtypes, as established in this study, were validated in subsequent studies and refined into the 4 major molecular subtypes : Luminal A, Luminal B, 2  HER2-enriched (HER2E) and basal-like 3,4,7-9 (Table 1.1). These four subtypes are distinct and display different clinical outcomes and response to therapy; a finding that has been repeatedly recapitulated using different gene expression platforms and different independent clinical datasets.10-14  Another major breast cancer study further classified into 10 different subtypes termed integrative clusters by analyzing various molecular features such as aberrations in DNA copy number and gene expression2. The study showed that luminal A and luminal B breast cancer intrinsic subtypes can be further divided into 8 different integrative clusters. The prognosis varies greatly among the 10 integrative clusters but shows similarity with breast cancer intrinsic subtypes2,15,16.    In clinical practice, partly due to high costs of molecular assays, immunohistochemistry surrogates are used to identify the intrinsic subtypes10,17,18(Table 1.1): For luminal A: estrogen receptor (ER)  and/or progesterone receptor (PR) positivity, low levels of  Ki67, and HER2 negative; for luminal B: ER and/or PR positive, high levels of Ki67 and/or HER2+ by immunohistochemistry or fluorescent in situ hybridization;  for HER2E: ER and PR negative and HER2+19 and for basal-like: ER, PR and HER2 negative and sometimes referred to an improved definition that adds CK5/6+ and/or EGFR+ to the triple negative phenotype20. Among these breast cancer subtypes, the triple negative immunohistochemistry phenotype (TNBC) is the most distinct and identifies, by gene expression, a heterogeneous group of breast cancers of which approximatively 50-75% are of basal-like subtype 21-24.  3  1.1.2 Evolution of breast cancer treatments Chemotherapy The efficacy of chemotherapy in treating breast cancer has been shown since the establishment of the first chemotherapy regimen made of cyclophosphamide, methotrexate and 5’fluorouracil (CMF)25,26 . This combination is composed of a DNA-damaging alkylating agent (cyclophosphamide) and 2 antimetabolites: an inhibitor of folic acid synthesis (methotrexate) and a nucleoside analogue (5’-fluorouracil). Chemotherapy regimens have evolved and improved breast cancer patients outcomes in both neoadjuvant and adjuvant settings with a current backbone composed of anthracyclines and taxanes26. Anthracyclines are anti-tumor nucleic acid intercalating agent and taxanes have anti-mitotic activity. Breast cancer intrinsic subtypes respond differently to CMF, and basal-like breast cancers are more responsive to CMF14,27,28 . Anthracyclines appear to be particularly efficacious in the treatment of HER2E breast cancers in part due to alterations in the topoisomerase 2 gene present in the same chromosomal location as ERBB2 gene (and thus often co-amplified) 14,29-31. More recent chemotherapy agents are used to treat metastatic breast cancer and include additional antimetabolites/nucleoside analogues (capecitabine, gemcitabine) and DNA-crosslinking agents (platinum-based therapies) with clinical responses dependent on breast cancer stage and subtype13,32,33.    4 Targeted therapy  Molecular profiling of breast cancers led to personalized therapies with direct effects on the prognosis of patients. ER expression in Luminal A and B breast cancer intrinsic subtypes is a predictive biomarker of response to anti-ER compounds such as letrozole and tamoxifen34. Similarly, HER2E breast cancers are treated with anti-HER2 therapies such as trastuzumab35. In contrast, the majority of basal-like breast cancers do not have proven targeted therapy options partly due to the heterogeneity of the subtype 22,23,36. Only BRCA-mutated basal-like/triple negative breast cancers (accounting for about 15%  of triple negative breast cancers)37 have been shown to respond to poly(ADP-ribose) polymerase inhibitors, agents that target the DNA damage repair system through synthetic lethality, which led to recent US health authority approvals38,39. However, accumulating evidence suggests that basal-like breast cancers might be the most amenable breast cancer subtype to immune-modulating therapies.  1.2 Immune system and cancer  The potential of anti-tumor immunity has been known for centuries, owing to pioneers such as Paul Ehrlich who introduced the immune surveillance concept in the 1900s40 – a concept that has been reinvigorated in recent years, with the rise of immunotherapy, and relies on the interplay between cancer cells and the tumor immune microenvironment in opposing tumor development and progression. The cancer-immunity interaction is described as a 3-phase process: Elimination, Equilibrium and Escape (Figure 1.1) 41-43  The elimination phase begins when transformed or cancer cells are recognized by the immune system. The immune response that proceeds is mediated by collective actions of the innate and the adaptive immune system. The latter includes cytotoxic T cells that target cancer cells expressing various types of antigens44. The actions of the innate immune system are mostly 5  antigen-independent with some exceptions including antibody-dependent cellular cytotoxicity and antibody-dependent cellular phagocytosis which provide a link between the innate and adaptive immune system.  The immunogenicity of the tumor is shaped during the equilibrium phase. It is at this phase that highly immunogenic cancer clones are successfully eliminated by the immune system leaving cancer clones with low immunogenic potential to grow unchecked by the immune system. A recent study reports the occurrence of immune-editing in metastatic sites45. It is believed that all cancers go through cancer immune-editing, and only become clinically detectable at the escape phase.   The tumor immune microenvironment includes polymorphonuclear cells, macrophages, T cells, natural killer cells, dendritic cells and B cells. These components vary in frequency and composition across different types of cancers. For instance, melanoma, renal cell, lung and colorectal carcinomas are known to have more extensive immune infiltrates in overall comparison to other types of cancers, such as breast or prostate carcinomas46 . This difference in immune infiltrates has been partly attributed to the degree of mutational burden observed prominently for example in melanoma47,48. Genetic instability and high mutation rates lead to increased production of neo-epitopes that induce a polyphenotypic immune response and generate a chronically inflamed tumor microenvironment 47,49.  Basal-like breast cancers are associated with genetic instability (via copy number alterations) and a relatively high mutational burden (partly due to mutations in p53 and deficiencies in DNA repair pathways) that make them the most immunogenic of all breast cancer intrinsic subtypes2,50 6  The presence of important cell populations of the innate and adaptive immune system has been reported in breast cancer 51. While the cells of the adaptive immune system have been the major focus in breast cancer, innate immune cells are frequently found in tumor specimens. 1.2.1 Innate immunity  Innate immunity constitutes a heterogeneous family and is the first immune response to an insult such as cellular transformation52. Following recognition of the insult, innate immune cells mediate their actions through distinct mechanisms such as phagocytosis for neutrophils and macrophages, cellular cytotoxicity for natural killer cells and γδ T cells, and antigen presentation for dendritic cells and macrophages.  Only innate immune cells of myeloid origin with reported immunosuppressive activity are presented below. Tumor-associated neutrophils Neutrophils are primarily known for their major role in the clearance of micro-organisms at sites of infection. In cancer, cytokines secreted in the tumor microenvironment such as IL-8 help in the recruitment and replenishment of these short-lived cells53. In preclinical studies tumor-associated neutrophils have been shown to be immunosuppressive by reducing T cell proliferation in breast cancer mouse models54,55. The role of tumor-associated neutrophils in breast cancer patients is still unclear.  Myeloid-derived suppressor cells (MDSCs), a heterogeneous population of immature myeloid cells with immunosuppressive activity constitute a phenotypically – but not histologically – distinct subset of neutrophils56,57 . MDSCs are observed in the tumor microenvironment of different types of cancer (including breast cancer), and have perhaps best been characterized in mouse models58,59 . Their phenotype varies but common markers used to recognize human MDSCs by flow cytometry include at least a CD11b+CD33+ phenotype60. 7  Importantly, the presence of MDSCs in breast cancer patient tumors correlates with late-stage disease and poor prognosis61. Macrophages Tumor-associated macrophages (TAM) found in cancer specimens are divided into two main categories, namely classically-activated M1 and alternatively-activated M2 macrophages, based on the main biological processes in which they participate. TAMs with an M1 phenotype are associated with a pro-inflammatory microenvironment and production of type I cytokines including TNFα and nitric oxide synthase. M1 macrophages appear to be involved in anti-tumor immune reactions but lack a unique biomarker to identify them by immunohistochemistry in clinical biopsy specimens62.  In contrast, the M2 phenotype is associated with an immunosuppressive microenvironment, producing cytokines such as IL-4 and IL-10. M2 macrophages have a role in establishing an immune environment that is permissive to tumor growth and spread; biomarkers such as CD163 have been used to identify this population63. By using CD68 as a pan-marker that can identify both M1 and M2 macrophages, and CD163 as a more specific marker for M2 macrophages, several studies have implicated a role for M2 macrophages in breast cancer patients as both high CD68 and CD163 counts are associated with poor outcome63-66. The interaction between TAMs and breast cancer cells might also be tumor subtype dependent, because in vitro data shows that co-culture of TAMs with estrogen receptor positive/luminal breast cancer cell lines generates M1 phenotype TAMs, whereas an M2 phenotype is generated when macrophages are co-cultured in the presence of triple negative/basal breast cancer cell lines67.  8  Recent studies have shown that the immunosuppressive activity of TAMs can also be mediated by the expression of an immune checkpoint receptor called Programmed cell Death-Ligand 1 (PD-L1) that leads to inhibition of cytotoxic tumor-infiltrating lymphocytes68 . PD-L1 expression is in turn driven by the secretion of IFNγ by activated T cells, representing a negative feedback loop that highlights the importance of cross-talk between T lymphocytes and macrophages within the tumor immune microenvironment68,69. 1.2.2 Adaptive immunity Tumor-infiltrating lymphocytes in breast cancer  In breast cancer, TILs composed primarily of B and T cells are observed in about 80% of patients and vary greatly by breast cancer subtypes70,71. In addition, tertiary lymphoid structures that include B cells can be identified in formalin-fixed paraffin-embedded  breast tumor whole sections, particularly those taken from the interface been carcinoma and adjacent normal tissues72,73. Measurement of TILs is typically done on standard hematoxylin and eosin (H&E)- stained slides made from core biopsies or surgical excision specimens from breast cancer patients (Figure 1.2). Even though research results from biomarker-assessed TILs on tissue microarray cores vs. whole sections are concordant (in terms of associations with clinicopathological parameters and prognosis), and 4 micron thick whole sections still only represent a minuscule fraction of the total tumor specimen, the international TIL working group recommends that TIL assessment be done on whole sections to better capture the spatial distribution of TILs on H&E slides74. In a seminal study investigating the prognostic value of TILs in breast cancer, the scoring of H&E TILs was assessed irrespective of their localization in the tumor and was categorized by density75. In an effort to further its clinical implementation, an international TIL working group 9  has set recommendations to standardize the assessment process for this biomarker74. Stromal TILs (sTILs) are defined as the percentage of the stromal area infiltrated by TILs that are not in direct contact with carcinoma cells, whereas intraepithelial TILs (iTILs) are defined as the percentage of the tumor area occupied by TILs within tumor nests in direct contact with carcinoma cells74 (Figure. 1.2). Although there is a positive correlation between sTIL and iTIL scores, only H&E sTILs are recommended to be assessed for prognostic or predictive analyses due to the higher analytical reproducibility of sTIL scoring among different pathologists74,76. The prognostic and predictive values of H&E TILs in breast cancer vary by subtype as reported in numerous studies70. The role of H&E TILs as a prognostic marker has been evaluated in several large studies that consistently show that among patients with triple negative breast cancer, 10% increments in H&E TILs are associated with an improved clinical outcome, especially for the endpoint of relapse-free survival77-81.  H&E TILs as a predictive factor have been evaluated in breast cancer neoadjuvant and adjuvant settings in multiple studies81-83. In neoadjuvant settings, a recent study analyzing breast cancers pooled from 6 randomized neoadjuvant clinical trials reported increasing H&E TIL concentration is a predictive factor for pathologic complete response in all breast cancers (irrespective of subtype)78. In contrast, in the adjuvant setting, a correlation between presence of high levels of H&E TILs (≥ 60%, a cut-off sometimes used to define  lymphocyte-predominant breast cancers) and response to particular adjuvant chemotherapy agents or anti-HER2 agents has been shown82,84; however discrepant results have also been reported 83. This suggests that the predictive value of H&E TILs might be immune cell-type specific. 10 B cells In comparison to T lymphocytes, the role of B cells in the context of breast cancer has not been studied as comprehensively. Although B cells have an established primary role in antibody production, they express MHC class II proteins that enable them to present antigens to CD4+ T cells; therefore, they do play a role in modulating T cell mediated anti-tumor activity 85. Several markers can be used to identify B cells by immunohistochemistry on formalin-fixed paraffin-embedded tumor tissues, but most studies have used CD2051,71,86,87. In the context of breast cancer, few studies have assessed the specific role of B cells in patient tumors86-89. One large study looked at the density of CD20+ B cells by immunohistochemistry using tissue microarrays representing 1470 breast carcinomas86. Results showed that the presence of CD20+ B cells was significantly associated with high grade tumors, hormone receptor negativity, basal-like subtypes and improved survival in the ER negative subtypes – associations very similar to those reported for most T-cell markers.  Moreover, the presence of antibody-secreting plasma cells in breast tumors and its prognostic implications have been reported90. Plasma cells are morphologically distinct from other B cells and can be visualized on hematoxylin and eosin-stained slides90. Immunohistochemical biomarkers have been used to identify plasma cells in breast tumors and include CD138 and immunoglobulin kappa C87. The prognostic value of plasma cells is still unsettled and depends on biomarkers used to identify plasma cells with some studies pointing to a favorable prognostic factor91,92 while others reporting the opposite87,90. Evidence suggests there is also a T-cell independent immunosuppressive role for a subset of B cells, through their secretion of immunosuppressive cytokines such as IL-10 and TGFβ. These B cells are sometimes referred as tumor-evoked B cells in breast cancer93, and promote 11  lung metastases in a breast cancer mouse model by converting helper T cells into regulatory T cells93. In one study, B-cell gene signatures were shown to improve progression- and metastasis- free survival in breast cancer, with a more pronounced effect in the basal-like subtype94.   Overall, these studies emphasize the potential importance of the role played by B-cells in tumor immune responses, but more studies investigating the role of these important TILs are needed. T cells CD8+ TILs The elimination of cancer cells by the immune system is in part mediated by CD8+ cytotoxic T lymphocytes. After recognition of their specific epitopes presented on MHC class I molecules expressed antigen-presenting cells and by cancer cells, CD8+ T cells release perforin and granzymes leading to cancer cell death. The presence of infiltrating CD8+ T cells measurable by immunohistochemistry in breast carcinomas has been reported in several studies87,95,96. The location of CD8+ TILs in the tumor is assessed in the same manner as H&E evident lymphocyte counts, with CD8+ iTILs defined as CD8+ lymphocytes in direct contact with carcinoma cells or as CD8+ sTILs when they are located outside of carcinoma cells nests. CD8+ TIL counts are usually reported as absolute numbers. CD8+ tumor infiltrating lymphocytes (iTILs and sTILs) are observed mostly in the HER2-enriched and basal-like / triple negative breast cancer subtypes, which are known to be associated with increased levels of genomic instability2,95-97. Genomic instability can lead to increased production of neo-antigens that can be recognized by these CD8+ T cells, possibly explaining their high prevalence in these subtypes98. 12  The localization of CD8+ T cells in the tumor can affect their function, as CD8+ T cells need to be in direct contact with cancer cells to kill them. Stromal factors such as abundant collagen have been identified in stromal gene signatures of breast cancer (using laser capture microdissection) and can play a role in limiting T cell infiltration within tumors99. Results from several studies evaluating CD8+ TILs as a prognostic marker reported that high levels of CD8+ TILs are associated with better clinical outcomes. A large study that included 12 439 tumors combining patients from 4 different cohorts showed that CD8+ TILs were associated with a significantly reduced relative risk of death from breast cancer97 . Moreover, the presence of stromal and intraepithelial CD8+ TILs was associated with significantly higher survival when compared with tumors not containing CD8+ TILs. Importantly, a subgroup analysis showed that this prognostic effect was observed in ER negative and in HER2 enriched tumors, but was not observed in ER+ (luminal subtype) cases97. Consistent with these findings, a retrospective analysis done on 1334 primary breast cancer tissue samples demonstrated a significant favorable association of CD8+TILs with breast cancer specific survival in patients with ER negative cancers and their component HER2 enriched and basal-like intrinsic molecular phenotypes 95.  While there is therefore strong evidence supporting the prognostic effect of CD8+ TILs in breast cancer, a predictive utility for CD8+ TILs in the adjuvant setting has been demonstrated for the value of adding anthracycline chemotherapy to other chemotherapeutic agents97,100. A prospective-retrospective analysis of the National Epirubicin Adjuvant Trial (NEAT) showed that the presence of CD8+ iTILs among ER negative tumors was associated with a higher benefit in the anthracycline containing arm (epirubicin added to cyclophosphamide-methotrexate-5’fluorouracil(E-CMF)) compared to tumors negative for CD8+ iTILs. A significant interaction 13  was observed between the presence of CD8+ iTILs and the relative benefit of adding anthracyclines, at least among the ER negative breast tumors; patients receiving the control arm of non-anthracycline chemotherapy (CMF alone) did not show improved outcomes among cases with CD8+ iTILs 97.  Furthermore, I worked on a study that showed a significant interaction between low levels of CD8+sTILs and improved progression-free survival in metastatic breast cancer patients receiving trastuzumab, an antibody-based anti-HER2 agent101. The study was conducted on clinical trial material from patients enrolled in the MA.31 Phase III clinical trial that randomized metastatic breast cancer patients to receive adjuvant trastuzumab or lapatinib (small molecule-based anti HER2 agent) in combination with taxane chemotherapy101 . Results from this study suggest that low levels of pre-existing CD8+ TILs could favor an enhanced anti-tumor immunity promoted by an antibody-based therapy in metastatic settings. Several studies have reported that higher levels of CD8+TILs correlate with higher pCR rates after neoadjuvant chemotherapy102-104 ; however, all of these studies were retrospective analyses of non-randomized studies. Therefore, high level of evidence studies of the type required to prove the clinical utility of CD8+ TILs for predicting response to chemotherapy are still lacking. CD4+ T cells  CD4+ T cells recognized MHC class II expressed on antigen-presenting cells and  are divided into many subsets based on their specific functions and include Th1, Th2, regulatory T cells and follicular helper T cells105. By secreting IFNγ, Th1 CD4+ T cells are essential for the activation of CD8+ cytotoxic T cells, and have been shown to correlate with favorable survival 14  in breast cancer106. A different primary function is ascribed to Th2 CD4+ T cells, which secrete humoral immunity-related cytokines such as IL-4 and IL-6 105; their role in breast cancer is less clear.  The function of regulatory T cells is to dampen the immune system so as to limit excessive immune responses that can cause collateral damage to normal tissue. By immunohistochemistry, these cells are most readily identified via the expression of a nuclear factor called Forkhead box P3 (FOXP3). FOXP3+ TILs are associated with high risk clinicopathological factors such as high grade and ER negativity95,107 108-110. Some studies report significantly increased survival within ER negative subtypes that have high FOXP3+ TILs, specifically for triple negative and basal breast cancers95,107,111. The CD8/FOXP3 ratio is a parameter that indicates an activated immune microenvironment in the tumor and reflects the interplay between activating cytotoxic immune responses through CD8+ and downregulating it through FOXP3+ T cells. A higher CD8/FOXP3 ratio was shown to be significantly associated with better survival in ER negative tumors110. However, it should be noted that the analysis of FOXP3 lacks independent prognostic significance in multivariate analyses95,107 and many of the studies that examined the prognostic value of FOXP3 have a limited power to derive conclusions regarding its clinical utility. Accordingly, the role of FOXP3 as a prognostic marker in breast cancer needs to be further established in large prospective-retrospective clinical trial study designs. While it may seem intuitive that high levels of immunosuppressive FOXP3 TILs would reflect a pro-tumoral immunosuppressive microenvironmental change, their presence actually represents a natural secondary consequence of an active immune response and using a ratio of CD8/FOXP3 may 15  therefore not be an appropriate parameter for representing the clinical implications of the underlying biology. Follicular helper T cells (Tfh) are another subset of CD4+ lymphocytes which function as mediators of B cell activation in germinal centers112,113 . Tfh cells, measured by IHC or by gene expression, are present in breast cancer immune infiltrates and are linked to a better prognosis, especially within HER2+ cases 106. One study showed that high expression of CXCL13 (an important chemokine responsible for T and B cell homing in germinal centers) by Tfh in breast cancer is associated with higher levels of pathological complete response to anthracycline-based treatment regimens, and with better subsequent disease-free survival106.  1.3 Cancer immunotherapy: A new strategy in breast cancer treatment Immunotherapy can be described as any therapy that promote directly or indirectly the immune system. In cancer, immunotherapy agents can be broadly classified by their basic mechanism of action: immune augmentation or immune restoration. 1.3.1 Immune augmentation Immune augmentation strategies in cancer involve treatments that directly or indirectly establish and enhance anti-tumor immune responses. A few of these strategies are presented below. Direct effects  Immune augmentation can be established by altering anti-tumor immunity  through different methods such as therapeutic vaccination, adoptive T cell therapy and with chimeric antigen T cell receptors.  Therapeutic cancer vaccines are vaccines that induce immunity against targeted-cancer antigens. Cancers can produce a plethora of antigens with different degrees of immunogenicity 16  that may be targeted by strategies under active investigation including viral-based and dendritic cell-based vaccines such as sipuleucel-T114,115. Although not exclusively based on dendritic cells, sipuleucel-T is a health authority-approved therapeutic cancer vaccine that is intended for the treatment of hormone-refractory prostate cancers by targeting the prostatic acid phosphatase116.  In breast cancer, the most promising therapeutic cancer vaccines have been those targeting shared tumor-associated antigens such as HER2, with results yet to be published for the HER2 peptide E75 vaccine being evaluated in a phase III clinical trial (NCT01479244)117.   In adoptive T cell therapy, autologous TILs are activated in a milieu containing fragments from patient’s own tumor118. This technique leads to the selection and expansion of tumor-specific T cells which are then transferred back into the patients and has been used mostly in melanoma patients where it appears to reliably work118. Another immune augmentation strategy is manufacturing T cells with chimeric antigen T cell receptors (CAR T cells) which is now approved by health authorities in the United States for treatment of some forms of lymphomas119 . With recombinant DNA, the antigen receptors for CAR T cells are engineered to target receptors expressed on cancer cells (such as CD20 or CD19 for B cell lymphoma) in addition to carrying additional immune modulating regions in their cytosolic tails119. Adoptive cell therapy holds promise for otherwise poorly-immunogenic tumors such as ER+ breast cancers and can be used in combination with immune restoration strategies discussed below 120,121. Indirect effects Particular conventional chemotherapy agents are known to elicit an anti-tumor immune response by creating a process termed immunogenic cell death in dying cancer cells. Immunogenic cell death elicits the presence of damage-associate molecular patterns in dying 17  cancer cells such as exposure of cell surface calreticulin and release of HMGB1, promotes antigen cross-presentation by dendritic cells in addition to the release of antigens122,123. Cyclophosphamide and anthracyclines are among the conventional breast cancer chemotherapy agents working via this mechanism of action122 . Additionally, particular chemotherapy agents such as cyclophosphamide and gemcitabine preferentially suppress T regulatory and myeloid-derived suppressor cells to a much greater extent than cytotoxic and helper T-cells, enhancing antitumor immune responses124. In addition to chemotherapy agents, radiation is thought to augment anti-tumor immunity through the release of antigens in irradiated tumors that also lead to elimination of non-irradiated tumors, a process termed the “abscopal effect”125,126. However, additional factors such as selection of combination treatment agents and radiation dosage modulate the induced- immune response125,127.  Antibody-dependent cellular cytotoxicity, antibody-dependent cellular phagocytosis, and complement-dependent cytotoxicity are immune-mediated effects of some antibody-based targeted therapies such as trastuzumab 101,128-130. Antibody-dependent cellular cytotoxicity and phagocytosis appear to be more frequent than complement-dependent cytotoxicity and is mediated by immune populations with receptors for the Fc portion of the antibody131. These cells include macrophages that engulf antibody-coated cells in antibody-dependent cellular phagocytosis and Natural Killer cells that become activated following binding to their respective Fc receptors132,133. 1.3.2 Immune restoration via immune checkpoint receptor inhibition Immune restoration is characterized by strategies that reinvigorate a patient’s own anti-tumor immunity in situ such as by targeting immune checkpoint receptors (Figure 1.3). 18  Following activation of a T cell, the expression of various inhibitory receptors, termed immune checkpoints, is induced as part of a normal feedback circuitry as another way to avoid excessive immune responses and limit damage to normal tissue. The term ‘exhaustion’ is also used to describe effector T cells exhibiting reduced proliferation and decreased production of IFNγ in the context of a chronic antigen stimulation, a characteristic of the tumor microenvironment 134. In cancer, such pathways appear upregulated, as part of cancer immuno-evasion that characterizes the Escape phase. Fortunately, these repressive pathways themselves are targeted by checkpoint inhibitor drugs. CTLA-4   CTLA-4 is a receptor expressed on T cell membranes that acts through binding to CD80/86 expressed on dendritic cells, blocking T cell activation signal 2 and thereby leading to T cell anergy, a normal biological negative feedback process. Dendritic cell interactions with T cells that may yield CTLA-4/CD80 binding events occur mostly within peripheral lymphoid organs such as lymph nodes, and for this reason CTLA-4+ TILs are not often observed on primary tumor specimens135. Few immune gene signature studies have evaluated CTLA4 mRNA expression in breast cancer, although recent conference presentations describe increased expression correlating with reduced survival136,137.   The approach of CTLA-4 blockade and its clinical success in melanoma therapy pioneered the field of immune checkpoint inhibition138-140, leading to even more recent advances on other targetable immune checkpoints such as PD-1141. However, a breast cancer clinical trial examining tremelimumab (a human monoclonal anti-CTLA-4 antibody) in addition to exemestane in metastatic ER+ breast cancer patients did not achieve any partial or complete responses, while exposing patients to some potential serious side effects of anti-CTLA-4 therapy 19  (Table 1.2) 142. Accordingly, the selection of breast cancer patients for clinical trials that would benefit from anti-CTLA-4 agents needs to be done carefully, and specifically investigated among the particular subgroups of breast cancer patients most likely to benefit from immune restoration strategies (who are unlikely to be ER+ populations due to their low immunogenicity). PD-1/PD-L1  PD-1 is a cellular receptor expressed on antigen-experienced T cells143. Binding to its ligands – PD-L1 or PD-L2, which are expressed normally on antigen-presenting cells and aberrantly on tumor cells – leads to inhibition of effector functions144.  PD-1 inhibition of effector functions results from targeting CD28, the co-stimulatory T cell receptor, for dephosphorylation145. Inhibitors targeting the PD-1/PD-L1 pathway have exhibited durable responses in multiple clinical trials leading to various health authorities’ approval for treatment of certain cancers such as melanoma, non-small cell lung cancer and bladder cancer146. PD-1/PD-L1 inhibitors have also been evaluated in several breast cancer clinical trials including a recent phase III clinical trial (IMpassion 130) that randomized 902 advanced triple negative breast cancer patients to receive PD-L1 inhibitor (atezolizumab) with chemotherapy, or chemotherapy alone 147-149(Table 1.2). Accumulating reports from these trials show that clinical responses are mostly limited to patients receiving immune checkpoint inhibitors as first-line treatment and/or with immunogenic tumors characterized by PD-L1 expression 148,150-152. In the IMpassion 130 clinical trial, 40.9% of randomized patients were PD-L1 positive. PD-L1+ breast cancer patients randomized to the atezolizumab+chemo arm had a significant longer median progression-free survival (7.5 months vs 5 months, hazard ratio:0.62, 95%CI 0.49-0.78) and overall survival (25 months vs 15.5 months, hazard ratio: 0.62, 95%CI 0.45-0.86) in comparison 20  to PD-L1+ patients treated with chemotherapy alone148.  PD-L1 assessment in clinical trials as a potential predictive marker is the subject of controversy and has not reached a high level of analytical validity due to several factors such as heterogeneity of PD-L1 expression, variable cut-offs for PD-L1 positivity and the variety of PD-L1 monoclonal antibodies in diagnostic use153. It is important to highlight that while tumors with a positive PD-L1 expression can clinically benefit from PD-1 or PD-L1 checkpoint inhibitor monotherapy, other tumors with negative or minimal expression for PD-L1 might need the combination of other agents with anti-PD-1 to elicit the immune system to fight against their tumors. An intriguing hypothesis is that the combination of immune modulating conventional chemotherapy agents or radiotherapy with checkpoint inhibitors could have special value in patients with a low immune response that exhibit low base-line levels of TILs and minimal expression of PD-L1. In support of such hypotheses, a pilot study assessing the interaction of the immune system with the combination of tumor cryoablation plus ipilimumab interestingly found that the combination of these two very different treatment approaches leads to activation of the immune system, as marked by increased production of plasma IFNγ and proliferation of T-cells 154. Since this combination therapy has been shown to be associated with upregulation of IFNγ, it could lead to a higher expression of PD-L166,67 and thus could benefit patients with little to no pre-existing immune response (based on the levels of TILs and PD-1/PD-L1 expression). More studies are needed to define the best dosage therapy and target population of patients most likely to benefit from combinations of immunotherapy with conventional therapies. 21  1.3.3 Beyond CTLA-4 and PD-1/PD-L1 Following the success of CTLA4 and PD1/PDL1 inhibitors, additional immune checkpoint targets have recently emerged and made their way into early phase clinical trials155. These include LAG-3 and TIM-3. LAG-3 Lymphocyte Activation Gene 3 (LAG-3) is an exhaustion marker with immunosuppressive activity expressed on activated T cells 156. Major histocompatibility complex class II (MHC-II) is a ligand for LAG-3; additional ligands (e.g., L-selectin and galectin-3) have also been identified156. Regulatory T cells (Tregs) expressing LAG-3 have enhanced suppressive activity, whereas cytotoxic CD8+ T cells expressing LAG-3 have reduced proliferation rates and effector cytokine production in cancer and autoimmune diabetes157-159. A splice variant of LAG-3 cleaved by metalloproteinases and secreted in the cellular microenvironment has immune-activating properties when bound to MHC-II on antigen presenting cells160. LAG-3+ tumor-infiltrating lymphocytes (TILs) have been reported in melanoma, colon, pancreatic, breast, lung, hematopoietic, and head and neck cancer patients161-167, in association with aggressive clinical features. Antibody-based LAG-3 blockade in multiple cancer mouse models restores CD8+ effector T cells and diminishes Treg populations, an effect enhanced when combined with anti-PD-1168,169. A recent study in a metastatic ovarian cancer mouse model showed that LAG-3 blockade leads to upregulation of other immune checkpoints (PD-1, CTLA-4, and TIM-3), and combination therapy targeting LAG-3, PD-1, and CTLA-4 increases functional cytotoxic T cell levels while reducing Tregs and myeloid-derived suppressor cells170. Multiple early phase clinical trials are testing antagonistic LAG-3 agents in combination with anti-PD-1 and/or anti-CTLA-4 therapy (>15 phase I or II clinical trials on  February 2019). In view of the activating properties of soluble secreted LAG-3, a soluble agonist LAG-3 antibody (IMP321) was tested in advanced solid malignancies as a single agent171, and demonstrated sufficient tolerability and efficacy to warrant advancement to phase II (NCT02614833). TIM-3 T-cell Immunoglobulin- and Mucin-domain-containing molecule 3 (TIM-3) is an immune-inhibitory molecule first identified on CD4+ Th1 (helper) T-cells and CD8+ Tc1 (cytotoxic) T-cells172, then later on Th17 T-cells,173 regulatory T-cells174,175, and innate immune cells176-178. TIM-3 is activated primarily by its widely-expressed ligand, galectin-9179, leading to effector T-cell death through calcium influx, cellular aggregation, and apoptosis180. When TIM-3 signaling is active, interferon-producing T-cells become exhausted, resulting in Th1 suppression and immune tolerance180-182. TIM-3 expression is commonly observed during chronic infection, as a characteristic marker of exhausted T cells.183-187. In cancer, tumor-infiltrating lymphocytes expressing TIM-3 have been observed in melanoma188,189, non-Hodgkin’s lymphoma190, lung174, gastric191, and other cancers192-195. In these studies, TIM-3 is co-expressed with PD-1 and associated with effector T-cell exhaustion and dysfunction. This phenomenon is also observed in mouse models of solid196 and hematologic197 cancers, where TIM3+PD1+CD8+ T-cells exhibit an exhausted phenotype characterized by reduced proliferation and defective production of IL-2, TNFα, and IFN-γ. In contrast, TIM-3 positive Treg display increased expression of effector molecules and are more immunosuppressive than their TIM-3 negative counterparts198,199.  Inhibition of TIM-3 alone tends to have little effect on tumor growth in pre-clinical mouse models, despite some evidence supporting a reversal of immune cell exhaustion188,196,200-23  202. However, combined targeting of PD-1 and TIM-3 leads to a substantial reduction in tumor growth – better than either pathway alone – in numerous preclinical in vivo models188,196,197,202, supporting the concept that malignant cells become resistant to PD-1 checkpoint blockade by activating another immune checkpoint. Indeed, mouse models partially responsive to PD-L1 inhibition upregulated TIM-3 expression in resistant tumors194,203, and addition of TIM-3 blockade was successful in overcoming that resistance. Upregulation of TIM-3 has also been observed in patients receiving PD-L1 monotherapy, suggesting it may represent a form of adaptive resistance to this therapy 203. At least seven early phase clinical trials are underway that attempt to combine anti PD-L1/PD-1 therapy with agents targeting TIM-3 (NCT03489343, NCT02817633, NCT03680508, NCT03311412, NCT03099109, NCT03744468, NCT03066648). 1.4 Rationale for thesis and research objectives As an increasing number of immune targets for cancer immunotherapy are being discovered in pre-clinical animal studies, the characterization of immune infiltrates in patient tumors and the investigation of these immunotherapy targets are crucial for clinical trial study design and assay development studies that may lead to clinical implementation of immune targets as potential prognostic or predictive biomarkers.  At the start of my thesis, immuno-oncology research in breast cancer was still in its infancy and early phase immune checkpoint inhibitor clinical trials in breast cancer were just opening. The main goal of my research was to investigate the presence and prognostic implications of clinically relevant immune infiltrates in well-annotated breast cancer pathology specimens.  My hypothesis is that immunotherapy targets will be expressed on infiltrating immune cells in a specific subset of breast cancers, detectable in breast cancer pathology specimens. 24  Furthermore, these biomarkers will have prognostic implications and may serve as a guide for breast cancers amenable to emerging immunotherapy strategies.  Specifically, my thesis is based on the following 2 aims:  1. Development and prognostic analyses of emerging immune checkpoint biomarkers by immunohistochemistry in an annotated breast cancer cohort, with subsequent validation using a larger breast cancer cohort. For this aim, my hypothesis is that emerging immunotherapy targets would be enriched in estrogen receptor negative breast cancer patients and their presence will be associated with unfavorable survival.   I reviewed the literature to identify emerging immune checkpoint biomarkers beyond PD-1/ PD-L1 that were being evaluated in clinical trials. I then organized evaluation of these candidates by immunohistochemistry on an initial breast cancer patient cohort consisting of 330 breast tumor excision specimens built into tissue microarrays and linked to clinical outcome. This cohort serves as a training set to screen for biomarkers compatible for immunohistochemistry assessment on formalin-fixed paraffin-embedded tumor tissues. In addition, I used the training set for cut-point determination for positivity and to generate hypotheses to be validated on an independent larger cohort. The latter consists of a tissue microarray series spread over 17 blocks, representing 3,992 breast cancer patients linked to clinical outcome and extensive biomarker data. Prior to the analyses on this larger cohort, my results on the training cohort set of 330 breast cancer patients and the specific hypotheses I generated were reviewed by the breast cancer outcomes unit at the BC Cancer Agency for approval to access the valuable large cohort.   2. Development of a new in situ multiplex methodology for immune profiling of breast cancer patients  25  My hypothesis for this second aim is that a signature composed of immune and tumor biomarkers would discriminate between immune-enriched and immune-desert profiles of breast cancer tumors. Immune infiltrates present in tumors represent multiple important immune cell types which likely mediate different activities within the tumor microenvironment that could promote or inhibit responses to immune-modulating therapies. Assessing and visualizing the presence of multiple biomarkers within a tissue remains a technical challenge. For this aim, I set out to evaluate a novel immunohistochemical multiplex technology called digital spatial profiling by Nanostring, capable of generating quantitative assessment of 31 targets in tumor tissues. I tested the feasibility of the digital spatial profiling technology using two different breast cancer tissue microarray cohorts.   The results generated in the first aim are presented in Chapter 2 and Chapter 3 whereas Chapter 4 presents the results of the Nanostring-based digital spatial profiling platform tested on two initial cohort-based sets. Chapter 5 summarizes the body of work presented in the thesis and provides perspectives for the future in breast cancer immunotherapy.  26     Figure 1.1 Cancer-immune interaction Simplified illustration depicting the 3 phases of cancer-immune interaction: Elimination, Equilibrium and Escape. The dynamic tumor immune microenvironment is depicted by the changes in the composition of cancer clones and immune populations at each phase.    27   Figure 1.2 Assessment of tumor-infiltrating lymphocytes in the breast cancer microenvironment. Hematoxylin and Eosin-stained section of a breast cancer specimen displaying the immune infiltrates in the tumor stroma compartments used for scoring TILs (representative scoreable areas are indicated by white circles) as per recommendations of the international TILs Working Group74. A, low magnification view of the tumor used to identify areas where stromal TIL count can be assessed. B, High magnification view of one stromal area included in the scoring, estimated at 20% sTILs. Reproduced with permission.28     Figure 1.3 Immune regulation in the breast cancer microenvironment Different populations of immune cells are observed in breast cancer specimens and are associated with both anti- and pro-tumorigenic effects. Immune checkpoint inhibitor-expressing immune cells are identified by the presence of PD-1, LAG-3 and TIM-3 on the cell 29  surface. Green arrows indicate interactions leading to activation whereas red arrows indicate inhibitory interactions. Reproduced with permission.30  Table 1.1 Description of breast cancer clinical intrinsic subtypes.  Luminal A Luminal B HER2E Basal-like Frequency (%) 70-80 10-15 10-15 Clinical IHC surrogate ER+, PR+  Ki67 (<14%) HER2– ER+, PR+ Ki67 (≥14%) (Luminal/HER2+) ER–, PR– HER2+ ER–, PR–, HER2–, CK5/6+ and/or EGFR+ TNBC (ER–, PR–, HER2–) Therapy options Hormonal therapy +/- chemotherapy or radiotherapy Hormonal therapy, chemotherapy, radiotherapy Anti-HER2,  Chemotherapy Chemotherapy, PARP inhibitors for BRCA-mutated  Prognosis Good Intermediate Intermediate Worse Immunogenicity Low Intermediate Intermediate High   31  Table 1.2 Breast cancer immunotherapy clinical trials   ORR: objective response rate by RECIST 1.1 criteria; CR: complete response, PR: partial response pCR: Pathologic complete response    Target Agent (company) Study Phase Population Reported resultsPD-1 Pembrolizumab (Merck) KEYNOTE-028 (single agent)Rugo HS et al., Clin Can Res, 2018Ib Advanced ER+/HER2-(N=25)ORR: 12% (3 PR)KEYNOTE-086 (single agent)Adams S et al.,Ann Oncol, 2018 II Metastatic triple negative breast cancer(N=254)Cohort A: previously treated, TNBC unselected (N=170) : ORR=5.3% (2 CR, 7 PR)Cohort B: previously untreated, TNBC PD-L1+ (N=84): ORR= 21.4% (4 CR, 14 PR)KEYNOTE-012 (single agent)Nanda R. et al., J Clin Oncol, 2016Ib Selected PD-L1+ metastatic triple-negative breast cancer(N=27)ORR=18.5 % (1CR, 4 PR)ENHANCE-1/KEYNOTE-150 (in combination with eribulin)Tolaney SM et al., Canc res, 2018.P6-13 abstrIb/II Advanced triple-negative breast cancer(N=106)ORR: 26.4% (3 CR, 25 PR)KEYNOTE-173 (in combination with various chemotherapy agents) Schmid P et al., J Clin Oncol, 556 (2017). abstr.Ib Advanced triple-negative breast cancer(N=20)pCR:Cohort A treated with pembro followed by pembro+nab-paclitaxelfollowed by pembro+doxorubin/cyclophosphamide: 60%Cohort B treated with pembro followed by pembro+ nab-paclitaxel + carboplatin and by pembro+ doxorubin/cyclophosphamide: 90%I-SPY 2 (in combination with paclitaxel or paclitaxel alone followed by doxorubicin/cyclo-phosphamide)Nanda R et al., J Clin Oncol, 506 (2017).abstr.II Primary breast cancer(N=249)pCR: TNBC: 60% pembro arm vs 20% controlHR+/HER2-: 34% pembro vs 13% controlPD-L1 Avelumab(Pfizer and EMD serono) JAVELIN (single agent)Dirix LY. et al, Breast Cancer Res Treat, 2018Ib Metastatic breast cancer(N=153)ORR=4.8% (1 CR, 7 PR)Atezolizumab (Roche) IMPassion 130 (in combination with nab-paclitaxel vs nab-paclitaxel alone)Schmid P et al., N Engl J Med, 2018 III Advanced triple-negative breast cancer(N=902; randomized 1:1)Primary endpoints in atezo+nab-pacli: Median PFS: 7.2 months vs 5.5 months, p=0.002Median OS : 21.3 months vs 17.6 monthsp=0.08Durvalumab (Imfizi) Yale studyPusztai L et al., J Clin Oncol, 572 (2017). abstrI/II Primary triple-negative breast cancer(N=7)pCR: 71.4%CTLA-4 Tremelimumab(Astrazeneca)Vonderheide RH et al.,Clin Cancer Res, 2010, (In combination with exemestane I Metastatic ER+/HER2- breast cancer (N=26)ORR=011 SD (42%)32  Chapter 2: LAG-3+ tumor infiltrating lymphocytes in breast cancer: clinical correlates and association with PD-1/PD-L1+ tumors SYNOPSIS Novel immune checkpoint blockade strategies are being evaluated in clinical trials and early phase studies are now including strategies targeting the lymphocyte activation gene 3 (LAG-3) checkpoint, alone or in combination with PD-1/PD-L1 blockade. In this chapter, I investigated LAG-3 expression and its prognostic value in a large series of breast cancer patients, and correlated LAG-3 expression with key biomarkers including PD-1 and PD-L1. LAG-3 expression was evaluated by immunohistochemistry on two tissue microarray series incorporating 4322 breast cancer primary excision specimens (N=330 in the training and N= 3,992 in the validation set) linked to detailed clinico-pathological, biomarker and long term clinical outcome data. PD-1 and PD-L1 expression were also evaluated by immunohistochemistry.  After locking down interpretation cutoffs on the training set, LAG-3+ intra-epithelial tumor-infiltrating lymphocytes (iTILs) were found in 11% of cases in the validation set. In both sets, LAG-3+iTILs were significantly associated with negative prognostic factors: young age, large tumor size, high proliferation, HER2E and basal-like breast cancer subtypes. In multivariate analyses, breast cancer patients with LAG-3+iTILs had a significantly improved breast cancer-specific survival (BCSS) (HR: 0.71,95%CI 0.56-0.90), particularly among ER–patients (HR: 0.50,95%CI 0.36-0.69). Furthermore, we found that 53% of PD-L1+ and 61% of PD-1+ cases are also positive for LAG-3+iTILs, supporting potential immune checkpoint blockade combination strategies as a treatment option for breast cancer patients.   33  2.1 Introduction Although breast cancer is not generally considered an especially immunogenic malignancy in comparison to melanoma and lung cancer47, recent studies show that some tumors, especially estrogen receptor (ER) negative breast cancers, do elicit an immune response 70. Immune checkpoint inhibitors targeting cytotoxic T-lymphocyte-associated protein 4(CTLA-4), the programmed cell death receptor 1(PD-1) and/or its ligand (PD-L1) have shown clinical efficacy, especially in melanoma and lung cancer 138,204. In breast cancer, results from early phase trials have suggested checkpoint inhibitor efficacy may be primarily seen in triple negative cases 205. While the immuno-oncology field is moving at a fast pace, large scale studies of immune checkpoint expression in breast cancer series are as yet few. The lymphocyte activation gene 3 (LAG-3) represents one example of a new immune checkpoint target.  LAG-3 is a cellular receptor expressed by activated T lymphocytes and is associated with T cell exhaustion 156. LAG-3 is commonly upregulated with PD-1; pre-clinical data suggests that LAG-3 blockade releases T cells effector functions and synergizes with other immune checkpoint inhibitors such as anti-PD-1168. Thus, LAG-3 represents an interesting target for immunotherapy and there are a number of ongoing early phase clinical trials testing anti-LAG-3 therapeutic antibodies in different types of cancer. Our objective was to assess the prognostic value of LAG-3 and its association with PD-1 and PD-L1 immune checkpoints in a large, well-characterized cohort of breast cancer specimens.  34  2.2 Materials and methods 2.2.1 Study cohorts The initial set (a training set used to finalize staining and scoring methodology) consisted of 330 invasive breast cancer cases from University of British Columbia hospitals diagnosed between 1989 and 2002 206. The validation set, previously described in detail 107, consisted of 3,992 female patients diagnosed with invasive breast cancer at centres of the British Columbia Cancer Agency across the province between 1986 and 1992 and for which formalin-fixed paraffin-embedded primary breast tumor excision specimens were collected from a central estrogen receptor testing laboratory. None of these patients received neoadjuvant treatment.  A further description of the study cohorts is provided in Section 3.2.1 of Chapter 3. Basic clinical and pathological parameters of the study populations are summarized in Table 2.1 and 2.2. The median follow-up time is 12.6 years for both sets. Access to the samples and corresponding de-identified clinico-pathologic, treatment and outcome data was approved by the Clinical Research Ethics Board of the University of British Columbia and by the BC Cancer Agency Breast Cancer Outcomes Unit.    2.2.2 Immunohistochemistry Formalin-fixed paraffin-embedded primary excision specimens were used to build tissue microarrays, represented as 0.6mm cores as previously described 206,207. Biomarkers previously stained by immunohistochemistry on the validation tissue microarray include ER, PR, HER2, Ki67, EGFR, CK5/6, and CD8 96. Immunohistochemistry for LAG-3 and PD-1 was performed using a Ventana Ultra automated stainer in accordance with manufacturer’s protocol using the following antibodies and concentrations: LAG-3 (Clone 17B4, dilution 1:100, Abcam, see Appendix A), PD-1 (Clone NAT105, dispenser, Roche). PD-L1 staining was performed at the 35  Deeley Research Centre (Victoria, Canada) using clone SP142, dilution 1:100 (Spring Bioscience) (see Figure 2.1 for representative staining of LAG-3, PD-1, and PD-L1). Breast cancer intrinsic subtypes from both cohorts were previously determined by immunohistochemical methods that were originally developed against gene expression gold standards 208. Briefly, Luminal A was defined as ER+ (≥1%) or PR+ (≥1%), HER2– and low Ki67, Luminal B as ER+ (or PR+) and HER2+ or high Ki67 (≥14%); HER2E as HER2+ and ER– and PR–; Basal-like as HER2–/ER–/PR– with EGFR+ or CK5/6+.    2.2.3 LAG-3, PD-1 and PD-L1 scoring Biomarkers were evaluated and reported following REMARK guidelines 209. TMA slides were digitally scanned and visually scored by an experienced pathologist blinded to clinical information. Scoring and quantification of LAG-3 and PD-1 were carried out as previously described for other lymphocyte biomarkers 107. In brief, stromal lymphocytes were defined as lymphocytes not in direct contact with the breast carcinoma nest whereas intra-epithelial lymphocytes were located within the carcinoma nest. Scores of lymphocyte biomarkers were reported as absolute counts, and any positive expression (≥1 TILs per TMA core) was used for dichotomization into positive and negative cases. PD-L1 scoring is a matter of controversy in the current literature; in this study we used the method of McDermott DF et al., 210 whereby expression was assessed as the percentage of carcinoma cells with membranous expression; any cores with ≥1% of PD-L1+ carcinoma cells were considered positive. We found only 24 cases out of 2918 cases with PD-L1+ immune cells and they were discarded for further analyses. Cutoffs determined using the training set were locked down for analysis on the validation set. Overall stromal tumor-infiltrating lymphocytes were assessed following the methods established by the international TIL working group74 on scanned images of hematoxylin and 36  eosin-stained tissue microarray slides and were only available on the training set at the time of the publication of this chapter. 2.2.4 Statistical Analysis Statistical analyses were carried out using IBM SPSS (version 24.0) and R (version 3.3.2) software. Breast cancer-specific survival (BCSS) was the primary outcome, defined as the time from date of diagnosis to date of death attributed to breast cancer. Patients who died from another cause or were alive at the end of follow-up were censored. Relapse-free survival (RFS) analyses were also performed, defined as time from date of diagnosis to date of any type of breast cancer relapse (local, regional, distant, or contralateral). In addition, patients were censored if they had not died from breast cancer or if they had not relapsed at the end of follow-up time for relapse-free survival analyses. Univariate associations between LAG-3 and survival were examined using Kaplan-Meier survival curves and log-rank test. Cox proportional hazards regression was used to estimate hazard ratios for LAG-3 adjusted for the following clinicopathological parameters: age (<50 vs ≥50), tumor grade (grade 3 vs grade 1,2), tumor size (>2cm vs ≤2 cm), lymphovascular invasion status and nodal status). Findings observed on the initial set, to be tested on the validation set, were prespecified in a formal written statistical plan, presented at the BC Cancer Agency Breast Cancer Outcomes Unit. Additionally, for analyses of lymphocytes biomarker relationships, due to low numbers of positive cases in the initial set, a training and validation approach was used by splitting the validation set in half. Findings to be validated on the other half of the set were prespecified in a written statistical plan prior to the analyses (see Appendix B). All statistical tests performed were two-sided at α=0.05.    37  2.3 Results 2.3.1 Selection of LAG-3+TILs cutoffs on the training set 278 (84%) of breast cancer cases in the training set were interpretable for LAG-3 (Table 2.1).  As relatively few cases had LAG-3+TILs in 0.3mm2 cores, any cases that had ≥1 lymphocyte expressing LAG-3 were deemed as positive. We found stromal lymphocytes expressing LAG-3 in 15% and intra-epithelial lymphocytes expressing LAG-3 in 14% of cases. Both LAG-3+sTILs and iTILs were significantly associated with negative prognostic factors including high grade tumor, ER negativity and high Ki67 proliferation (Table 2.1). Following our previous publications on lymphocyte biomarkers 96,107, we set on LAG-3+iTILs≥1 per 0.3mm2 core as our primary analysis for the validation set to allow comparison among lymphocyte biomarkers. 2.3.2 LAG-3+iTILs are enriched in ER negative subtypes and associated with improved survival 2,921 (73%) cases were interpretable for LAG-3 staining on the validation set. LAG-3+iTILs≥1 per 0.3mm2 core were observed in 11% of breast cancer patients with a distribution range of 0-45 (Table 2.2). The presence of LAG-3+iTILs was significantly associated with younger age, large tumor size, ER/PR negativity, and high Ki67 proliferation index (Table 2.2; interaction test by subtype shown in Table 2.3). ER negative tumor subtypes more commonly contained LAG-3+iTILs, present in 33% of basal-like and 27% of HER2E samples (versus 3% and 11% in luminal A and luminal B tumor subtypes respectively).  In the whole cohort and in the ER+ breast cancer subset (81% of cases), the presence of LAG-3+iTILs was not significantly associated with survival in univariate analyses (Figure 2.2A-B). In contrast, ER negative breast cancer patients with LAG3+iTILs had significantly improved 38  breast cancer-specific survival and relapse-free survival, an association present in both HER2E and basal-like subtypes (Figure 2.2C-E, Figure 2.3). In multivariate analyses that adjusted for breast cancer-specific clinicopathalogical factors, LAG-3+iTILs were a significant favorable prognostic factor in the whole cohort and among ER negative breast cancer patients (Table 2.4A). Due to the correlation between the other immune checkpoints markers, the presence of LAG-3+iTILs is not an independent prognostic factor in a multivariate analysis that include CD8, PD-1 and PD-L1 as covariates (Table 2.4B); only the presence of PD-1+iTILs represented a significant prognostic factor for improved breast cancer-specific survival in this model. 2.3.3 LAG-3+iTILs are strongly associated with PD-L1/PD-1+ tumors Expression of immune checkpoint markers is regulated in a time-dependent manner and can lead to cells expressing multiple immune checkpoint markers such as LAG-3 and PD-1156. We investigated the association between tumors containing LAG-3+TILs, PD-1+TILs, and carcinoma cells expressing PD-L1.  As expected, immune checkpoint markers were significantly associated with each other (Table 2.5) and with total H&E stromal TIL levels (Table 2.6). 53% of PD-L1+tumors and 61% of tumors with PD-1+TILs were also positive for LAG-3+iTILs on equivalent-sized tissue microarray cores, whereas only 38% of tumors with CD8+iTILs were co-infiltrated with LAG-3+TILs. The percentage of PD-L1+tumors infiltrated with PD-1+TILs was similar to that of LAG-3+TILs but only ~20% of tumors with CD8+iTILs had carcinoma cells positive for PD-L1 (Table 2.5). Scatter plots relating the investigated biomarkers are presented in Figure 2.4 39  2.3.4 Exploratory survival analyses of concurrent tumor infiltration with LAG-3+ and CD8+T cells Given that LAG-3 can be expressed on different lymphocyte subsets (CD8 or CD4)168, we investigated the association between the presence of cytotoxic CD8+iTILs and the prognostic value of LAG-3+iTILs. In the absence of CD8+iTILs, with relatively few cases positive for LAG-3, we found that the presence of LAG-3+iTILs was no longer associated with survival among ER negative breast cancer patients (Figure 2.5A). In contrast, concurrent infiltration of LAG-3+iTILs and CD8+iTILs was significantly associated with improved survival, suggesting an important role of CD8+TILs expressing LAG-3 in breast cancer. Furthermore, we observed similar findings with concurrent infiltration of PD-1+iTILs and CD8+iTILs (Figure 2.5B; LAG-3 and PD-1 co-infiltration shown in Figure 2.6). Surprisingly, in ER negative breast cancer patients with no detectable CD8+iTILs per 0.3mm2 core, the presence of PD-L1+ carcinoma cells was associated with a trend for poor BCSS, whereas patients with both CD8+iTILs and PD-L1+ carcinoma cells had significantly improved BCSS (Figure 2.5C).  2.4 Discussion In a large set of breast cancer patients, we report on the presence of an immune checkpoint biomarker targetable with new drugs. We found that LAG-3+TILs can be scored quantitatively by immunohistochemistry on tissue microarrays with consistent results supporting its analytical validity as a biomarker. By this methodology, LAG-3+TILs were observed in a limited subset of patients (11%) that mostly fall into the HER2E and basal-like breast cancer subtypes and is a favorable prognostic factor in ER negative breast cancers. The presence of 40  immune infiltrates most prominently in ER negative breast cancers, detectable with lymphocyte biomarker assays, is now supported by a large body of evidence 70. In contrast, LAG-3+iTILs were not prognostic in ER+ patients. This finding may not be surprising as ER+ breast cancers appear generally to be less immunogenic and have lower levels of TILs than ER negative patients 70. In the FinHER trial, the presence of TILs was not associated with survival in early stage ER+/HER2‒ breast cancer patients 80, and in a combined analysis of two French Phase III clinical trials TILs were only prognostic in triple negative and HER2 positive patients 83. High pretreatment serum levels of soluble LAG-3 have been associated with improved survival among ER+ breast cancer patients 211. Soluble LAG-3 represents a splice variant with activating functions when bound to major histocompatibility complex II protein expressed on dendritic cells 134. By contrast, our study measured the expression of LAG-3 on TILs in breast carcinoma tissues, and agrees with a recently published smaller study that found a subset of triple negative breast cancers had concurrent infiltration of LAG-3+/PD-1+ TILs and that this was associated with a (non-significant) trend for improved survival  163.  In pre-clinical studies, LAG-3 expressing CD8+TILs are exhausted and do not function properly, whereas LAG-3 expressing CD4+ regulatory T cells exhibit an enhanced immune suppressive function, supporting a hypothesis that the presence of LAG-3+TILs in cancer patients should lead to poor survival. Instead, our results suggest that the presence of LAG-3 expressing TILs may in fact indicate that there is an ongoing cancer-immune interaction, a phenotype that is described as an inflamed tumor 212 and usually implies an improved prognosis. Indeed, we found that >50% of breast cancers that are PD-L1+ or are infiltrated with PD-1+TILs have concurrent infiltration of LAG-3+TILs. 41  In the training set where H&E sTILs counts were available, we found that high levels of H&E sTILs (>50%, a level termed “lymphocyte predominant breast cancer” or LPBC) correlated with increased co-infiltration of LAG-3+/PD-1+ TILs. H&E sTILs are prognostic in ER- breast cancers and can be predictive of response to trastuzumab in HER2+ patients 80,213; emerging evidence also suggests TIL count may predict response to anti-CTLA4 (ipilimumab) and anti-PD-1(pembrolizumab) 214,215. TIL counts performed on H&E sections have the advantage of being applicable to existing slide sets or incident cases without requiring new immunohistochemical assays. However, this method necessarily combines all lymphocytes – including activating, suppressing and anergic populations, as well as NK and B cells. Theoretically, direct assessment of LAG-3 should be more likely to predict response to LAG-3-targeted checkpoint inhibitors; our study is a first step in defining an assessment method and a description of expression patterns over a large number of cases with detailed follow-up. The tissue microarray format limits the capacity to directly compare with standardized H&E TIL scores 74, and an assessment of the predictive capacity of LAG-3 in breast cancer will of course require application to randomized clinical trials.  The clinical activity of immune checkpoint inhibitors as single agents in breast cancer appears limited based on recent and varying results. In the KEYNOTE-012 trial of 32 women with advanced triple negative breast cancer (TNBC) and ≥1% PD-L1 expression treated with pembrolizumab (a PD-1 inhibitor), the response rate was 18.5% 205, with the median duration of response not yet reached at study publication. In a more recent presentation of 115 metastatic TNBC patients treated with atezolizumab (a PD-L1 inhibitor), a 10% overall response rate was seen irrespective of PD-L1 status, with most of this activity in the 1st line setting216. The median progression free survival was 1.4 months, but in those patients who achieved a response, the 42  median duration of response was 21.1 months. Of note, in the exploratory biomarker analyses, there was a suggestion for better clinical outcomes in those tumors with higher levels of CD8+ T cell infiltrates. Lastly in a phase Ib study of previously taxane / anthracycline-treated, advanced breast cancer (N=168) treated with avelumab (a PD-L1 inhibitor), the response rate was only 4.8% in the overall population, and 8.6% in the TNBC cohort217. It seems clear that the majority of advanced breast cancers do not achieve a response to single agent immune checkpoint inhibition, although in the minority that do, the response appears relatively durable – as has been seen in other malignancies treated with these agents. The results of our study suggest considering stratification of patients in these clinical trials for both PD-1/PD-L1 status and LAG-3 status. As the field is now studying combinations of immune checkpoint inhibitors, our results raise the question whether the cohort of tumors with co-expression of PD-1/PD-L1 and LAG-3+iTILs should be excluded from these trials as the natural history of this cohort is relatively favorable. However, this could be due to the effects of adjuvant therapy that may have reinvigorated a de novo immune response. Although this study has major strengths, such as the use of analytically validated antibodies and multiple, large, well-annotated sets of breast cancer specimens, it also has some limitations. While long term follow-up is a strength, the study cohort does date from a time prior to use of HER2 targeted therapies or taxanes, which may affect generalization to contemporary patients. The breast cancer specimens used in the study were retrospectively collected and assembled into TMAs and as such, the area of the tumor analyzed reflects only a minute sampling of the original tumor tissue. In addition, co-infiltration of immune subsets (LAG-3+, PD-1+, CD8+) had to be inferred from results of individually-stained tissue sections and therefore cannot directly assess co-expression of immune checkpoints on the same cell.  43  In conclusion, LAG-3 is an immune checkpoint marker targeted by emerging treatments and is most commonly expressed among ER- breast cancer patients – including in one third of basal-like breast cancers, an aggressive subtype where checkpoint inhibitors have great promise and potential. Although our study does not directly measure biomarker expression in the metastatic setting, the strong association between tumors positive for PD-1/PD-L1+ and LAG-3+ biomarkers suggests a potential for the combination of therapies targeting these immune checkpoint markers, a concept currently being evaluated in clinical trials in metastatic disease 218.        44     Figure 2.1 Representative immunohistochemical staining of LAG-3, PD-1, and PD-L1 in breast tumor tissue microarray cores Brown staining on intra-epithelial and stromal tumor-infiltrating lymphocytes can be observed for LAG-3 and PD-1 micrographs whereas PD-L1 staining can mainly be observed on carcinoma cell membranes. Micrographs were taken under X20 objective magnification.LAG-3 PD-1 PD-L1 45      Figure 2.2 LAG-3+iTILs association with breast cancer-specific survival. (A) whole cohort, (B) ER+, (C) ER‒, (D) HER2E and (E) Basal-like.  46   Figure 2.3 Association with relapse-free survival for LAG-3+ intra-epithelial tumor-infiltrating lymphocytes. Association with relapse-free survival for LAG-3+iTILs. (A) Whole cohort, (B) ER+, (C) ER‒, (D) HER2E and (E) Basal-like.   47   Figure 2.4 Correlation between LAG-3, PD-1, PD-L1 and CD8 scores. Scatter plots with spearman’s rho correlation coefficient and p values for each pair of immune markers. A, LAG-3+iTILs and CD8+iTILs; B, LAG-3+iTILs and PD-1+iTILs; C, LAG-3+iTILs and PD-L1+ carcinoma cells; D, PD-1+iTILs and CD8+iTILs    Spearman’s rho: 0.399p=<0.001Spearman’s rho: 0.334p=<0.001Spearman’s rho: 0.499p=<0.001Spearman’s rho: 0.423p=<0.001ACBD48     Figure 2.5 Association between CD8+iTILs and the prognostic value of immune checkpoint markers. Breast cancer-specific survival Kaplan-Meier curves of patients in the whole validation set stratified by the presence of immune checkpoint markers (A) LAG-3, (B) PD-1, (C) PD-L1 in the absence (Left panel) or presence (Right panel) of concurrent CD8+iTILs.   A B C 49   Figure 2.6 Association with breast cancer-specific survival for LAG-3+/PD-1+ intra-epithelial tumor-infiltrating lymphocytes in the presence or absence of CD8+ intra-epithelial tumor-infiltrating lymphocytes among ER negative patients.  50   Table 2.1 Association of LAG-3+intra-epithelial and stromal tumor-infiltrating lymphocytes with clinicopathological parameters in the training set   Training set (N=278) Parameters Negative  N=240 LAG-3+iTILs ≥1 N=38 (14%)   P value* Negative  N =235 LAG-3+sTILs ≥1 N =43 (15%)   P value* Age at diagnosis (year) <50 ≥50   90 150   17(16%) 21(12%) 0.47   89 146   18 (17%) 25 (15%) 0.61 Tumor size (cm) ≤2 >2  143 97  16 (10%) 22 (19%) 0.05  138 97  21 (13%) 22 (18%) 0.24 Grade 1&2 3 Unknown  157 81 2  6 (4%) 32 (28%) - <0.001  155 78   8 (5%) 35 (31%)  <0.001 ER Negative Positive Unknown  36 202 2  23 (39%) 15 (7%) - <0.001   37 196 5  22 (37%) 21 (10%) - <0.001 PR Negative Positive Unknown  66 171 3  26 12 - <0.001   66 166 3  26 (28%) 17 (9%) - <0.001 Ki67 <14% ≥14% Unknown  149 86 5  8 30 - <0.001   144 86 5  13 (8%) 30 (26%) - <0.001 Subtypes Luminal A Luminal B HER2E Basal-like NOS Unknown  135 56 7 14 3 25  6 (4%) 7 (11%) 4 (36%) 11 (44%) 0 10 <0.001  131 53 7 16 3 25  10 (7%) 10 (16%) 4 (36%) 9 (36%) 0  10 <0.001 *Fisher’s exact tests were computed for 2x2 association. Association between subtypes and TIM-3+TILs was analyzed by Chi square test      51  Table 2.2 Association of LAG-3+ intra-epithelial tumor-infiltrating lymphocytes and clinicopathological parameters.  Validation set (N =2921)  Parameters  Negative   N=2594  LAG-3+iTILs ≥1  N=327 (11%)     P value* Age at diagnosis (years)  <50 ≥50   719 1875   131(15%) 196 (9%) <0.001 Tumor size (cm) ≤2 >2  1371 1211  139 (9%) 186 (13%) <0.001 Grade 1&2 3 Unknown  1206 1283 105  60 (5%) 255 (17%) 12 <0.001 Ki67 negative positive (≥14%) Unknown  1386 989 219  66 (4%) 237 (19%) 24 <0.001 ER Negative Positive (>1%) Unknown  620 1965 9  195 (24%) 132 (6%) 0 <0.001 PR Negative Positive (>1%) Unknown  1094 1354 146  234 (18%) 84 (6%) 9  <0.001 Subtypes Luminal A Luminal B HER2E Basal-like NOS Unknown  1144 744 159 186 - 224  38 (3%) 95 (11%) 41 (20%) 90 (33%) - 13 <0.001 *Chi square p value      52  Table 2.3 Differences in survival by LAG-3 expression and breast cancer subtypes, as assessed by interaction test  Survival endpoint Interaction test   P value* Breast cancer-specific survival  0.002 Relapse free survival  0.004 *interaction test was conducted using the likelihood ratio test to assess the interaction between a cox regression model with and without the LAG-3 x breast cancer subtype (luminal vs non-luminal) interaction term.                 53  Table 2.4 Multivariate analyses of LAG-3+iTILs in the whole cohort and among ER- patients for breast cancer-specific survival. A), only clinico-pathological parameters and LAG-3+iTILs as covariates, and B), clinico-pathological parameters and all immune biomarkers as co-variates.  A) Whole cohort  (# of events/n: 805/2702) Among ER-   (# of events/n: 283/764)   Hazard Ratio for BCSS (95% CI) LRT P-value Hazard Ratio for BCSS (95% CI) LRT P-value Age at diagnosis (Reference group:<50) ≥50 1.00 (0.87-1.16) 0.96 1.04 (0.82-1.32) 0.75 Tumor grade (Reference group: grade 1-2) Grade 3 1.83 (1.57-2.13) <0.0001 1.96 (1.42-2.72) <0.0001 Tumor size (Reference group: ≤2cm) >2 1.70 (1.47-1.97) <0.0001 1.54 (1.19-200) <0.0001 Lymphovascular invasion status (Reference group: negative) Positive 1.32 (1.12-1.56) 0.0007 1.45 (1.09-1.93) 0.01 Nodal status (Reference group: negative) Positive 2.00 (1.69-2.35) <0.0001 2.21 (1.68-2.91) <0.0001 LAG-3+iTILs (Reference group: 0) ≥1 0.71 (0.56-0.90) 0.003 0.50 (0.36-0.69) <0.0001  B) Whole cohort       (# of events/n: 697/2384) Among ER-      (# of events/n: 224/634)   Hazard Ratio for BCSS (95% CI) LRT P-value Hazard Ratio for BCSS (95% CI) LRT P-value Age at diagnosis (Reference group:<50) ≥50 0.99 (0.85-1.16) 0.92 0.97 (0.75-1.27) 0.85 Tumor grade (Reference group: grade 1-2) Grade 3 1.89 (1.61-2.23) <0.0001 2.10 (1.45-3.05) <0.0001 Tumor size (Reference group: ≤2cm) >2 1.61 (1.38-1.89) <0.0001 1.43 (1.08-1.91) 0.012 Lymphovascular invasion status (Reference group: negative) Positive 1.36 (1.14-1.63) 0.00063 1.41 (1.02-1.95) 0.035 Nodal status (Reference group: negative) Positive 1.95 (1.63-2.32) <0.0001 2.20 (1.61-3.01) <0.0001 CD8+iTILs (Reference group: 0) ≥1 0.98 (0.83-1.16) 0.84 1.04 (0.76-1.42) 0.81 PD-1+iTILs (Reference group: 0) ≥1 0.65 (0.47-0.91) 0.0082 0.50 (0.32-0.80) 0.002 PD-L1 (Reference group: <1) ≥1% 0.80 (0.58-1.10) 0.16 0.81 (0.52-1.28) 0.36 LAG-3+iTILs (Reference group: 0) ≥1 0.95 (0.71-1.28) 0.75 0.73 (0.49-1.09) 0.12 BCSS: Breast cancer-specific survival, LRT: Likelihood ratio test   54  Table 2.5 Associations among immune response biomarkers in breast cancer patients Immune markers LAG3+iTILs=0 LAG-3+iTILs  Positive (≥1) P-value (χ2) PD-L1* Negative Positive (≥1%)  2306 109  189 (7.6%) 125 (53.4%)  <0.0001 PD-1+iTILs* Negative Positive (≥1)  2382 97  170 (6.6%) 147 (60.7%)  <0.0001 CD8+iTILs* Negative Positive (≥1)  1775 693  69 (4%) 245 (26%)  <0.0001   PD-1+iTILs=0 PD-1+iTILs  Positive (≥1)  PD-L1 Negative Positive (≥1%)  2332 141  151 (6.1%) 89 (38.6%)  <0.0001 CD8+iTILs Negative Positive (≥1)  1785 762  63 (3.4 %) 178 (18.9 %)  <0.0001  PD-L1=0 PD-L1 Positive (≥1%)  CD8+iTILs Negative Positive (≥1)  1793 763  62 (3.3%) 174 (18.6%)  <0.0001 *Frequency in the whole cohort: PD-L1 ≥1%= 241/2918 (8.3%); PD-1+iTILs ≥1 = 246/2908 (8.5%); CD8+iTILs ≥1=1089/3403 (32%)  55  Table 2.6 Association of hematoxylin and eosin-stained (H&E) stromal TILs with PD-1+, LAG-3+, and PD-1+/LAG-3+ concurrent infiltration.   H&E sTILs categories*    Low (<10% sTILs)  n (%) Intermediate (≥10%- 50% sTILs)  n (%) High (LPBC) (>50% sTILs) n (%) Total  PD-1+iTILs(≥1) 34/225 (15%) 24/59 (41%) 9/10 (90%) n=67 LAG-3+iTILs(≥1) 11/205 (5%) 23/57 (40%) 4/9 (44%) n=38 PD1+/LAG3+** 8/197 (4%) 14/54 (26%) 4/9 (44%) n=26 Abbreviations: sTILs: stromal tumor infiltrating lymphocytes; LPBC: lymphocyte predominant breast cancer. * sTILs were assessed on the training set by the method of Salgado R et al.74  **Concurrent infiltration of PD-1+iTILs and LAG-3+iTILs in TMA cores.56  Chapter 3: TIM-3 expression in breast cancer SYNOPSIS Upregulation of additional immune checkpoint markers is one mechanism of resistance to current inhibitors that might be amenable to targeting with newer agents. T-cell Immunoglobulin and Mucin domain-containing molecule 3 (TIM-3) is an immune checkpoint receptor that is an emerging target for cancer immunotherapy. In this chapter, I investigated TIM-3 immunohistochemical expression in 3,992 breast cancer specimens assembled into tissue microarrays, linked to detailed outcome, clinico-pathological parameters and biomarkers including CD8, PD-1, PD-L1 and LAG-3.  We found that breast cancer patients with TIM-3+ intra-epithelial tumor-infiltrating lymphocytes (iTILs) (≥1) represented a minority of cases (11%), with a predilection for basal-like breast cancers (among which 28% had TIM-3+iTILs). The presence of TIM-3+iTILs highly correlated with hematoxylin and eosin-stained stromal TILs and with other immune checkpoint markers (PD-1+iTILs, LAG-3+iTILs and PD-L1+ tumors). In prognostic analyses, early breast cancer patients with TIM-3+iTILs had significantly improved breast cancer-specific survival. In multivariate analyses, the presence of TIM-3+iTILs was an independent favorable prognostic factor in the whole cohort as well as among ER negative patients. This study supports TIM-3 as a target for breast cancer immunotherapy.  3.1 Introduction The presence of small round dark mononuclear cells characteristic of tumor-infiltrating lymphocytes (TILs) on hematoxylin and eosin (H&E) - stained breast cancer specimens has garnered increased attention with the emergence of immune checkpoint inhibitors and has led to a 57  re-examination of the role of the immune system in breast tumors. Accumulating evidence shows that the presence of an immune response in breast cancers correlates with estrogen receptor negative (ER-) subtypes (i.e. the HER2 and basal-like intrinsic subtypes) among whom there is an association with favorable outcomes 79-81. In contrast, the more common ER+ breast cancer subtypes rarely display such heightened immune responses, which when present are associated with unfavorable prognosis78,219,220.  Immune checkpoint inhibitors targeting cytotoxic T-Lymphocyte-associated antigen 4 (CTLA-4),  programmed cell death-1 (PD-1) or its ligand (PDL-1) perform best in immunogenic cancers such as melanoma and non-small cell lung cancer138,221, but responses have recently been reported in triple negative / basal-like breast cancers205,222,223 (for reviews see refs. 224,225). However, even among such potentially immunogenic cancers, immune checkpoint inhibitors benefit only a relatively small number of patients138,224,226-230. As resistance may be due to the activation of alternative checkpoint pathways, additional immune checkpoint targets have become a subject of active research, including the T-cell Immunoglobulin and Mucin-domain- containing molecule 3 (TIM-3)155.       TIM-3 is an immune receptor discovered in 2002 that is expressed on a variety of immune cells including dendritic cells, macrophages, and T cells172,231,232.  TIM-3 mediates its suppressive activity on immune cells via its ligands that include phosphatidylserine, CEACAM-1 and the widely expressed ligand galectin-9 180,233. TIM-3 is expressed on activated T cells and its signaling on cytotoxic T cells leads to an exhausted phenotype, characterized by a reduction in proliferation, decreased production of effector cytokines and apoptosis of effector T cells234. In addition, TIM-3+ TILs can co-express PD-1, with blockade of both receptors leading to a more pronounced tumor regression than either agent alone, at least in pre-clinical studies 188,196. 58  Multiple studies have now reported on the presence of TIM-3+TILs in human tumors194,235-238. However, in breast cancer, TIM-3+TILs have been evaluated by immunohistochemistry in a limited number of patients, with one recent study reporting positive associations with lymph node metastases239,240. The objective of our study is to evaluate the expression of TIM-3 on TILs in a large series of breast cancers powered for multivariate correlation with clinico-pathological parameters, survival, and other important immune biomarkers.   3.2 Material and methods 3.2.1 Study cohorts  The study cohorts were the same as in Chapter 2. The initial set consisting of 330 breast cancer patients was used to finalize biomarker staining and interpretation conditions for an initial analysis of TIM-3. These patients were diagnosed with invasive breast cancer at University of British Columbia hospitals between 1989 and 2002 and have been previously described206. A detailed description of the validation set consisting of 3,992 breast cancer patients has been previously published107,241. In brief, newly diagnosed invasive breast cancers from centres across the province of British Columbia performing breast cancer excision surgery, referred to the British Columbia cancer agency between 1986 and 1992 and for which both blocks from a central estrogen receptor testing laboratory and detailed de-identified clinico-pathologic, treatment and outcome data collected by the British Columbia cancer agency breast cancer outcomes unit were available were assembled into 17 single core tissue microarray blocks. None of these patients (training and validation cohorts) received neoadjuvant treatment. The median follow-up for both cohorts is 12.6 years. The Clinical Research Ethics Board of the University of British Columbia and the British 59  Columbia Cancer Agency Breast Cancer Outcomes Unit approved the access to the samples and corresponding de-identified outcome data.   3.2.2 Immunohistochemistry and scoring  Tissue microarrays (TMAs) were built from formalin-fixed paraffin-embedded primary excision specimens from patients in the training and validation cohorts and represented as 0.6mm cores across 3 blocks for the training cohort and 17 blocks for the validation cohort. These TMAs have been previously stained and scored for multiple biomarkers including ER, PR, HER2, Ki67, EGFR, CK5/6, CD8, LAG-3, PD-1 and PD-L1242. Breast cancer intrinsic subtypes were previously determined from both cohorts by immunohistochemistry (IHC) benchmarked against a gene expression gold standard (the PAM50 intrinsic subtype classifier) 208 . Briefly, ER+ (≥1%) or PR+ (≥1%), HER2– (including IHC 2+ cases that were HER2– by fluorescence in situ hybridization) and low (<14%) Ki67 were defined as Luminal A; hormone receptor positive cases which were also either HER2+ or had high Ki67 were defined as Luminal B; HER2+/ER–/PR– cases were defined as HER2E, and triple negative cases that were positive for EGFR+ or CK5/6+ were defined as basal-like.  Overall stromal TILs were scored on H&E-scanned images of the TMA cores using the assessment recommendations of the International TILs Working Group74, whereby stromal TILs are scored as the percentage of intertumoral stromal surface area (i.e. excluding areas occupied by carcinoma cells) containing mononuclear lymphocytic infiltrates. TIM-3 immunohistochemistry was conducted with anti-TIM-3 rabbit monoclonal antibody clone D5D5R from Cell Signaling (Cat# 45208) as employed in other publications 231,239,243,244, here using a Ventana Ultra automated stainer (Ventana Medical Systems) in concordance with manufacturer’s protocol. In brief, slides underwent antigen retrieval with Standard Cell 60  Conditioning 1 reagent (Ventana Medical Systems) followed by 60 minutes of primary antibody incubation (applied at 1:50 dilution) with no heat, and visualized using a chromoMap DAB detection kit (Ventana Medical Systems). Membranous staining in tonsil tissue served as a positive control in each staining run. TIM-3+ lymphocytes scores were reported as absolute counts per TMA core for intra-epithelial or stromal locations. TIM-3+ intra-epithelial lymphocytes (TIM-3+iTILs) were defined as TIM-3+ lymphocytes located within carcinoma nests whereas TIM-3+ stromal lymphocytes (TIM-3+sTILs) were those not in direct contact with the carcinoma nest.  3.2.3 Statistical analysis  IBM SPSS software (version 24.0) and R (version 3.3.2) were used to conduct all the statistical analyses. TIM-3+ iTILs scores were dichotomized ≥1 (as positive) vs. 0 (as negative). In addition, TIM-3 expression on other immune cells (non-lymphocytes) was assessed, but as only 1% of cases were positive on the training set this staining pattern was not further analyzed.  For prognostic analyses, the primary end-point, breast cancer-specific survival, was defined as the time from date of diagnosis to date of death attributed to breast cancer. Patients were censored at death from another cause or if alive at end of follow-up. Relapse-free survival and overall survival were secondary end-points. Relapse-free survival was defined as time from date of diagnosis to date of any type of breast cancer relapse (local, regional, distant, or contralateral) and overall survival as time from date of diagnosis to date of death, irrespective of the cause of death. In addition, patients were censored if they had not died from breast cancer or if they had not relapsed at the end of follow-up time for relapse-free survival analyses. Correlation with survival was conducted using Kaplan-Meier curves, log-rank test and Cox regression models. Proportional hazard assumptions were assessed by visual examinations of Kaplan-Meier plots. In the case where the proportional hazard assumption was violated (for the basal-like subgroup, beyond 5 years), the 61  follow-up time and hazard ratios were modified accordingly in the multivariate Cox regression model. The effect size was adjusted in multivariate Cox regression models by taking into account significant clinicopathological parameters (age, tumor grade, tumor size, lymphovascular invasion and nodal status).  Clinico-pathological and prognostic associations for TIM-3+iTILs were analyzed first on the training cohort (n=330) and further tested on the validation cohort (n=3,992) in a pre-specified formal written statistical plan, presented at the British Columbia Cancer Agency Breast Cancer Outcomes Unit. Furthermore, half of the validation cohort served for a training and a validation approach specifically for correlations and combinatorial analyses among immune biomarkers (TIM-3, PD-L1, PD-1, LAG-3, CD8) due to the low number of positive cases observed in the training set. In addition, 40% of cases in the validation cohort were considered TIM-3+sTIL positive based on a ≥1 positive TIL cut-point. A cut-point of ≥2 positive TILs for TIM-3+sTILs, representing 20% of cases, was selected following testing of various cut-points (≥1, ≥2) based on the distribution on half of the validation cohort set in prognostic analyses. In these cases, a pre-specified written statistical plan for validation on the other half of the set was presented prior to statistical analyses (see Appendix C). Prognostic analyses of co-infiltrated immune checkpoint markers were nevertheless considered exploratory. All statistical tests performed were two-sided at α=0.05.      3.3 Results 3.3.1 Distribution of TIM-3+TILs in breast cancers  To define staining conditions and interpretation, we conducted an initial evaluation of TIM-3 staining and correlation with clinico-pathological parameters on a tissue microarray 62  consisting of 330 breast cancer patients (representing a training set). We observed 12% of breast cancer cases with TIM-3+ intratumoral tumor infiltrating lymphocytes (≥1 iTIL per 0.6 mm diameter core) whereas stromal TIM-3+sTILs (≥1) were present in 48% of cases (Results for TIM-3+iTILs shown in Table 3.1). We then proceeded with TIM-3 staining on a TMA comprising an independent cohort of 3,992 breast cancer cases, of which 3,148 cases were interpretable for TIM-3 immunohistochemistry staining (Figure 3.1). The results were consistent with the training set as 11% of cases had ≥1 TIM-3+ iTILs and 40% of cases had ≥1 TIM-3+ sTILs (Figure 3.2). TIM-3 expression on macrophages was only observed in 1% of cases and was not analyzed further. As there were a large number of cases with TIM-3+sTILs, a cut-off for dichotomization of ≥2 sTIL/0.6 mm core, a level reached in 20% of breast cancers, was selected based on analyses of Kaplan Meier curves of different TIM-3+sTILs cut points (as described in Methods: Statistics). However, TIM-3+iTILs were selected as the primary analysis parameter, to allow comparison with previously published immune biomarkers in this breast cancer cohort96,242 . 3.3.2 The presence of TIM-3+iTILs in breast cancer is associated with unfavorable clinico-pathological factors  Consistent with the results in the initial cohort of 330 patients, breast cancer cases with TIM-3+iTILs in the validation cohort were significantly associated with younger age at presentation, higher grade, hormone receptor (ER/PR) negativity, and high Ki67 proliferation index [defined as ≥14%] (Table 3.2). In addition, the presence of TIM-3+iTILs was much more common in the basal-like subtype relative to other subtypes (28% in basal-like vs 6% in luminal A). The results for TIM-3+sTILs reflected similar associations with clinico-pathological parameters to TIM-3+iTILs findings (Table 3.3).  63  3.3.3 TIM-3+iTILs correlate with the presence of other immune checkpoint markers (LAG-3, PD-1, PD-L1) and overall H&E sTILs Because this large cohort had been previously assessed for key immune biomarkers including PD-1, PD-L1 and LAG-3, we were able to analyze their correlations with TIM-3. In addition, we scored overall H&E stromal TILs to allow a parallel evaluation with immune checkpoint markers. We found that breast tumors with TIM-3+iTILs were highly significantly associated with the presence of additional immune checkpoint markers (Table 3.4). Indeed, nearly half of breast cancers that were positive for PD-L1 or PD-1+iTILs or LAG-3+iTILs were also infiltrated with TIM-3+iTILs in the same 0.6 mm TMA core. However, only 3% (91/2736 interpretable cases) expressed all three immune checkpoint markers (TIM-3+/PD-1+/LAG-3+) (Table 3.4) when assessed by this method. No particularly unique association pattern was observed between the presence of TIM-3+iTILs and any of the other individual immune checkpoint markers tested, suggesting that the TIM-3 checkpoint expression on TILs occurs in tumors containing T cells positive for other exhausted markers. Furthermore, we found that all immune checkpoint markers correlated positively (p<0.001) with H&E sTILs (Figure 3.3). In this cohort, less than 1% of cases were categorized as lymphocyte-predominant breast cancer (LPBC, defined as ≥50% H&E sTILs). 3.3.4 TIM-3+iTILs are associated with good prognosis in early breast cancer In univariate analyses, the presence of TIM-3+iTILs in early breast tumors was associated with improved breast cancer-specific survival (BCSS) in the whole cohort (HR: 0.76, 95%CI 0.61-0.96, Log Rank p=0.02) (Figure 3.4). When breast cancer subtypes were stratified in the analysis, only HER2+ and basal-like breast cancer patients with TIM-3+iTILs displayed 64  significantly improved BCSS (HER2+: HR: 0.27, 95%CI 0.10-0.72, Log Rank p=0.005; Basal-like: HR: 0.48, 95%CI 0.29-0.78, Log Rank p=0.003) (Figure 3.4). These results were similar using overall survival and relapse-free survival secondary endpoints (Figure 3.5 for overall survival; Figure 3.6 for relapse-free survival). In contrast, the presence of TIM-3+sTILs had a trend for favorable prognosis for BCSS and relapse-free survival and reached significance in the whole cohort for overall survival (Figure 3.7 for BCSS; Figure 3.8 for overall survival and Figure 3.9 for relapse-free survival).  In multivariate analyses that included H&E sTILs as a covariate, the presence of TIM-3+iTILs remained a favorable prognostic factor in the whole cohort and among ER- breast cancer patients (Table 3.5 – Whole cohort: HR: 0.64, 95%CI 0.48-0.85, p=0.001; ER-: HR: 0.58, 95%CI 0.39-0.86, p=0.004, Basal-like: HR: 0.58, 95%CI 0.32-1.03, p=0.052). Similar findings were observed for TIM-3+sTILs albeit not reaching significance for basal-like breast cancer patients (Table 3.6). We also found that ER- breast cancer patients with tumors that were co-infiltrated with TIM-3+, PD-1+ and LAG-3+ TILs had a significant improved breast cancer specific survival, in univariate and multivariate analyses, relative to patients with a single positive, dual positive, or complete absence of these three immune checkpoint markers (Figure 3.10, Table 3.7).   3.4 Discussion We report the first study of TIM-3 expression in a large (>1000 case) series of early breast cancers. TIM-3 expression in this cohort was restricted to tumor-infiltrating lymphocytes and was present in about 12% of cases when 0.6 mm cores were evaluated for expression on intra-epithelial TILs, or 20% of cases when assessed on stromal TILs. The presence of TIM-3+iTILs was 65  associated with younger age, high grade and high Ki67 proliferation index and was enriched in the basal-like breast cancer subtype. Moreover, TIM-3+iTILs highly correlated with co-infiltration of additional immune checkpoint markers PD-L1 (on carcinoma cells), PD-1 and LAG-3+ (on TILs). In prognostic analyses, early breast cancer patients with TIM-3+ iTILs had significantly improved survival for all assessed endpoints, as compared to patients whose tumors lacked TIM-3+iTILs. In multivariate analyses, the prognostic effect was maintained in the whole cohort as well as among ER- and basal-like breast cancer patients.    Studies from our group and from others have been consistent in finding that the presence of immune checkpoint markers on intra-epithelial TILs in breast tumors is an uncommon event, and mostly restricted to ER- breast cancers163,239,242. However, TILs positive for immune checkpoint markers are able to discriminate breast cancer patients with favorable survival, consistent with an active anticancer immune microenvironment. Indeed, we found that breast tumors infiltrated with TIM-3+iTILs highly correlate with tumors positive for other checkpoint markers (PD-1, PD-L1 and LAG-3). Results are consistent with other reported studies and imply that the expression of multiple different immune checkpoints can occur during tumor progression, reflecting an ongoing battle between cancer cells and the immune system212,245. In our cohort, coexpression of TIM-3 with PD-1 and LAG-3 is associated with a particularly favorable prognosis, perhaps reflecting an underlying robust immune recognition of the cancer cells that is difficult for the tumor to evade. Furthermore, other studies have reported TIM-3 expression on carcinoma cells to be associated with poor prognosis (for meta-analysis see ref. 246), which we did not observe in our large breast cancer cohort. These apparently conflicting results may be due to the different types of tumor and possible confounding by stage or other factors, as the smaller studies in other tumors were not powered for multivariate analyses. 66  Strengths of our study include the use of a large cohort of early breast cancer patients, treated consistently according to provincial guidelines, linked to detailed long-term outcome data and assessed using a training and validation approach to biomarker interpretation. Some limitations include, first, the necessity in such a large series to rely on TMA cores, representing a 0.28 mm2 surface area sampling of a tumor for assessment of the tumor immune microenvironment. Second, infiltration of TILs bearing multiple immune biomarkers could only be inferred from single stains and therefore does not directly identify co-expression on the same lymphocyte. Third, breast cancer patients in the cohort received what would now be considered older treatments (predating trastuzumab, taxanes and aromatase inhibitors) which may affect extrapolation of some of the observed prognostic and predictive associations to more contemporary treatment regimens.     Accumulating evidence suggests resistance to anti- CTLA-4 or anti- PD-1/PD-L1 inhibitors can occur in otherwise immunogenic cancers through compensatory upregulation of additional immune checkpoints170,245. TIM-3 has recently emerged as a target for cancer immunotherapy following pre-clinical studies suggesting its non-redundant functions in comparison to the better-characterized checkpoint markers PD-1/PD-L1, and efficacious treatment synergy when TIM-3 is targeted in combination with anti-PD1/PDL1 antibodies188,196,197,202. Although ER- breast cancer, in particular basal and triple negative breast cancer, is considered the most immunogenic subtype, reports from immune checkpoint inhibitor clinical trials are not as encouraging. Early reports suggest metastatic breast cancer patients may benefit most from PD-1/PD-L1 blockade monotherapy in the first-line setting, or in combination with chemotherapy agents for second or third-line therapy with an objective response rate ranging from 10%-40% 205,222,223 (for review, see ref. 224). The findings from our study imply that TIM-3 inhibitors could potentially help to treat PD-1 refractory or metastatic tumors. Currently, four early phase clinical 67  trials testing the efficacy of anti- TIM-3 in combination with anti- PD-1/PDL1 in advanced tumors have opened [NCT03066648, NCT02608268, NCT02817633, and NCT03099109]. Our data support that this appears to be a relevant combinatorial strategy to assess in breast cancer, particularly in patients with non-BRCA mutated basal-like tumors, an aggressive subtype for which targeted therapies are not currently available. 68    Figure 3.1TIM-3 staining in breast cancer patients Representative images of TIM-3 immunohistochemistry staining on formalin-fixed paraffin-embedded breast cancer tissue microarray cores. (A) TIM-3 negative staining; (B) TIM-3+ intra-epithelial tumor-infiltrating lymphocytes; (C) TIM-3+ stromal tumor-infiltrating lymphocytes. Red arrows indicate positive TIM-3+ intra-epithelial (in B) or stromal (in C) tumor-infiltrating lymphocytes.           A B C 69   Figure 3.2 Distribution of TIM-3+ tumor-infiltrating lymphocytes in the whole (validation) cohort Histograms illustrating the absolute count per TMA core for TIM-3+ intra-epithelial TILs (A) and stromal TILs (B). Scores were available for 3,148 cores for which immunohistochemistry staining was interpretable. 70   Figure 3.3 H&E sTIL correlation with immune biomarkers. Scatter plots depicting the correlation between the percentage of H&E sTILs on the x-axis and immune biomarker scores (A, PD-L1; B, PD-1+iTILs; C, LAG-3+iTILs; D, TIM-3+iTILs) per TMA core on the y-axis, with the corresponding Spearman rho and p values.71    Figure 3.4 TIM-3+intra-epithelial tumor-infiltrating lymphocytes association with breast cancer-specific survival in the whole (validation) cohort and by breast cancer subtype. Kaplan Meier curves of breast cancer-specific survival in breast cancer patients stratified by the presence or absence of TIM-3+iTILs. Kaplan-Meier curves in (A) the whole cohort, (B) Luminal A cases, (C) Luminal B, (D) HER2+ and (E) basal-like cases are shown with their corresponding numbers of patients, events and log rank p values. The number of patients still at risk at the end of each 5 years of follow-up is shown at the bottom of each panel.  Luminal B HER2+Basal-likeWhole cohortHR: 0.764, 95%CI (0.61-0.96), Log Rank p=0.02Luminal AHR: 1.155, 95%CI (0.71-1.89), Log Rank p=0.58HR: 0.68, 95%CI (0.44-1.05), Log Rank p=0.09 HR: 0.27, 95%CI (0.10-0.72), Log Rank p=0.005HR: 0.48, 95%CI (0.29-0.78), Log Rank p=0.00372   Figure 3.5 Overall survival for TIM-3+intra-epithelial tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes. Kaplan Meier curves (KM) of overall survival in breast cancer patients stratified by the presence or absence of TIM-3+iTILs. KM curves in the whole cohort (A), HER2+ (B) and basal-like (C) are shown with corresponding number of patients, events and log rank p values.               73     Figure 3.6 Relapse-free survival for TIM-3+ intra-epithelial tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes. Kaplan Meier (KM) curves of relapse-free survival in breast cancer patients stratified by the presence or absence of TIM-3+iTILs. KM curves in the whole cohort (A), HER2+ (B) and basal-like (C) are shown with corresponding number of patients, events and log rank p values.               74    Figure 3.7 TIM-3+ stromal tumor-infiltrating lymphocytes association with breast cancer-specific survival in the whole cohort and by breast cancer subtype. Kaplan Meier curves of breast cancer-specific survival in breast cancer patients stratified by the presence or absence of TIM-3+sTILs. KM curves in (A) the whole cohort, (B) Luminal A cases, (C) Luminal B, (D) HER2+ and (E) basal-like cases are shown with their corresponding numbers of patients, events, hazard ratios and log rank p values.   75   Figure 3.8 Overall survival for TIM-3+ stromal tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes. Kaplan Meier curves (KM) of overall survival in breast cancer patients stratified by the presence or absence of TIM-3+sTILs. KM curves in the whole cohort (A), HER2+ (B) and basal-like (C) are shown with corresponding numbers of patients, events, hazard ratios and log rank p values.         76   Figure 3.9 Relapse-free survival for TIM-3+ stromal tumor-infiltrating lymphocytes in the whole cohort, in HER2+ and basal-like breast cancer subtypes.  Kaplan Meier (KM) curves of relapse-free survival in breast cancer patients stratified by the presence or absence of TIM-3+sTILs. KM curves in the whole cohort (A), HER2+ (B) and basal-like (C) are shown with corresponding numbers of patients, events, hazard ratios and log rank p values. 77   Figure 3.10 Prognostic value of TIM-3, PD-1 and LAG-3+iTILs co-infiltration among ER negative breast cancer patients. Kaplan Meier curve of breast cancer-specific survival among ER- breast cancer patients stratified by the presence or absence of one or more immune checkpoint markers is shown with corresponding number of patients, events and a log rank p value. Blue: All negative (TIM3-/PD1-/LAG3-), green: Single positive (TIM3+ or PD1+ or LAG3+), grey: Double positive (TIM3+/PD1+ or TIM3+/LAG3+ or PD1+/LAG3+), purple: All positive (TIM3+/PD1+/LAG3+).    78  Table 3.1 Association of TIM-3+ iTILs with clinico-pathological parameters on the initial training cohort.  Initial training cohort (n=330; interpretable staining n=234)  Parameters  Negative   n=206  TIM-3+iTILs (≥1)  n=28 (12%)     P value* Age at diagnosis (years) <50 ≥50   71 135   13 (15%) 15 (10%) 0.22 Tumor size (cm) ≤2 >2  118 88  13 (10%) 15 (15%) 0.31 Grade 1&2 3 Missing  122 83 1  8 (6%) 19 (19%) 1 0.004 Ki67 Negative (<14%) Positive (≥14%) Missing  121 82 3  9 (7%) 19 (19%) 0 0.006 ER Negative Positive (>1%)  40 166  15 (27%) 13 (7%) <0.001 Subtypes Luminal A Luminal B HER2E Basal-like Unknown  104 49 8 18 27  8 (7%) 6 (11%) 3 (27%) 7 (28%) 4 <0.001 * Chi-square p value       79   Table 3.2 TIM-3+ intra-epithelial tumor-infiltrating lymphocytes association with clinico-pathological parameters in breast cancer   Parameters  Negative   N=2816  TIM-3+iTILs ≥1  N=332 (11%)  P-value* (χ2) Age at diagnosis (years)  <50 ≥50   787 2029   119 (13%) 213 (10%) 0.003 Tumor size (cm) ≤2 >2  1467 1334  163 (10%) 166 (11%) 0.331 Grade 1&2 3 Unknown  1286 1414 116  96 (7%) 219 (13%) 12 <0.001 Ki67 Negative (<14%) Positive (≥14%) Unknown  1469 1085 262  100 (6%) 209 (16%) 23 <0.001 ER Negative Positive (>1%) Unknown  716 2091 9  159 (18%) 172 (8%) 1 <0.001 Subtypes Luminal A Luminal B HER2E Basal-like Triple negative, non-basal Unknown  1209 631 184 205 182 240  73 (6%) 80(11%) 25 (12%) 81 (28%) 40 (18%) 17 <0.001 * Chi-square test      80  Table 3.3 Association of TIM-3+ stromal tumor-infiltrating lymphocytes with clinico-pathological parameters.   Validation set (N =3,148)  Parameters  Low (<2)  N=2516  TIM-3+sTILs (≥2)  N=632 (20%)     P value* Age at diagnosis (years) <50 ≥50   683 1833   223 (25%) 409 (18%) <0.001 Tumor size (cm) ≤2 >2 Missing  1313 1189 14  317 (19%) 311 (21%) 4 0.370 Grade 1&2 3 Missing  1133 1276 107  249 (18%) 357 (22%) 26 0.009 Ki67 Negative (<14%) Positive (≥14%) Missing  1307 982 227  262 (17%) 312 (24%) 58 <0.001 ER Negative Positive (>1%) Missing  649 1858 9  226 (26%) 405 (18%) 1 <0.001 Subtypes Luminal A Luminal B HER2E Basal-like Triple negative, non-basal Unknown  1074 561 150 207 172  352  208 (16%) 150 (21%) 59 (28%) 79 (28%) 50 (23%)  86 <0.001 * Chi-square p value    81   Table 3.4 Association of TIM-3+ intra-epithelial tumor-infiltrating lymphocytes with other immune biomarkers in breast cancer  Immune biomarkers TIM-3+iTILs=0 (n=2816) TIM-3+iTILs ≥1 (n=332) P-value (χ2) PD-L1* Negative Positive (≥1%)  2374 133  215 (8%) 97 (42%)  <0.0001 PD-1+iTILs* Negative Positive (≥1)  2388 127  198 (8%) 113 (47%)  <0.0001 LAG-3+iTILs* Negative Positive (≥1)  2344 175  169 (7%) 146 (45%) <0.0001 CD8+iTILs* Negative Positive (≥1)  1881 778  111 (6%) 213 (22%)  <0.0001 *Frequency in the whole cohort: PD-L1 ≥ 1%= 241/2918 (8.3%); PD-1+iTILs ≥1 = 246/2908 (8.5%); LAG-3+iTILs= 327/2921 (11%), from Burugu S et al.242 CD8+iTILs ≥1=1089/3403 (32%) from Liu S et al.96         82  Table 3.5 Multivariate analyses of TIM-3+ intra-epithelial tumor-infiltrating lymphocytes in the whole cohort, among estrogen receptor negative and in basal-like patients for breast cancer-specific survival including H&E sTILs as a covariate.   Whole cohort      (# of events/n: 705/2379)   Hazard Ratio for BCSS (95% CI) LRT P-value Age at diagnosis (Reference group:<50) ≥50 1.06 (0.86-1.30) 0.61 Tumor grade (Reference group: grade 1-2) Grade 3 1.49 (1.26-1.77) <0.001 Tumor size (Reference group: ≤2cm) >2 1.63 (1.39-1.91) <0.001 Lymphovascular invasion status (Reference group: negative) Positive 1.33 (1.11-1.60) 0.002 Nodal status (Reference group: negative) Positive 2.29 (1.86-2.82) <0.001 Adjuvant systemic therapy (Reference group: no AST) TAM only Chemo only TAM+chemo 0.73 (0.57-0.93) 0.74 (0.56-0.99) 0.69 (0.49-0.97) 0.05   Breast cancer subtypes (Reference group: Luminal A)  Luminal B/Ki67  Luminal/HER2+ HER2+ Basal-like  1.81 (1.50-2.19) 2.16 (1.64-2.84) 2.54 (1.93-3.35) 2.28 (1.74-2.99)   <0.001 H&E sTILs (10% increments)  0.98 (0.98-0.99) <0.001 TIM-3+iTILs (Reference group: 0)  ≥1  0.64 (0.48-0.85)  0.001 Among ER-*  (# of events/n: 255/705)   Hazard Ratio for BCSS (95% CI)  LRT P-value Age at diagnosis (Reference group:>50) ≥50 0.90 (0.66-1.23) 0.50 Tumor grade (Reference group: grade 1-2) Grade 3 1.91 (1.35-2.70) <0.001 Tumor size (Reference group: ≤2cm) >2 1.62 (1.23-2.12) 0.001 83  Lymphovascular invasion status (Reference group: negative) Positive 1.32 (0.99-1.77) 0.06 Nodal status (Reference group: negative) Positive 2.44 (1.76-3.38) <0.001 Adjuvant systemic therapy (Reference group: no AST) TAM only Chemo only TAM+chemo 0.89 (0.59-1.34) 0.81 (0.55-1.19) 1.02 (0.59-1.75) 0.64    H&E sTILs (10% increments)   0.98(0.97-0.99)  0.002 TIM-3+iTILs (Reference group: 0)  ≥1  0.58 (0.39-0.86)  0.004  Among basal-like (# of events/n: 94/263)   Hazard Ratio for BCSS (95% CI) LRT P-value Age at diagnosis (Reference group:<50) ≥50 0.86 (0.50-1.46) 0.57 Tumor grade (Reference group: grade 1-2) Grade 3 1.39 (0.72-2.71) 0.31 Tumor size (Reference group: ≤2cm) >2 1.40 (0.91-2.16) 0.13 Lymphovascular invasion status (Reference group: negative) Positive 1.31 (0.82-2.10) 0.26 Nodal status (Reference group: negative) Positive 2.01 (1.19-3.38) 0.008 Adjuvant systemic therapy (Reference group: no AST) TAM only Chemo only TAM+chemo 1.65 (0.80-3.42) 1.21 (0.65-2.25) 1.40 (0.47-4.19) 0.60     H&E sTILs (10% increments)  0.97 (0.95-0.99) 0.002 TIM-3+iTILs (Reference group: 0)  ≥1  0.58 (0.32-1.03)  0.052 * including HER2 positive and negative   84  Table 3.6 Multivariate analyses of TIM-3+stromal tumor-infiltrating lymphocytes in the whole cohort, among estrogen receptor negative and in basal-like patients for breast cancer-specific survival.   Whole cohort      (# of events/n: 705/2379)   Hazard Ratio for BCSS (95% CI) LRT P-value Age at diagnosis (Reference group:>50) ≥50 1.05 (0.85-1.29) 0.66 Tumor grade (Reference group: grade 1-2) Grade 3 1.48 (1.25-1.74) <0.001 Tumor size (Reference group: ≤2cm) >2 1.63 (1.39-1.91) <0.001 Lymphovascular invasion status (Reference group: negative) Positive 1.34 (1.11-1.60) 0.001 Nodal status (Reference group: negative) Positive 2.29 (1.86-2.81) <0.001 Adjuvant systemic therapy (Reference group: no AST) TAM only Chemo only TAM+chemo 0.73 (0.58-0.93) 0.75 (0.57-1.00) 0.69 (0.49-0.98) 0.06 Breast cancer subtypes (Reference group: Luminal A)  Luminal B/Ki67  Luminal/HER2+ HER2+ Basal-like  1.80 (1.49-2.18) 2.21 (1.68-2.90) 2.58 (1.96-3.41) 2.15 (1.64-2.80)   <0.001 H&E sTILs (10% increments)  0.98 (0.97-0.99) <0.001 TIM-3+sTILs (Reference group: <2)  ≥2  0.79 (0.65-0.96)  0.017 Among ER-  (# of events/n: 255/705)   Hazard Ratio for BCSS (95% CI)  LRT P-value Age at diagnosis (Reference group:>50) ≥50 0.93 (0.68-1.27) 0.63 Tumor grade (Reference group: grade 1-2) Grade 3 1.87 (1.32-2.64) <0.001 Tumor size (Reference group: ≤2cm) >2 1.57 (1.20-2.06) 0.001 85  Lymphovascular invasion status (Reference group: negative) Positive 1.38 (1.03-1.85) 0.029 Nodal status (Reference group: negative) Positive 2.40 (1.73-3.32) <0.001 Adjuvant systemic therapy (Reference group: no AST) TAM only Chemo only TAM+chemo 0.93 (0.62-1.40) 0.84 (0.57-1.23) 1.05 (0.61-1.81) 0.73   H&E sTILs (10% increments)  0.98 (0.97-0.99) <0.001 TIM-3+sTILs (Reference group: <2)  ≥2  0.73 (0.53-1.00)  0.047  Among basal-like  (# of events/n: 94/263)   Hazard Ratio for BCSS (95% CI) LRT P-value Age at diagnosis (Reference group:<50) ≥50 0.86 (0.50-1.46) 0.57 Tumor grade (Reference group: grade 1-2) Grade 3 1.40 (0.72-2.71) 0.30 Tumor size (Reference group: ≤2cm) >2 1.36 (0.88-2.11) 0.16 Lymphovascular invasion status (Reference group: negative) Positive 1.41 (0.88-2.25) 0.15 Nodal status (Reference group: negative) Positive 2.06 (1.22-3.48) 0.006 Adjuvant systemic therapy (Reference group: no AST) TAM only Chemo only TAM+chemo 1.65 (0.79-3.43) 1.20 (0.64-2.24) 1.17 (0.39-3.50) 0.63   H&E sTILs (10% increments)  0.97 (0.95-0.99) <0.001 TIM-3+sTILs (Reference group: <2)  ≥2  1.05 (0.61-1.78)  0.87    86  Table 3.7 Multivariate analyses of TIM-3/PD-1/LAG-3+iTILs among ER negative breast cancer patients for breast cancer-specific survival.   *Wald-test Among ER-  (# of events/n: 249/686)   Hazard Ratio for BCSS (95% CI) P-value* Age at diagnosis (Reference group:<50) ≥50 0.89 (0.64-1.24) 0.49 Tumor grade (Reference group: grade 1-2) Grade 3 2.22 (1.56-3.16) <0.001 Tumor size (Reference group: ≤2cm) >2 1.50 (1.14-1.98) 0.004 Lymphovascular invasion status (Reference group: negative) Positive 1.355 (0.998-1.840) 0.052 Nodal status (Reference group: negative) Positive 2.654 (1.887-3.731) <0.001 Adjuvant systemic therapy (Reference group: no AST) TAM only Chemo only TAM+chemo 0.851 (0.566-1.279) 0.739 (0.495-1.103) 1.193 (0.683-2.084) 0.436   TIM-3/PD-1/LAG-3+ iTILs (Reference group:  TIM3-/PD1-/LAG3-)     TIM3-/PD1+/LAG3-  TIM3-/PD1-/LAG3+  TIM3-/PD1+/LAG3+  TIM3+/PD1-/LAG3-  TIM3+/PD1+/LAG3-  TIM3+/PD1-/LAG3+  TIM3+/PD1+/LAG3+   0.499 (0.251-0.989)  0.498 (0.289-0.861)  0.586 (0.299-1.148)  0.615 (0.314-1.203)  0.348 (0.086-1.410)  0.959 (0.519-1.772)  0.165 (0.073-0.375)  0.046  0.012  0.119  0.155  0.139  0.893  <0.001 87  Chapter 4: Quantitative in situ multiplex immune profiling of breast cancer patients using digital spatial profiling technology SYNOPSIS:  Predictive biomarkers of immunotherapies are the focus of intensive research that has dramatically increased in a short amount of time. Immune biomarkers presented in Chapter 2 and 3 represent only some of the immunotherapy targets that have advanced from pre-clinical to clinical phase. Studies show that more than one biomarker will likely be required to identify immunotherapy-responsive tumors. The tumor immune microenvironment is complex and requires a multiplex detection system to distinguish the various immune cell populations. In this chapter, I profiled the tumor immune microenvironment of two breast cancer cohorts (a total of 59 breast cancers enriched for the basal-like subtype) using a novel multiplex technology by Nanostring called Digital Spatial Profiling (DSP). DSP enables the characterization and expression profiling of breast cancer tumors, in my work using a 31-marker immuno-oncology panel. I then validated DSP digital counts for CD8 and PD1 by immunohistochemistry, and CD45 digital counts by assessing hematoxylin and eosin-stained stromal tumor-infiltrating lymphocyte counts. Furthermore, I was able to identify a profile of patients with favorable prognosis.  Lastly, I identified a 4-biomarker signature that was indicative of a pre-existing intratumoral immune response. A proposal to evaluate the capacity of the 4-biomarker signature to predict response to immune-modulating chemotherapy is presented at the end of the chapter.  88  4.1 Introduction Immunotherapy has drastically changed the cancer treatment landscape, including now in breast cancer. Clinical trials evaluating immune checkpoint inhibitors in breast cancer have mostly targeted patients with basal-like triple negative breast cancers as it represents the most immunogenic subtype [as reviewed by Wein et al.147 ]. A recent phase III clinical trial (IMpassion130) of a PD-L1-targeted immune checkpoint inhibitor reported improved progression-free survival in treatment-naive metastatic basal-like breast cancer patients148. Predictive biomarkers of response to immune checkpoint inhibitors are needed (to avoid unnecessary toxicities and high costs) but the identity of these biomarkers is still a subject for debate. Genomic characteristics of tumors that are associated with response include the presence of high microsatellite instability (MSI-H), mismatch-repair deficiency (dMMR), a high mutational burden, and copy number alterations247-249. However, MSI-H/dMMR tumors are rare, especially in breast cancer (~2% of breast cancers250 ) and no standardized means of measurement nor cut-offs for mutational burden are yet firmly established.  Tumors heavily-infiltrated with immune cells are more likely to respond to immune checkpoint inhibitors as demonstrated in earlier studies212,215,251 [reviewed by Chen DS and Mellman I212]. Anti-tumor immunity is mainly mediated by T cells, which represent a heterogeneous population that includes effector, exhausted and regulator phenotypes252. In addition, other immune cells play key roles in anti-tumor immunity such as Natural Killer cells and macrophages52,252. The concurrent infiltration of the various immune cell populations in a patient’s tumor, as well as their intratumoral localization pattern, may be associated with impeded or improved responses to immunotherapies. Nonetheless, to analyze multiple immune biomarkers concurrently in a patient’s tumor tissue represents a technical challenge; methods that 89  have been tried include sequential staining of multiple antibodies or digital segmentation of tumors for image analyses253,254.  In my study, I employed a novel quantitative spatially-resolved multiplexed antibody-based method called digital spatial profiling (DSP) using NanoString’s GeoMx™ technology. This method allows for simultaneous quantitative measurement of an immuno-oncology panel (representing 31 biomarkers) in user-defined areas of formalin-fixed paraffin-embedded breast tumor tissues. The 31 biomarkers included immune cell type identification markers (CD14, CD19, CD3, CD4, CD45, CD45RO, CD56, CD68, CD8A, FOXP3, GZMB, CD20, PD1, PD-L1, VISTA), tissue and cancer biomarkers (Pan-cytokeratins, PTEN, β-catenin, Ki67, Bcl-2) and other immune-related biomarkers (B7-H3, STAT3,Beta-2 microglobulin, CD44) . We surveyed the expression of the 31 biomarkers in two basal-like enriched tissue microarray cohorts representing a total of 59 breast cancer patients.    4.2 Material and methods 4.2.1 Study cohort  Tumor excision specimens from two British Columbia breast cancer patient cohorts (herein referred to as Cohort A and Cohort B) were built into two tissue microarrays comprised of 0.6mm duplicated cores.  Cohort A consisted of 39 patients diagnosed with invasive breast cancer in the period of 2008-2011, details of which have already been published in the context of a study from our laboratory of basal biomarker expression in ER negative and weakly ER positive cases255 . Breast cancer patients from cohort B (N=20) were diagnosed between 2013-2015, with clinical triple negative breast cancer (ER–/PR–/HER2–) and treated accordingly. None of the patients (from Cohort A or B) received neoadjuvant therapy, so the samples under 90  investigation represent primary tumor tissue that was not exposed to radiation, chemotherapy or endocrine therapy. Clinicopathological parameters for Cohort A and B are summarized in Table 4.1. Identification of breast cancer subtypes in Cohort A was done using PAM50 gene expression profiling as a gold standard breast cancer subtyping assay as previously published255. Clinical outcome data were available for cohort A with a median follow-up time of 48 months (cohort B cases are too recent to have mature outcome data).  The clinical research ethics board of the University of British Columbia and BC Cancer Agency approved access to clinical outcome and de-identified data. 4.2.2 Digital spatial profiling Breast cancer tissue microarrays were subjected to digital spatial profiling using Nanostring GeoMx™ technology as recently published256,257(See schematic in Figure 4.1).  In brief, tissue microarrays were incubated with a cocktail of antibodies that included fluorescent visualization markers and an immuno-oncology biomarker panel. To select regions of interest, fluorescent visualization markers consisted of CD45 for immune cells, Pan-cytokeratins for carcinoma cells and DAPI as a nuclear stain. Each TMA core served as a defined region of interest. The immuno-oncology biomarker panel consisted of 31 antibodies (see Figure 4.1) linked with UV-cleavable DNA oligos that can be quantified on the Nanostring nCounter platform. Biomarker expression counts were normalized using Nanostring internal spike-in controls for hybridization (called ERCC). Biomarker expression counts from duplicated cores were averaged for each patient prior to further analyses. There were 3 cases in Cohort A and 1 case in Cohort B for which biomarker expression counts were interpretable in only one of the 2 cores.  91  DSP counts were divided by the geometric mean values of isotype controls to generate the signal to noise ratio for each biomarker.    4.2.3 Immunohistochemistry and scoring  Anti-CD8 mouse monoclonal antibody (clone 28/144B) and anti-PD1 rabbit monoclonal antibody (clone EPR4877) were applied to the Cohort B TMA according to manufacturer’s protocol, at the Deeley Research Centre (Victoria, BC). A pathologist blinded to clinical and DSP data scored CD8+ and PD1+ lymphocytes using a previously published scoring system developed in our laboratory242. In brief, absolute counts of CD8+ or PD1+ lymphocytes were reported based on their localization in the tumor. Intra-epithelial tumor-infiltrating lymphocytes (iTILs) were defined as CD8+ or PD1+ lymphocytes in direct contact with carcinoma nests whereas stromal tumor-infiltrating lymphocytes (sTILs) were outside of carcinoma nests. To allow comparison with CD8 and PD1 DSP digital counts, iTIL and sTIL counts were combined to generate a total count of CD8+ or PD1+TILs for each tissue microarray core. H&E sTILs were scored on both TMA cohorts as described in Chapter 3 based on an internationally-standardized scoring system our lab contributed to developing74,258.  4.2.4 Statistical analysis Unsupervised hierarchical clustering was performed on DSP biomarker expression normalized counts for each patient using ComplexHeatmap on R/Bioconductor and provided by Nanostring. Associations with clinicopathological parameters and outcome were performed in IBM SPSS software version 25. Kaplan-Meier curve estimates were built for breast cancer-specific survival, defined as the time between invasive breast cancer diagnosis and time of death attributed to breast cancer. Breast cancer patients alive at last follow-up or that had died of other causes were censored.   92  An intraclass correlation was calculated to assess core to core agreement of biomarker expression digital counts between two duplicated cores. Chi-square tests compared the statistical significance of the pathological parameter associations among the DSP immune expression profiles. Biomarker correlation analyses (for CD8, PD1 IHC vs DSP counts and H&E sTILs vs CD45 DSP counts) were assessed by Spearman Rho. A one-way analysis of variance (ANOVA) test was conducted to identify biomarkers differentially expressed between the three levels of H&E sTILs (low:<10%,  intermediate: ≥10-<50% and high or lymphocyte-predominant breast cancer: ≥50%). To reduce Type I errors across multiple tests generated in ANOVA, a Bonferroni-correction was used to set a cut-off of 0.001 for significant p-values. Post-hoc multiple comparisons using Tukey Honest Significant Difference (HSD) test were used to assess significant differences in pairwise group comparisons. A multiple regression analysis in Cohort A (due to a larger sample size compared to Cohort B) was used to build a predictive model for identifying immune-enriched breast tumors based on the 4-biomarker signature. For this analysis, H&E sTILs counts were entered as the variable to predict. The predictive model generated, for each patient, a continuous immune score based on the 4-biomarker signature. High values of this score represented immune-enriched breast tumors.  4.3 Results 4.3.1 Overview of DSP digital counts of the 31 biomarkers A single tissue microarray slide for each cohort (represented by duplicated cores) was stained and analyzed by using the Nanostring Digital Spatial Profiler (Figure 4.1). Tissue microarray cores representing 37 out of 39 cases from Cohort A and all 20 cases from Cohort B were interpretable by DSP. DSP digital counts were obtained for all the biomarkers included in 93  the immuno-oncology panel (Appendix Table B.1). All counts were normalized using hybridization technical controls as described in Material and methods. For cases with interpretable digital spatial profiling counts on duplicated cores (N=34 in Cohort A and N=19 in Cohort B), there was a good to excellent core-to-core agreement (intraclass correlation greater than 0.75) in the 31 biomarker expression counts for the majority of cases (68% for cohort A and 58% for cohort B) with only a few cases (less than16% in both cohorts) having poor (intra class correlation <0.5) core-to-core agreement (Figure 4.2).  In both cohorts, DSP digital counts for tissue microarray cores were highly variable ranging from less than 100 counts for biomarkers including antibody isotype controls and FOXP3 to high values of more than 80,000 counts for biomarkers such as Pan-cytokeratin and β-catenin (Figure 4.3, bottom tables). DSP digital counts for mouse isotype control was significantly higher in Cohort B in comparison to Cohort A and a similar trend for the rabbit isotype control (p=0.07) was observed, suggesting a possible experimental variation (Figure 4.4A). However, there was no significant difference in digital counts for reference biomarkers (Histone H3 and Ribosome S6), indicating that cellularity was comparable between the tissue microarray cores (Figure 4.4B). For further DSP analyses, a signal to noise ratio normalization was applied to DSP counts using isotype controls counts to remove non-biological background variance (Appendix Table B.2).   4.3.2 Validation of CD8 and PD1 DSP counts by immunohistochemistry and correlations with tumor infiltrating lymphocyte counts To validate the biomarker digital counts generated by the digital spatial profiling, we analyzed immunohistochemistry scores assessed visually by a pathologist for CD8 and PD1 previously stained on the Cohort B tissue microarray. Based on our established, published cut-94  points96,242, CD8+ iTILs (≥1) were present in 55% of cases whereas PD1+iTILs (≥1) were observed in 50% of cases. CD8 and PD1 immunohistochemistry continuous scores significantly (p<0.001) correlated with the CD8 and PD1 DSP counts performed on sections from the same tissue microarray (Spearman rho r=0.674 for CD8 and r=0.838 for PD1: Figure 4.5A; B). We found that there were 8 cases (40%) with PD1+ lymphocytes that were only detected by DSP, although the PD1 digital counts for these specific cases were in the lower range of distribution of PD1 DSP counts (Figure 4.5B).  Similar findings were observed for differences between CD8 DSP counts and immunohistochemistry scores.  As H&E sTILs scores were available for both cohorts, we also analyzed the Spearman rank correlation between visual counting of lymphocytes with CD45 DSP counts (a general marker for lymphocytes) (Figure 4.5C-D). H&E sTILs scores and CD45 DSP counts significantly correlated (p<0.001) with each other, supporting that the level of immune cells detected by CD45 using DSP directly reflects the number of morphologically-characterized lymphocytes visually identifiable in a core.  4.3.3 Distinct breast cancer immune expression profiles are illustrated by DSP      Normalized counts of the 31 biomarker DSP immuno-oncology panel were subjected to an unsupervised hierarchical clustering analysis for each cohort (Figure 4.6). Cohort A (n=37) was clustered into 4 groups whereas Cohort B (n=20) partitioned into 3 groups (Figure 4.6).  The level of immune infiltration based on the immune biomarkers’ expression counts varied greatly between these groups. In Cohort A, breast cancer patients in Groups A2 and A3 showed the highest immune infiltration including high levels of CD45, CD3 and CD20 counts in comparison to patients represented in Group A1 and A4 (Figure 4.6A). However, patients in Group A2 had the most immune-enriched tumors among all the groups (Figure 4.6A). In Cohort 95  B, Group B2 included breast cancer patients with high levels of CD45, CD45RO, CD8, and CD20 counts whereas patients in Groups B1 and B3 had low to intermediate level of immune infiltration in their basal-like tumors (Figure 4.6B).   As expected, pan-cytokeratin counts were elevated  in patients with low immune infiltration in their tumors for both cohorts (i.e., in groups A1/A4 compared to groups A2/A3 for cohort A and in groups B1/B3 compared to group B2 for cohort B, Figure 4.6) although it did not reach statistical significance. Similar non-significant trends were observed for additional non-immune biomarkers elevated in immune-desert breast tumors including p-AKT and beta-catenin in Cohort B (Figure 4.6B). Moreover, counts for B7-H3, an emerging immune checkpoint molecule259, appeared to be inversely correlated with immune infiltration in Cohort A (Figure 4.6A).   No significant associations with any of the clinicopathological parameters tested were found, likely due to the limited sample size in each cohort (Tables 4.2 and 4.3). However, in survival analyses (on Cohort A), improved breast cancer-specific survival was associated with immune-enriched DSP profiles (Figure 4.7). All the patients with death attributed to breast cancer in this 7-year follow-up were among the DSP profile groups with low tumor-immune infiltration, namely A1 and A4. Due to the absence of events in the immune-enriched DSP profiles, hazard ratios could not be computed. 4.3.4 Identification of a set of biomarkers most associated with immune-enriched profiles in both cohorts To determine which were the key biomarkers included within the DSP immuno-oncology panel that would discriminate between immune-enriched and immune-desert breast tumors, I 96  conducted a one-way analysis of variance among patients stratified by the different levels of H&E sTILs shown in Table 4.2 and 4.3.     After performing log transformation of the DSP counts for all 31 biomarkers as per standard procedure to allow one-way analysis of variance tests statistics to be calculated, I found 8 biomarkers out of 31 with significant Bonferroni-corrected p-values showing a difference in expression between patients with different levels of H&E sTILs in Cohort A. These were: CD20, CD45RO, CD3, PD1, CD4, CD45, CD8 and CD19 (Table 4.4).  In Cohort B, only 4 out of 31 biomarkers were differentially expressed among patients with different levels of H&E sTILs and all 4 biomarkers (CD20, CD3, PD1 and CD45) were also shared with Cohort A (Table 4.4). Post-hoc pairwise comparison analyses using Tukey honest significant difference test indicated which H&E sTILs groups had significant differences in expression of the selected biomarkers (Table 4.4). The B-lymphocyte biomarker CD20 had the highest fold change in expression between patients with lymphocyte-predominant breast cancers and patients with low levels of H&E sTILs. There was a more than 100-fold difference for Cohort A and a nearly 50-fold difference in expression for Cohort B (Table 4.4). I settled on the 4 immune biomarkers (CD20, CD3, PD1 and CD45) differentially expressed in both cohorts as a potential candidate protein biomarker-signature for immunogenic breast cancers. The 4 biomarker-signature was entered in a multiple regression analysis to build a predictive model for immune-enriched breast tumors (see methods section for details-Table 4.5). The equation yielded in the predictive model showed a higher coefficient factor for PD1 DSP counts in comparison to the other 3 biomarkers (1.412 for PD1 vs 0.035 for CD3 counts), suggesting a significant contribution of PD1 counts to predict immune-enriched tumors. In 97  contrast, as CD45 expression is not restricted to lymphocytes, a negative coefficient factor was associated with CD45 DSP counts in the predictive model.  In both cohorts, this predictive model yielded a good to excellent area under the receiver operating characteristic curve (AUC=0.829 for Cohort A and AUC= 1.00 for Cohort B), indicating that this model accurately identified immune-enriched breast tumors (Table 4.5). In cohort A, the 4-biomarker signature (treated as a continuous variable) correlated with improved breast cancer-specific survival albeit not reaching statistical significance (HR: 0.65 , 95% CI:0.35-1.09, Wald test p= 0.10).  4.4 Discussion Using the digital spatial profiling technology, I was able to profile 31 protein biomarkers for their expression in 59 breast cancer patient surgical samples using just two TMA slides. Digital counts obtained by DSP on tissue microarray cores show a good core-to-core agreement between duplicated cores and exhibited a high dynamic range from biomarkers with low counts such as FOXP3 to highly abundant proteins such as cytokeratins. Digital counts from isotype controls showed variation between cohort A and cohort B whereas cellularity remained similar between the two cohorts. Furthermore, I directly validated CD8 and PD1 DSP digital counts by single stain immunohistochemistry and indirectly validated CD45 DSP digital counts by assessing H&E sTIL scores. The immune biomarker expression profiles illustrated by DSP were significantly associated with survival in one cohort. Our discovery-based study identified a signature of 4 key biomarkers (CD20, CD3, PD1 and CD45) that best discriminate immune-enriched from immune-cold breast tumors in the tested study cohorts. This represents the first 98  study to investigate immune expression profile in breast cancer tissue microarrays using the digital spatial profiling Nanostring-based technology.  In situ detection of multiple biomarkers in tumor tissues can also be done by fluorescent multiplex IHC or by mass cytometry. One of the most commonly used fluorescent multiplex IHC methods compatible with autostainers is the Opal system by Perkin Elmer. The Opal system offers the visualization of up to six target biomarkers in tumor tissues and is a preferred tool for proteins of low abundance as it has increased sensitivity due to tyramide-based signal amplification260. However, some of the disadvantages of this technique include: 1) the need for signal amplification which does not provide the relative abundance of the target in the tissue; 2) a requirement for particular planning of sequential staining of compatible antibodies in a panel, which is a laborious task and limits the number of targets that can be detected and 3) this technique requires building and training algorithms for multispectral image analysis253,261-264.  Multiplex detection by mass spectrometry such as mass cytometry or multiplex-ion beam imaging techniques can identify more biomarkers than conventional multiplex IHC and offers quantitative measurement but still requires training for digital image analysis and can be time-consuming 254,260. In contrast, the 31 biomarker expression DSP digital counts were generated automatically without signal amplification steps and without an image analysis training step. Thus, the Nanostring-based digital spatial profiling technology offers a way to quickly and directly profile the tumor immune microenvironment and select key biomarkers for in-depth analyses. Indeed, DSP was recently used in two different studies to profile immune biomarker expression in tumor biopsies from melanoma patients receiving immune checkpoint inhibitors (a combination of anti-CTLA4+ anti-PD1 agents or single anti-PD1 agent) and few immune 99  biomarkers were found to correlate with response in the neoadjuvant setting and with increased relapse-free survival in the adjuvant setting256,257.         In addition to identifying discrete breast cancer immune profiles, I was also able to associate the DSP-derived immune expression profiles with clinical outcome in an exploratory analysis. In Cohort A, the groups of patients with high tumor immune infiltration (A2 and A3) were associated with significantly improved breast cancer-specific survival.   Using one-way analysis of variance statistics, I narrowed the DSP immuno-oncology panel down to 8 biomarkers that discriminated the groups into high-vs low immune infiltrations. These 8 biomarkers include a general lymphocyte marker (CD45) and more specifically encompass immune cell populations including T cells (CD3, CD4, CD8), B cells (CD19, CD20), exhaustion-immune checkpoint states (PD1), and memory state (CD45RO). In Cohort B, the immuno-oncology panel was narrowed down to 4 immune biomarkers that associated with those cancer patients with the highest tumor immune infiltration (grouped in lymphocyte-predominant breast cancers group), all 4 of which were also incorporated within the 8-biomarker signature from Cohort A. These 4 common biomarkers are CD20, CD3, PD1, and CD45. Furthermore, I identified a predictive model to identify immune-enriched breast tumors based on the 4-biomarker signature.    As this 4-biomarker signature illustrates, evaluating one immune biomarker at a time leaves out important information that may have clinical utility (prognostic or predictive value). Indeed, studies on predictive biomarkers of response to immune checkpoint inhibitors show the need for assessing more than one biomarker215,249,265 [Reviewed by Gibney et al.266]. DSP digital counts from non-immune biomarkers such as beta-catenin and p-AKT appeared to positively correlate with breast tumors with low immune infiltration (although it did not reach statistical 100  significance in the two small cohorts). Beta-catenin has been reported to be associated with immune-desert tumors [ Reviewed by Spranger S et al.,267] which further shows the importance of assessing tumor intrinsic properties in multiplex panels as in the DSP.  Assessing H&E sTIL scores evaluates all the immune populations included in the 4-biomarker signature, namely B cells, and T cells in different activation states. In my study, H&E sTIL scores correlate significantly with CD45 DSP counts. However, using this method, the information of the individual contribution of each immune cell type to the overall immune population is not evaluated.  As this discovery-based study shows, biomarkers such as CD20, on an average, have the highest DSP counts in tumor immune-enriched breast cancer patients. This can be interpreted as B cells being the largest contributor to the total lymphocyte population within the tumor, since DSP digital counts are directly reflective of the number of CD20 molecules present. However, this should be taken with caution as the abundance of each antibody-targeted molecule per cell might be due to a difference in sensitivity of detection of certain epitopes (e.g. the CD20 epitope targeted in the panel could be particularly highly accessible in formalin-fixed paraffin-embedded tissues and/or could have an especially high affinity for its antibody).  Immuno-oncology biomarker studies have largely focused on T cells265,268, as PD-1 (expressed on T cells) represents the target for the regulator-approved immunotherapy drugs in greatest clinical use. In contrast, studies on the presence of B cells in breast cancer have been mostly prognostic86,88,89 with one study suggesting B cells are predictive of response to anthracyclines, an immune-modulating chemotherapy agent100. Certainly, B cells have pleiotropic functions within tumors, including antigen presentation, which can enable a de novo immune response within the tumor microenvironment269.  101  Although the study I describe in Chapter 4 successfully identified immune expression profiles from breast cancer patient specimens that have potential clinical relevance, I have to acknowledge several limitations: 1) As it is a first study applying this novel technology to breast cancer tissue microarray materials, the sample size was limited which can affect the variability of DSP profiles if analyzed in a larger cohort; 2) DSP profiles in each cohort were selected in a data-driven fashion based on the visual clusters generated by unsupervised hierarchical analysis and data-driven model-building and thus, conclusions drawn from the biomarkers and panels identified from this approach will require further validation on independent materials; and 3) due to the limited sample size, the whole tissue microarray core was selected as a region of interest without discriminating intraepithelial from stromal immune infiltration which prevented the ability to assess the importance of biomarker localization in the tissue.   As the number of clinically relevant immune biomarkers discovered is likely to continue rising, multiplex detection systems such as DSP will be needed to perform future studies. Moreover, as Chapter 4 illustrates, DSP can generate a substantial amount of protein measurements using only tissue microarray cores, therefore sparing precious patient tumor tissues. This first study on breast cancer tissue microarrays supports DSP as a technology that could be used to survey the immune context in tissue samples representing tumors responding versus non-responding to immune-modulating therapies. A proposal to use DSP to evaluate the capacity of immune biomarkers to predict response to immune-modulating chemotherapy in breast cancer is presented at the end of this chapter with a future goal of applying the technology on breast cancer immune checkpoint inhibitor clinical trials.    102  4.5 Proposal to apply the Nanostring digital spatial profiling technology to clinical trials material  Assessing the predictive value of immune biomarkers for response to anthracyclines, an immune-modulating chemotherapy, in the MA.5 clinical trial Tumor immunogenicity or the ability of cancer cells to produce antigens recognized by the body’s immune system is a major biological feature underlying the success of immunotherapy270,271. A growing body of evidence suggests that some chemotherapy drugs, such as anthracyclines, are immune-modulating as their effects include promotion of immune activation, inhibition of immunosuppressive cells, and/or release of antigens from tumors in a process termed immunogenic cell death122,272,273.  Pre-existing immunity in breast cancer patients treated with neoadjuvant anthracycline-containing chemotherapy is associated with pathologic complete response in hormone-receptor negative subtypes100. However, in the adjuvant setting, the predictive value of a pre-existing anti-tumor immune response is not clear. Some studies report solely a prognostic value for high levels of H&E stained tumor-infiltrating lymphocytes in hormone-receptor negative breast cancer patients treated with anthracycline-containing chemotherapy; others report a predictive value81,83. The Canadian Cancer Trials Group (CCTG) MA.5 clinical trial randomized premenopausal node-positive breast cancer patients to receive adjuvant Cyclophosphamide-Methotrexate-5’Fluorouracil (CMF) or Cyclophosphamide-Epirubicin-5’Fluorouracil (CEF: anthracycline substitution) and therefore provides an opportunity to evaluate the value of pre-existing immunity in breast cancer patients to predict benefit from adjuvant anthracyclines, an immune-modulating chemotherapy. 103  Problem: H&E stromal tumor-infiltrating lymphocytes represent a heterogeneous population of immune cells and lack precision as a biomarker, highlighting a need to test more advanced, specific biomarkers. However, individually testing the large number of biomarkers required to identify immune populations and activation phenotypes in formalin-fixed paraffin-embedded tissues would consume a considerable amount of precious Phase III trial material.  We propose to evaluate the predictive value of immune biomarkers of response to immune modulating chemotherapy in patients from the MA.5 clinical trial by undertaking digital spatial profiling (DSP) using NanoString’s new GeoMx technology, a method for which we have generated relevant preliminary data on breast cancer tissue microarrays.  4.5.1 Hypothesis:  Breast cancer patients with pre-existing immune-enriched tumors assessed by the 4-biomarker signature will benefit from adjuvant anthracycline-containing chemotherapy (CEF) more than from CMF. Our study objectives are:  1. Assess the predictive value of the 4-immune biomarker signature of response to anthracycline-containing chemotherapy regimens 2.  Characterize the pre-existing tumor immune microenvironment of a large set of breast cancer patients using a 56 antibody multiplexed immuno-oncology panel on the DSP Nanostring platform.  4.5.2 Materials and methods:  For this study, we will use the existing TMAs built from surgical excision tumor specimens from the patients in MA.5. Our laboratory has experience using the MA.5 TMAs which comprise 4 blocks representing a total of 511 cases with duplicated cores. One 4μm FFPE 104  unstained TMA section will be required for this study, which will be performed using a Nanostring Digital Spatial Profiler instrument being installed on site. The DSP immuno-oncology panel will be purchased from Nanostring and current iteration of this panel includes 56 biomarkers (Table 4.6).     Fluorescent visualization markers for selection of regions of interest will include a tumor marker (Pan-cytokeratins), a lymphocyte marker (CD45) and DAPI as a nuclear stain. To allow the assessment of intraepithelial vs stromal immune infiltration,  a pan-cytokeratin and CD45 mask will be used to select two regions of interest (ROIs) per TMA core per patient to be analyzed by DSP (a total of 2044 ROIs). Pan-cytokeratin positive/CD45 negative will be defined as intraepithelial regions whereas stromal regions will be identified as pan-cytokeratin negative/CD45 positive. Analysis of biomarker expression counts in each ROI will be conducted in our lab using the Nanostring DSP platform. Internal spike-in hybridization controls and a signal to noise ratio will be applied for data normalization prior to statistical analyses.  4.5.3  Statistical design:  MA.5 randomized 710 node-positive premenopausal breast cancer patients to receive adjuvant anthracycline-containing chemotherapy (CEF, n=351) versus CMF (n=359). The 10-year follow-up trial results showed superiority of CEF over CMF for the primary endpoint, relapse-free survival (HR=1.31; 95%CI, 1.06 to 1.61; stratified log-rank, p=0.007)274. The MA.5 TMA consists of 511 breast cancer patients representing 72% of the clinical trial’s population.  We hypothesize that there will be a significant interaction observed between MA.5 study arm (CEF vs CMF) and a pre-existing antitumor immune response for the trial’s primary endpoint, relapse-free survival.  105  Drawing from our recently-completed discovery-based study on breast cohort TMAs, we will evaluate the predictive value of the 4-biomarker signature (CD20, CD3, PD1 and CD45) in the intraepithelial region as our primary hypothesis.  Precisely, the level of immune infiltration in breast tumors will be based on the score generated by the 4-biomarker predictive model as presented in the initial cohort-based study. The Canadian cancer trials group statisticians following a formal pre-specified written statistical plan will conduct all clinical analyses. The initial analyses will include a) a table of patient characteristics (age, nodal status, tumor stage, intrinsic subtypes previously defined by PAM5014, according to 4-biomarker signature (as a continuous variable); b) a distribution of the 4-biomarker signature by treatment regimen (CEF vs CMF) to assess any improper balances between patients in CEF and in CMF treatment arms . If there are no significant imbalances, Kaplan-Meier estimates in univariate analyses and multivariate cox regression analyses will use relapse-free survival as a primary endpoint and overall survival as secondary endpoint stratified by the 4-biomarker signature. Our analyses will first focus on the CEF arm alone as CMF can potentially have an immune modulating effect that could lower our power to see an effect of pre-existing immunity in response to CEF. Benefit of CEF will be estimated for all patients stratified by the 4-biomarker signature. As a secondary analysis, we will analyze benefit of CEF vs CMF for all patients stratified by the 4-biomarker signature. A treatment x 4-biomarker immune signature interaction term will be calculated using the likelihood ratio test to assess benefit of CEF vs CMF in all patients stratified by the 4-biomarker signature. Prognostic and unsupervised hierarchical clustering analyses using the full DSP 56 immune biomarker panel will be conducted as exploratory analyses.   106  A finalized specific statistical plan will be discussed with the Canadian cancer trials group statisticians following initial gathering and processing of DSP results and reassessment of study power based on the biomarker distribution and number of cases with interpretable data.  4.5.4 Significance of research:  This study proposes to evaluate, quantitatively, the complex tumor immune microenvironment of breast cancer patient surgical specimens using the cores on existing tissue microarrays via novel digital spatial profiling technology. As a result, large amounts of protein-level biomarker data will be generated from tiny amounts of precious clinical trial material.  Furthermore, we will assess the predictive value of a pre-existing antitumor immune response signature for benefit from anthracycline-containing chemotherapy. This could support hypotheses that particular chemotherapy regimens act in part via the immune system. Biomarker results could potentially help to better identify those breast cancer patients most amenable to immune-modulating anthracycline chemotherapies vs. those who might be spared their sometimes severe side effects.          107     Figure 4.1 Schematic of DSP analysis process on Cohort A and B TMAs (1) An antibody cocktail is applied to formalin-fixed paraffin-embedded tumor tissue from a 4 micron section of a tissue microarray. Green-highlighted biomarkers in the antibody cocktail represent immune cell populations. (2) The slide is visualized using 3 fluorescent markers included in the antibody cocktail (Pan-CK in green, CD45 in red and a nuclear stain in blue) and is used to select the regions of interest (ROIs). 37/39 cases were interpretable for Cohort A. All 20 cases were interpretable for Cohort B. (3) DNA oligos attached to each antibody are cleaved by UV in the selected ROI and aspirated oligos are dispensed into a 96-well plate. The process repeats for each ROI.  (4) 96-well plate is put into the Nanostring nCounter for hybridization to corresponding capture/reporter probes and digital counts for each antibody-targeted biomarker are generated.  Cohort A Cohort BAKT, B7-H3, Bcl-2, B2M, B-catenin, CD14, CD19, CD3, CD4, CD44, CD45, CD45RO, CD56, CD68, CD8A, FOXP3,GZMB, Histone H3, IgG RbCtrl, IgG Ms Ctrl, Ki67, CD20, p-AKT, Pan-CK, PD-1, PD-L1, VISTA, PTEN, p-STAT3, STAT3, S6Hybridize and count using Nanostring nCounterAntibody cocktailSelection of ROIsAspiration of DNA oligos for each ROI96-well plate (1)(2)(3)(4)108   Figure 4.2 Core-to-core agreement on DSP measurements for biomarker expression counts  Graphs depicting the intraclass correlation coefficient computed for the 31 biomarkers expression counts measured in each pair of duplicated tissue microarray cores for each patient in Cohort A (top) and Cohort B (bottom) . Numbers shown in each brackets represent the tissue microarray core number of duplicates.  109   Figure 4.3 Distribution of DSP digital counts. Top, boxplots illustrating DSP digital normalized counts (minimum-maximum) obtained for each DSP target in each cohort.  Bottom, tables representing median counts and interquartile range for each DSP target in each cohort.   Cohort BCohort A110      Figure 4.4 Comparison of background variance and sample cellularity between Cohort A and Cohort B. A, boxplots depicting DSP normalized counts for each isotype antibody control (Mouse and Rabbit) in each cohort and used to evaluate background variance. B, boxplots illustrating DSP normalized counts for 2 reference protein biomarkers (Histone H3 and Ribosome S6) used to evaluate sample cellularity. A Mann-Whitney test was used to assess significant difference in DSP counts for isotype controls or reference proteins between Cohort A and Cohort B.  ns: non-significant p value       111    Figure 4.5 Validation of digital spatial profiling counts by immunohistochemistry and H&E staining. Correlation between CD8 (A), PD1 (B) digital spatial profiling counts and immunohistochemistry scores. (C, D) Correlation analysis between stromal tumor-infiltrating lymphocytes (sTILs) scored on H&E-stained slides and CD45 DSP counts in Cohort A and B. Spearman rho correlation coefficient and p values are displayed for each correlation analysisA BC D112    Cohort A Cohort B113  Figure 4.6 Unsupervised hierarchical clustering of patients in Cohort A and B based on biomarker expression counts analyzed by DSP. Top, heatmaps generated based on DSP technical control-normalized counts for each region of interest (or tissue microarray core) in each cohort. Bottom, DSP counts for regions of interest representing duplicated tissue microarray cores were averaged and a heatmap representing DSP counts per patient was generated for each cohort.  Bottom of each heat map shows the clinico-pathological parameter associated with each patient. Color bar in the middle of the 2 heatmaps denotes the biomarkers counts.  H&E sTILs: Hematoxylin and eosin-stained stromal tumor-infiltrating lymphocytes; LVI: Lymphovascular invasion; LN: Lymph node status   114        Figure 4.7 Prognostic value of DSP immune profiles in breast cancer patients from Cohort A. Kaplan Meier curves of breast cancer-specific survival (BCSS) among patients with interpretable DSP counts (n=37/39) stratified by DSP clusters. The median follow-up time was 43 months. Number of events for each curve and the log rank p value of the Kaplan Meier are displayed.    115   Table 4.1 Cohort description Cohort A  Cohort B Parameters N  Parameters N Age (yrs) <50 ≥50  6 33 (85%)  Age (yrs) <40 ≥40-47  4 16 (80%) Tumor size (cm) ≤2 >2  15 33 (61%)  Tumor size (cm) ≤2 >2  10 8 (44%) Tumor Grade Grade 1 or 2 Grade 3  12 27 (69%)  Tumor Grade Grade 1 or 2 Grade 3  0 19 (100 %) PAM50 subtypes Luminal A Luminal B HER2E Basal-like   6 12 5 15 (40%)  Lymph node status Negative Positive  6 10 (63%)  Lymphovascular invasion  Absent Present   13 5 (28%)    116  Table 4.2 Cohort A DSP immune expression profile associations with clinicopathological parameters*  Cluster: A1 (n=6) A2 (n=6) A3 (n=11) A4 (n=14) p** Age (yrs) <50 ≥50  2 (33%) 4 (67%)  0 6 (100%)  0 11 (100%)  4 (29%) 10 (71%) 0.10 Tumor size (cm) ≤2 >2  1 (17%) 5 (83%)  4 (67%) 2 (33%)  4 (40%) 6 (60%)  5 (36%) 9 (64%) 0.35 Tumor Grade Grade 1 or 2 Grade 3  2 (33%) 4 (67%)  2(33%) 4(67%)  4 (36%) 7 (64%)  3 (21%) 11 (79%) 0.86 PAM50 subtypes Luminal A Luminal B HER2E Basal-like  0 3 (60%) 2 (40%) 0  1 (17%) 1 (17%) 1 (17%) 3 (50%)  2 (18%) 3 (27%) 0 6 (55%)  2 (14%) 4(29%) 2(14%) 6 (43%) 0.40 H&E sTILs level Low (<10%) Intermediate (≥10-<50%) High or LPBC (≥50%)   3 (50%) 3 (50%) 0  0 3 (50%) 3 (50%)  4(36%) 7 (64%) 0  10 (71%) 4 (29%) 0 0.001 LPBC: Lymphocyte-Predominant Breast Cancer * Lymphovascular invasion and nodal status information were not available for the analysis.**Chi-square test            117  Table 4.3 Cohort B DSP immune expression profile associations with clinicopathological parameters Cluster: B1 (n=5) B2 (n=8) B3 (n=7) p* Age <50  ≥50  5 0  8 0  7 0 n/a Tumor grade Grade 1 or 2 Grade 3  5 0  8 0  7 0 n/a Tumor size (cm) ≤2 >2  3 (60%) 2 (40%)  4 (67%) 2 (33%)  3 (43%) 4 (57%) 0.67 Lymphovascular invasion absent present  2 (50%) 2 (50%)  6 (75%) 2 (25%)  5 (83%) 1 (17%) 0.5 Lymph node status negative positive  0 3(100%)  4 (57%) 3 (43%)  2 (33%) 4 (67%) 0.22 H&E sTILs level Low (<10%) Intermediate (≥10-<50%) High or LPBC (≥50%)   4 (80%) 1(20%) 0  0 2 (25%) 6 (75%)  2 (29%) 3 (43%) 2 (28%) 0.02 n/a: Not applicable LPBC: Lymphocyte-Predominant Breast Cancer  *Chi-square test           118  Table 4.4 Identification of key biomarkers in breast tumors stratified by H&E sTILs levels from Cohort A and B. Cohort A Biomarker F  value p  Post-hoc multiple comparison among H&E sTILs levels LOG2 mean difference (95%CI) p  1. CD20 22.186 <0.001 LPBC intermediate 4.78 (2.04-7.52) <0.001 Low 7.07 (4.33-9.81) <0.001 Intermediate Low 2.29 (0.79-3.79) 0.002 2. CD45RO 21.220 <0.001 LPBC intermediate 2.89 (1.41-4.36) 0.001 Low 3.85 (2.38-5.33) <0.001 Intermediate Low 0.96 (0.16-1.77) 0.02 3. CD3 20.144 <0.001 LPBC intermediate 3.46 (1.58-5.34) <0.001 Low 4.74 (2.86-6.62) <0.001 Intermediate Low 1.28 (0.25-2.31) 0.01 4. PD1 18.824 <0.001 LPBC intermediate 1.97 (0.81-3.14) 0.001 Low 2.81 (1.64-3.98) <0.001 Intermediate Low 0.84 (0.20-1.48) 0.008 5. CD4 15.746 <0.001 LPBC intermediate 2.29 (1.091-3.50) <0.001 Low 2.76 (1.55-3.96) <0.001 Intermediate Low 0.46 (-0.198-1.12) 0.22 6. CD45 15.679 <0.001 LPBC intermediate 3.46 (1.36-5.56) 0.001  Low 4.69 (2.60-6.79) <0.001 Intermediate Low 1.23 (0.08-2.38) 0.03 7. CD8A 14.466 <0.001 LPBC intermediate 2.90 (1.09-4.71) 0.001 Low 3.90 (2.09-5.71) <0.001 Intermediate Low 0.99 (0.001-1.98) 0.05 8. CD19 13.091 <0.001 LPBC intermediate 2.98 (1.38-4.59) <0.001 Low 3.34 (1.74-4.95) <0.001 Intermediate Low 0.36 (-0.52-1.24) 0.58 Cohort B Biomarker F value p  Post-hoc multiple comparison among H&E sTILs levels LOG2 mean difference (95%CI) p  1. PD1 16.757 <0.001 LPBC intermediate 0.60 (-0.46-1.65) 0.34 Low 2.34 (1.28-3.39) <0.001 Intermediate Low 1.74 (0.61-2.87) 0.003 2. CD20 14.777 <0.001 LPBC intermediate 2.38 (-0.3-5.02) 0.08 Low 5.60 (2.96-8.24) <0.001 Intermediate Low 3.22 (0.40-6.05) 0.02 3. CD3 10.913 0.001 LPBC intermediate 0.07 (-1.62-1.77) 0.99 Low 2.82 (1.13-4.52) 0.001 Intermediate Low 2.75 (0.93-4.56) 0.003 4. CD45 1.455 0.001 LPBC intermediate 0.51 (-1.53-2.56) 0.80 Low 3.48 (1.43-5.52) 0.001 Intermediate Low 2.97 (0.78-5.15) 0.008 LPBC: Lymphocyte-Predominant Breast Cancer 119  Table 4.5 Assessment of the 4-biomarker signature for detecting immune-enriched breast tumors        Predictive model: 4.051+ (0.035*[CD3 DSP counts]) + (1.412*[PD1 DSP counts])-(0.079*[CD45 DSP counts])+(0.055*[CD20 DSP counts]) AUC: Area under the receiver operating characteristics curve N/A: No outcome data was available to compute hazard ratios in Cohort B    AUC (95% CI) Hazard ratio for breast cancer-specific survival (95%CI), p value 4-biomarker signature in Cohort A 0.828 (0.697-0.962) 0.65 (0.39-1.09), p=0.10 4-biomarker signature in Cohort B 1.00 N/A 120   Table 4.6 List of biomarkers in the commercially available DSP immuno-oncology panel that will be used for the study on MA.5 clinical trial     Biomarkermodule Biomarker listImmune cell profiling core (20 markers)Beta-2-microglobulin, CD11c, CD20, CD3, CD4, CD45, CD56, CD68, CD8, CTLA4, GZMB, Histone H3, Ki-67, PD1, PD-L1, Pan-cytokeratin, HLA-DR, SMA, Fibronectin, TGF-BImmuno-oncologydrug target module (11 markers)4-1BB, B7-H4, LAG-3, OX40L, TIM-3, VISTA, ARG1, B7-H3, GITR, IDO-1, STING Immune activationstatus module (10 markers)CD127, CD25, CD80, CD86, ICOS, PD-L2, CD40, CD40L, CD27, CD44Immune cell typing module(8 markers)CD45RO, FOXP3, CD34, CD66b, gamma delta TCR, CD14, FAPalpha, CD163Pan-tumor module (7 markers)MART1, NY-ESO-1, S100B, Bcl-2, EpCAM, Her2/ERBB2, PTEN121  Chapter 5: Overall conclusions and future directions 5.1 Summary of findings Although breast cancer has not been generally viewed as an immunogenic cancer, the body of work presented in this thesis and accumulating evidence from others illustrates that this view is only applicable to a portion of breast cancers. Out of 3,992 breast cancer pathology specimens profiled in this thesis, hormone receptor negative breast cancers (specifically HER2+ and basal-like breast cancer subtypes) were frequently enriched in immune infiltrates.  The immune populations evaluated in Chapter 2 (LAG-3) and Chapter 3 (TIM-3) express immunotherapy targets for agents currently in early phase clinical trials, guided by the remarkable results reported from agents targeting the PD-1/PD-L1 pathway in multiple cancer types. Thus, this work provides important pre-clinical data on the selection of breast cancer patients amenable to testing such new immunotherapy agents.  Although Chapter 2 and Chapter 3 are quite similar, there were some differences in statistical analyses presented in each chapter such as conducting an interaction test between LAG-3 and breast cancer intrinsic subtypes for predicting survival in Chapter 2 and prognostic analyses of TIM-3 (Chapter 3) using overall survival as an endpoint. The different statistical analyses were provided to address reviewers’ comments for each publication.  Predictive biomarkers for cancer immunotherapy are still being investigated and results from immune checkpoint clinical trials point to multiple candidates rather than a single biomarker. In Chapter 4, I profiled the immune expression of 57 breast cancers in a discovery-based study using breast cancer tissue microarrays to test a new multiplex technology called digital spatial profiling by Nanostring. This technology simultaneously and quantitatively assessed the expression of 31 biomarkers, representing predominantly an immuno-oncology 122  panel.  I identified a signature of 4 biomarkers highly expressed in immune-enriched breast cancers. In addition, I present, at the end of Chapter 4, a proposal to apply the digital spatial profiling technology onto breast cancer clinical trial specimens to validate the prognostic aspect of this signature and to assess the predictive value of the 4-biomarker signature for response to anthracyclines, a type of chemotherapy drug with immune-modulating effects. Results from the proposed study will serve as a framework for evaluating the predictive value of immune biomarkers for response to immune checkpoint inhibitors in breast cancer immunotherapy clinical trials.   5.1.1 Perspectives for LAG-3-targeted agents  In Chapter 2, I investigated the presence of a newly recognized targetable immune checkpoint called LAG-3 in a large cohort of breast cancer patients. LAG-3 expression on tumor-infiltrating lymphocytes was enriched in basal-like breast cancers, highly associated with breast cancers positive for PD-L1 expression and was an independent favorable prognostic factor. This work constitutes the largest evaluation of the expression of LAG-3, PD-1, and PD-L1 at the protein level in breast cancer excision specimens.     A recent single-cell sequencing study of breast cancer excision specimens identified LAG3 gene expression as one of the characteristic markers for tissue resident memory T cells, cells that are thought to be the key players in mediating the effects of immune checkpoint inhibitors275.  LAG-3 can also be secreted in a soluble form with activating or inhibitory function on various immune populations but its cell surface expression on effector T cells leads to exhaustion134,276 . This apparent contradiction may be partly due to the variety of LAG-3 ligands276.  There is an ongoing Phase II clinical trial evaluating the efficacy of a LAG-3 agonist (IMP321) in hormone receptor-positive breast cancer patients which reported (as interim-results) 123  that no dose limiting toxicities in the 15 patients enrolled were observed277 (NCT02614833). At the time of this writing, there were few anti-LAG-3 antagonistic agents being evaluated in clinical trials. An early report from a phase I/II clinical trial evaluated an anti-LAG3 agent (LAG525) alone or in combination with an anti-PD1 agent in advanced malignancies, and found that the combination was well tolerated with promising signs of response in triple negative breast cancer patients278 .  The enthusiasm for finding new immune checkpoint targets for cancer immunotherapy and early results from LAG3 clinical trials point to LAG3 as a promising new candidate in breast cancer immunotherapy.  Furthermore, results presented in Chapter 2 can help guide combination strategies with other immune checkpoint targets. 5.1.2 Perspectives for TIM-3-targeted agents  In Chapter 3, I evaluated the expression of TIM-3, an emerging immunotherapy target, on tumor-infiltrating lymphocytes in our large cohort of 3,992 breast cancer patients. I found that in contrast to LAG-3, breast cancers infiltrated with TIM-3+ tumor-infiltrating lymphocytes were not particularly associated with the presence of a specific immune checkpoint biomarker but rather were strongly associated with all of them (PD-1, LAG-3, PD-L1). Moreover, TIM-3 remained an independent favorable prognostic factor in a multivariate analysis that included H&E sTILs as a covariate.   Some properties that contribute to make TIM-3 unique include its expression on several immune populations (resident memory T cells, macrophages and dendritic cells) and having multiple ligands (Galectin-9, CEACAM-1, Phosphatidyl serine)234,275.  The significance of TIM-3 as an immune checkpoint was again put in evidence in a recent study that identified rare germline mutations in HAVCR2 (TIM-3 coding gene) associated with TIM-3 deficiency that lead 124  to hematological malignancies in humans279. Results from early phase clinical trials of various anti-TIM-3 agents are yet to be published. 5.1.3 Digital Spatial profiling  Chapter 4 is quite distinct from the other chapters as it describes the application of a new technology, called digital spatial profiling, that I tested on two distinct, smaller tissue microarray cohorts. As the immuno-oncology field continues to expand and a variety of clinically relevant immune biomarkers are discovered, it is becoming important to have the tools to identify and ideally quantify multiple markers within formalin-fixed, paraffin-embedded patient tumor tissues removed at biopsy or surgery.  Immunohistochemistry is the workhorse in research or clinical pathology and although it is a simple methodology as an in situ detection system, conventional multiplex immunohistochemistry is restricted to antibody species that can work together sequentially as primary and secondary antibodies260. In contrast, the latest fluorescent multiplex immunohistochemistry systems (such as the Opal system) circumvent the restriction of antibody species by employing sequential antibody-directed attachment of fluorophores to the target epitopes253. However, this necessitates repeated antibody stripping and sequential hybridization steps that are unwieldy to run at scale; the most elaborate Opal detection system can measure up to six target biomarkers and thus would serve best as a validation tool for short lists of key biomarkers discovered by more highly multiplexed methods such as Nanostring-based digital spatial profiling. To measure multiple biomarkers quantitatively while preserving their spatial distribution in tumor tissue, conventional or fluorescent multiplex immunohistochemistry offers more highly resolved spatial distribution, but not such precise quantitative measurement. 125  Mass spectrometry imaging such as mass cytometry is another in situ multiplex detection system that offers quantitative measurement. It uses antibodies labeled with elemental isotopes thereby bypassing the requirement for different species’ primary antibodies and fluorophore emission spectra when assaying multiple biomarkers and has recently been used to evaluate immune infiltrates in breast cancer tissue microarrays254 but the technology requires the building of algorithms for multispectral imaging.  The digital spatial profiling technology allowed me to generate immune profiles of 57 breast cancer patients by analyzing the expression of 31 biomarkers using tissue microarray cores. Although there were only 4 out of 31 biomarkers that significantly distinguished immune-enriched from immune-desert breast tumors in the initial data sets, there were interesting trends observed with additional biomarkers such as beta-catenin and B7-H3 which show the value of assessing multiple biomarkers at once.  As outlined in the proposal study at the end of Chapter 4, the clinical value of the 4-biomarker signature is intended, in future studies, to be evaluated using clinical trials specimens from breast cancer patients that received immune-modulating therapies including immune checkpoint inhibitors. 5.2 Significance of the research and perspectives for the future of breast cancer immunotherapy Immunotherapy for breast cancer is just at its beginning and results from therapeutic agents being evaluated in ongoing clinical trials will provide more insights into factors enabling or preventing breast tumor responses to immunotherapies. This body of work brings a significant contribution into guiding the selection of breast cancer patients for clinical evaluation of emerging targets in immunotherapy. Furthermore, Chapter 4 provides a framework and tools for 126  efficiently and quantitatively profiling immune expression in breast cancers by simultaneously screening multiple biomarkers using only a tiny fraction of tissue in precious tumor samples. As I demonstrate in this thesis, current strategies for immune restoration appear most applicable to a fraction of immunogenic breast tumors. Most breast cancers have low immunogenicity (i.e., hormone receptor positive breast cancers) and will likely require different or a variation of current therapeutic strategies such as immune-modulating combination treatments. An example of that approach includes cyclin-dependent kinase 4/6 inhibitors that have shown impressive results in breast cancer clinical trials and have now been approved by health authorities280. A recent study showed that in addition to inhibiting breast cancer growth, abemaciclib, a cyclin-dependent kinase 4/6 inhibitor, provoked anti-tumor innate and adaptive immune responses in mice and thus provides pre-clinical evidence that supports combining CDK4/6 inhibition with anti-PD-L1 agents281. This strategy is being evaluated in early phase I/II clinical trials (NCT01676753; NCT02778685; NCT02779751; NCT03294694; NCT03573648) with promising preliminary results already reported in conference proceedings280. Additional immune-modulating combination treatments strategies can include radiotherapy and DNA-demethylating agents282.     In conclusion, the body of work and studies presented in this thesis contributes to the rapidly-expanding field anti-tumor immunity in breast cancer and supports the clinical evaluation of immune checkpoint combination immunotherapy that includes LAG-3 and TIM-3-targeted agents. The variety of therapeutic strategies and targets in cancer immunotherapy will continue to broaden; all with the hope of achieving personalized immunotherapy options for breast cancer patients, likely facilitated by technologies such as DSP.   127  Bibliography  1. Perou, C.M., Sorlie, T., Eisen, M.B., van de Rijn, M., Jeffrey, S.S., Rees, C.A., Pollack, J.R., Ross, D.T., Johnsen, H., Akslen, L.A., Fluge, O., Pergamenschikov, A., Williams, C., Zhu, S.X., Lonning, P.E., Borresen-Dale, A.L., Brown, P.O. & Botstein, D. Molecular portraits of human breast tumours. Nature 406, 747-752 (2000). 2. Curtis, C., Shah, S.P., Chin, S.F., Turashvili, G., Rueda, O.M., Dunning, M.J., Speed, D., Lynch, A.G., Samarajiwa, S., Yuan, Y., Graf, S., Ha, G., Haffari, G., Bashashati, A., Russell, R., McKinney, S., Group, M., Langerod, A., Green, A., Provenzano, E., Wishart, G., Pinder, S., Watson, P., Markowetz, F., Murphy, L., Ellis, I., Purushotham, A., Borresen-Dale, A.L., Brenton, J.D., Tavare, S., Caldas, C. & Aparicio, S. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 486, 346-352 (2012). 3. Goldhirsch, A., Winer, E.P., Coates, A.S., Gelber, R.D., Piccart-Gebhart, M., Thurlimann, B. & Senn, H.J. Personalizing the treatment of women with early breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol 24, 2206-2223 (2013). 4. Comprehensive molecular portraits of human breast tumours. Nature 490, 61-70 (2012). 5. Enerly, E., Steinfeld, I., Kleivi, K., Leivonen, S.K., Aure, M.R., Russnes, H.G., Ronneberg, J.A., Johnsen, H., Navon, R., Rodland, E., Makela, R., Naume, B., Perala, M., Kallioniemi, O., Kristensen, V.N., Yakhini, Z. & Borresen-Dale, A.L. miRNA-mRNA integrated analysis reveals roles for miRNAs in primary breast tumors. PLoS One 6, e16915 (2011). 6. Mertins, P., Mani, D.R., Ruggles, K.V., Gillette, M.A., Clauser, K.R., Wang, P., Wang, X., Qiao, J.W., Cao, S., Petralia, F., Kawaler, E., Mundt, F., Krug, K., Tu, Z., Lei, J.T., Gatza, M.L., Wilkerson, M., Perou, C.M., Yellapantula, V., Huang, K.L., Lin, C., McLellan, M.D., Yan, P., Davies, S.R., Townsend, R.R., Skates, S.J., Wang, J., Zhang, B., Kinsinger, C.R., Mesri, M., Rodriguez, H., Ding, L., Paulovich, A.G., Fenyo, D., Ellis, M.J. & Carr, S.A. Proteogenomics connects somatic mutations to signalling in breast cancer. Nature 534, 55-62 (2016). 7. Sorlie, T., Perou, C.M., Tibshirani, R., Aas, T., Geisler, S., Johnsen, H., Hastie, T., Eisen, M.B., van de Rijn, M., Jeffrey, S.S., Thorsen, T., Quist, H., Matese, J.C., Brown, P.O., Botstein, D., Lonning, P.E. & Borresen-Dale, A.L. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 98, 10869-10874 (2001). 8. Sorlie, T., Tibshirani, R., Parker, J., Hastie, T., Marron, J.S., Nobel, A., Deng, S., Johnsen, H., Pesich, R., Geisler, S., Demeter, J., Perou, C.M., Lonning, P.E., Brown, P.O., Borresen-Dale, A.L. & Botstein, D. Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci U S A 100, 8418-8423 (2003). 9. Hu, Z., Fan, C., Oh, D.S., Marron, J.S., He, X., Qaqish, B.F., Livasy, C., Carey, L.A., Reynolds, E., Dressler, L., Nobel, A., Parker, J., Ewend, M.G., Sawyer, L.R., Wu, J., Liu, Y., Nanda, R., Tretiakova, M., Ruiz Orrico, A., Dreher, D., Palazzo, J.P., Perreard, L., 128  Nelson, E., Mone, M., Hansen, H., Mullins, M., Quackenbush, J.F., Ellis, M.J., Olopade, O.I., Bernard, P.S. & Perou, C.M. The molecular portraits of breast tumors are conserved across microarray platforms. BMC genomics 7, 96 (2006). 10. Prat, A., Pineda, E., Adamo, B., Galvan, P., Fernandez, A., Gaba, L., Diez, M., Viladot, M., Arance, A. & Munoz, M. Clinical implications of the intrinsic molecular subtypes of breast cancer. Breast (Edinburgh, Scotland) 24 Suppl 2, S26-35 (2015). 11. Prat, A. & Perou, C.M. Deconstructing the molecular portraits of breast cancer. Molecular oncology 5, 5-23 (2011). 12. Parker, J.S., Mullins, M., Cheang, M.C., Leung, S., Voduc, D., Vickery, T., Davies, S., Fauron, C., He, X., Hu, Z., Quackenbush, J.F., Stijleman, I.J., Palazzo, J., Marron, J.S., Nobel, A.B., Mardis, E., Nielsen, T.O., Ellis, M.J., Perou, C.M. & Bernard, P.S. Supervised risk predictor of breast cancer based on intrinsic subtypes. J Clin Oncol 27, 1160-1167 (2009). 13. Rouzier, R., Perou, C.M., Symmans, W.F., Ibrahim, N., Cristofanilli, M., Anderson, K., Hess, K.R., Stec, J., Ayers, M., Wagner, P., Morandi, P., Fan, C., Rabiul, I., Ross, J.S., Hortobagyi, G.N. & Pusztai, L. Breast cancer molecular subtypes respond differently to preoperative chemotherapy. Clin Cancer Res 11, 5678-5685 (2005). 14. Cheang, M.C., Voduc, K.D., Tu, D., Jiang, S., Leung, S., Chia, S.K., Shepherd, L.E., Levine, M.N., Pritchard, K.I., Davies, S., Stijleman, I.J., Davis, C., Ebbert, M.T., Parker, J.S., Ellis, M.J., Bernard, P.S., Perou, C.M. & Nielsen, T.O. Responsiveness of intrinsic subtypes to adjuvant anthracycline substitution in the NCIC.CTG MA.5 randomized trial. Clin Cancer Res 18, 2402-2412 (2012). 15. Aure, M.R., Vitelli, V., Jernstrom, S., Kumar, S., Krohn, M., Due, E.U., Haukaas, T.H., Leivonen, S.K., Vollan, H.K., Luders, T., Rodland, E., Vaske, C.J., Zhao, W., Moller, E.K., Nord, S., Giskeodegard, G.F., Bathen, T.F., Caldas, C., Tramm, T., Alsner, J., Overgaard, J., Geisler, J., Bukholm, I.R., Naume, B., Schlichting, E., Sauer, T., Mills, G.B., Karesen, R., Maelandsmo, G.M., Lingjaerde, O.C., Frigessi, A., Kristensen, V.N., Borresen-Dale, A.L. & Sahlberg, K.K. Integrative clustering reveals a novel split in the luminal A subtype of breast cancer with impact on outcome. Breast Cancer Res 19, 44 (2017). 16. Mukherjee, A., Russell, R., Chin, S.F., Liu, B., Rueda, O.M., Ali, H.R., Turashvili, G., Mahler-Araujo, B., Ellis, I.O., Aparicio, S., Caldas, C. & Provenzano, E. Associations between genomic stratification of breast cancer and centrally reviewed tumour pathology in the METABRIC cohort. NPJ breast cancer 4, 5 (2018). 17. Hammond, M.E., Hayes, D.F., Dowsett, M., Allred, D.C., Hagerty, K.L., Badve, S., Fitzgibbons, P.L., Francis, G., Goldstein, N.S., Hayes, M., Hicks, D.G., Lester, S., Love, R., Mangu, P.B., McShane, L., Miller, K., Osborne, C.K., Paik, S., Perlmutter, J., Rhodes, A., Sasano, H., Schwartz, J.N., Sweep, F.C., Taube, S., Torlakovic, E.E., Valenstein, P., Viale, G., Visscher, D., Wheeler, T., Williams, R.B., Wittliff, J.L. & Wolff, A.C. American Society of Clinical Oncology/College of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer (unabridged version). Archives of pathology & laboratory medicine 134, e48-72 (2010). 18. Wolff, A.C., Hammond, M.E., Hicks, D.G., Dowsett, M., McShane, L.M., Allison, K.H., Allred, D.C., Bartlett, J.M., Bilous, M., Fitzgibbons, P., Hanna, W., Jenkins, R.B., 129  Mangu, P.B., Paik, S., Perez, E.A., Press, M.F., Spears, P.A., Vance, G.H., Viale, G. & Hayes, D.F. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update. J Clin Oncol 31, 3997-4013 (2013). 19. Goldhirsch, A., Wood, W.C., Coates, A.S., Gelber, R.D., Thurlimann, B. & Senn, H.J. Strategies for subtypes--dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011. Ann Oncol 22, 1736-1747 (2011). 20. Nielsen, T.O., Hsu, F.D., Jensen, K., Cheang, M., Karaca, G., Hu, Z., Hernandez-Boussard, T., Livasy, C., Cowan, D., Dressler, L., Akslen, L.A., Ragaz, J., Gown, A.M., Gilks, C.B., van de Rijn, M. & Perou, C.M. Immunohistochemical and clinical characterization of the basal-like subtype of invasive breast carcinoma. Clin Cancer Res 10, 5367-5374 (2004). 21. Perou, C.M. Molecular stratification of triple-negative breast cancers. The oncologist 15 Suppl 5, 39-48 (2010). 22. Lehmann, B.D., Jovanovic, B., Chen, X., Estrada, M.V., Johnson, K.N., Shyr, Y., Moses, H.L., Sanders, M.E. & Pietenpol, J.A. Refinement of Triple-Negative Breast Cancer Molecular Subtypes: Implications for Neoadjuvant Chemotherapy Selection. PLoS One 11, e0157368 (2016). 23. Burstein, M.D., Tsimelzon, A., Poage, G.M., Covington, K.R., Contreras, A., Fuqua, S.A., Savage, M.I., Osborne, C.K., Hilsenbeck, S.G., Chang, J.C., Mills, G.B., Lau, C.C. & Brown, P.H. Comprehensive genomic analysis identifies novel subtypes and targets of triple-negative breast cancer. Clin Cancer Res 21, 1688-1698 (2015). 24. Garrido-Castro, A.C., Lin, N.U. & Polyak, K. Insights into Molecular Classifications of Triple-Negative Breast Cancer: Improving Patient Selection for Treatment. Cancer discovery (2019). 25. Anampa, J., Makower, D. & Sparano, J.A. Progress in adjuvant chemotherapy for breast cancer: an overview. BMC Med 13, 195 (2015). 26. Waks, A.G. & Winer, E.P. Breast Cancer Treatment: A Review. Jama 321, 288-300 (2019). 27. Nielsen, T.O., Jensen, M.B., Burugu, S., Gao, D., Jorgensen, C.L., Balslev, E. & Ejlertsen, B. High-Risk Premenopausal Luminal A Breast Cancer Patients Derive no Benefit from Adjuvant Cyclophosphamide-based Chemotherapy: Results from the DBCG77B Clinical Trial. Clin Cancer Res 23, 946-953 (2017). 28. Munzone, E., Curigliano, G., Burstein, H.J., Winer, E.P. & Goldhirsch, A. CMF revisited in the 21st century. Ann Oncol 23, 305-311 (2012). 29. Pritchard, K.I., Messersmith, H., Elavathil, L., Trudeau, M., O'Malley, F. & Dhesy-Thind, B. HER-2 and topoisomerase II as predictors of response to chemotherapy. J Clin Oncol 26, 736-744 (2008). 30. Nielsen, K.V., Muller, S., Moller, S., Schonau, A., Balslev, E., Knoop, A.S. & Ejlertsen, B. Aberrations of ERBB2 and TOP2A genes in breast cancer. Molecular oncology 4, 161-168 (2010). 31. Di Leo, A., Desmedt, C., Bartlett, J.M., Piette, F., Ejlertsen, B., Pritchard, K.I., Larsimont, D., Poole, C., Isola, J., Earl, H., Mouridsen, H., O'Malley, F.P., Cardoso, F., Tanner, M., Munro, A., Twelves, C.J., Sotiriou, C., Shepherd, L., Cameron, D., Piccart, 130  M.J., Buyse, M. & Group, H.T.A.M.-a.S. HER2 and TOP2A as predictive markers for anthracycline-containing chemotherapy regimens as adjuvant treatment of breast cancer: a meta-analysis of individual patient data. The Lancet. Oncology 12, 1134-1142 (2011). 32. Asleh, K., Lyck Carstensen, S., Tykjaer Jorgensen, C.L., Burugu, S., Gao, D., Won, J.R., Jensen, M.B., Balslev, E., Laenkholm, A.V., Nielsen, D.L., Ejlertsen, B. & Nielsen, T.O. Basal biomarkers nestin and INPP4B predict gemcitabine benefit in metastatic breast cancer: Samples from the phase III SBG0102 clinical trial. International journal of cancer (2018). 33. Fujii, T., Le Du, F., Xiao, L., Kogawa, T., Barcenas, C.H., Alvarez, R.H., Valero, V., Shen, Y. & Ueno, N.T. Effectiveness of an Adjuvant Chemotherapy Regimen for Early-Stage Breast Cancer: A Systematic Review and Network Meta-analysis. JAMA Oncol 1, 1311-1318 (2015). 34. Nasrazadani, A., Thomas, R.A., Oesterreich, S. & Lee, A.V. Precision Medicine in Hormone Receptor-Positive Breast Cancer. Front Oncol 8, 144 (2018). 35. Costa, R.B., Kurra, G., Greenberg, L. & Geyer, C.E. Efficacy and cardiac safety of adjuvant trastuzumab-based chemotherapy regimens for HER2-positive early breast cancer. Ann Oncol 21, 2153-2160 (2010). 36. Lehmann, B.D., Bauer, J.A., Chen, X., Sanders, M.E., Chakravarthy, A.B., Shyr, Y. & Pietenpol, J.A. Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest 121, 2750-2767 (2011). 37. Sharma, P., Klemp, J.R., Kimler, B.F., Mahnken, J.D., Geier, L.J., Khan, Q.J., Elia, M., Connor, C.S., McGinness, M.K., Mammen, J.M., Wagner, J.L., Ward, C., Ranallo, L., Knight, C.J., Stecklein, S.R., Jensen, R.A., Fabian, C.J. & Godwin, A.K. Germline BRCA mutation evaluation in a prospective triple-negative breast cancer registry: implications for hereditary breast and/or ovarian cancer syndrome testing. Breast Cancer Res Treat 145, 707-714 (2014). 38. Lee, A. & Djamgoz, M.B.A. Triple negative breast cancer: Emerging therapeutic modalities and novel combination therapies. Cancer treatment reviews 62, 110-122 (2018). 39. Geenen, J.J.J., Linn, S.C., Beijnen, J.H. & Schellens, J.H.M. PARP Inhibitors in the Treatment of Triple-Negative Breast Cancer. Clinical pharmacokinetics 57, 427-437 (2018). 40. Ribatti, D. The concept of immune surveillance against tumors. The first theories. Oncotarget 8, 7175-7180 (2017). 41. Hanahan, D. & Weinberg, R.A. Hallmarks of cancer: the next generation. Cell 144, 646-674 (2011). 42. Heinrich, E.L., Walser, T.C., Krysan, K., Liclican, E.L., Grant, J.L., Rodriguez, N.L. & Dubinett, S.M. The inflammatory tumor microenvironment, epithelial mesenchymal transition and lung carcinogenesis. Cancer Microenviron 5, 5-18 (2012). 43. Pardoll, D.M. Immunology beats cancer: a blueprint for successful translation. Nat Immunol 13, 1129-1132 (2012). 44. Kelderman, S. & Kvistborg, P. Tumor antigens in human cancer control. Biochimica et biophysica acta 1865, 83-89 (2016). 131  45. Angelova, M., Mlecnik, B., Vasaturo, A., Bindea, G., Fredriksen, T., Lafontaine, L., Buttard, B., Morgand, E., Bruni, D., Jouret-Mourin, A., Hubert, C., Kartheuser, A., Humblet, Y., Ceccarelli, M., Syed, N., Marincola, F.M., Bedognetti, D., Van den Eynde, M. & Galon, J. Evolution of Metastases in Space and Time under Immune Selection. Cell 175, 751-765 e716 (2018). 46. Beatty, G.L. & Gladney, W.L. Immune escape mechanisms as a guide for cancer immunotherapy. Clin Cancer Res 21, 687-692 (2015). 47. Alexandrov, L.B., Nik-Zainal, S., Wedge, D.C., Aparicio, S.A., Behjati, S., Biankin, A.V., Bignell, G.R., Bolli, N., Borg, A., Borresen-Dale, A.L., Boyault, S., Burkhardt, B., Butler, A.P., Caldas, C., Davies, H.R., Desmedt, C., Eils, R., Eyfjord, J.E., Foekens, J.A., Greaves, M., Hosoda, F., Hutter, B., Ilicic, T., Imbeaud, S., Imielinski, M., Jager, N., Jones, D.T., Jones, D., Knappskog, S., Kool, M., Lakhani, S.R., Lopez-Otin, C., Martin, S., Munshi, N.C., Nakamura, H., Northcott, P.A., Pajic, M., Papaemmanuil, E., Paradiso, A., Pearson, J.V., Puente, X.S., Raine, K., Ramakrishna, M., Richardson, A.L., Richter, J., Rosenstiel, P., Schlesner, M., Schumacher, T.N., Span, P.N., Teague, J.W., Totoki, Y., Tutt, A.N., Valdes-Mas, R., van Buuren, M.M., van 't Veer, L., Vincent-Salomon, A., Waddell, N., Yates, L.R., Australian Pancreatic Cancer Genome, I., Consortium, I.B.C., Consortium, I.M.-S., PedBrain, I., Zucman-Rossi, J., Futreal, P.A., McDermott, U., Lichter, P., Meyerson, M., Grimmond, S.M., Siebert, R., Campo, E., Shibata, T., Pfister, S.M., Campbell, P.J. & Stratton, M.R. Signatures of mutational processes in human cancer. Nature 500, 415-421 (2013). 48. Vogelstein, B., Papadopoulos, N., Velculescu, V.E., Zhou, S., Diaz, L.A., Jr. & Kinzler, K.W. Cancer genome landscapes. Science 339, 1546-1558 (2013). 49. Shalapour, S., Font-Burgada, J., Di Caro, G., Zhong, Z., Sanchez-Lopez, E., Dhar, D., Willimsky, G., Ammirante, M., Strasner, A., Hansel, D.E., Jamieson, C., Kane, C.J., Klatte, T., Birner, P., Kenner, L. & Karin, M. Immunosuppressive plasma cells impede T-cell-dependent immunogenic chemotherapy. Nature 521, 94-98 (2015). 50. Pereira, B., Chin, S.F., Rueda, O.M., Vollan, H.K., Provenzano, E., Bardwell, H.A., Pugh, M., Jones, L., Russell, R., Sammut, S.J., Tsui, D.W., Liu, B., Dawson, S.J., Abraham, J., Northen, H., Peden, J.F., Mukherjee, A., Turashvili, G., Green, A.R., McKinney, S., Oloumi, A., Shah, S., Rosenfeld, N., Murphy, L., Bentley, D.R., Ellis, I.O., Purushotham, A., Pinder, S.E., Borresen-Dale, A.L., Earl, H.M., Pharoah, P.D., Ross, M.T., Aparicio, S. & Caldas, C. The somatic mutation profiles of 2,433 breast cancers refines their genomic and transcriptomic landscapes. Nat Commun 7, 11479 (2016). 51. Ruffell, B., Au, A., Rugo, H.S., Esserman, L.J., Hwang, E.S. & Coussens, L.M. Leukocyte composition of human breast cancer. Proc Natl Acad Sci U S A 109, 2796-2801 (2012). 52. Woo, S.R., Corrales, L. & Gajewski, T.F. Innate immune recognition of cancer. Annu Rev Immunol 33, 445-474 (2015). 53. Gregory, A.D. & Houghton, A.M. Tumor-associated neutrophils: new targets for cancer therapy. Cancer Res 71, 2411-2416 (2011). 54. Casbon, A.J., Reynaud, D., Park, C., Khuc, E., Gan, D.D., Schepers, K., Passegue, E. & Werb, Z. Invasive breast cancer reprograms early myeloid differentiation in the bone 132  marrow to generate immunosuppressive neutrophils. Proc Natl Acad Sci U S A 112, E566-575 (2015). 55. Eruslanov, E.B., Bhojnagarwala, P.S., Quatromoni, J.G., Stephen, T.L., Ranganathan, A., Deshpande, C., Akimova, T., Vachani, A., Litzky, L., Hancock, W.W., Conejo-Garcia, J.R., Feldman, M., Albelda, S.M. & Singhal, S. Tumor-associated neutrophils stimulate T cell responses in early-stage human lung cancer. J Clin Invest 124, 5466-5480 (2014). 56. Weber, R., Fleming, V., Hu, X., Nagibin, V., Groth, C., Altevogt, P., Utikal, J. & Umansky, V. Myeloid-Derived Suppressor Cells Hinder the Anti-Cancer Activity of Immune Checkpoint Inhibitors. Front Immunol 9, 1310 (2018). 57. Budhwar, S., Verma, P., Verma, R., Rai, S. & Singh, K. The Yin and Yang of Myeloid Derived Suppressor Cells. Front Immunol 9, 2776 (2018). 58. Chafe, S.C., Lou, Y., Sceneay, J., Vallejo, M., Hamilton, M.J., McDonald, P.C., Bennewith, K.L., Moller, A. & Dedhar, S. Carbonic anhydrase IX promotes myeloid-derived suppressor cell mobilization and establishment of a metastatic niche by stimulating G-CSF production. Cancer Res 75, 996-1008 (2015). 59. Vincent, J., Mignot, G., Chalmin, F., Ladoire, S., Bruchard, M., Chevriaux, A., Martin, F., Apetoh, L., Rebe, C. & Ghiringhelli, F. 5-Fluorouracil selectively kills tumor-associated myeloid-derived suppressor cells resulting in enhanced T cell-dependent antitumor immunity. Cancer Res 70, 3052-3061 (2010). 60. Yu, J., Du, W., Yan, F., Wang, Y., Li, H., Cao, S., Yu, W., Shen, C., Liu, J. & Ren, X. Myeloid-derived suppressor cells suppress antitumor immune responses through IDO expression and correlate with lymph node metastasis in patients with breast cancer. J Immunol 190, 3783-3797 (2013). 61. Markowitz, J., Wesolowski, R., Papenfuss, T., Brooks, T.R. & Carson, W.E., 3rd. Myeloid-derived suppressor cells in breast cancer. Breast Cancer Res Treat 140, 13-21 (2013). 62. Tang, X. Tumor-associated macrophages as potential diagnostic and prognostic biomarkers in breast cancer. Cancer Lett 332, 3-10 (2013). 63. Medrek, C., Ponten, F., Jirstrom, K. & Leandersson, K. The presence of tumor associated macrophages in tumor stroma as a prognostic marker for breast cancer patients. BMC Cancer 12, 306 (2012). 64. Mahmoud, S.M., Lee, A.H., Paish, E.C., Macmillan, R.D., Ellis, I.O. & Green, A.R. Tumour-infiltrating macrophages and clinical outcome in breast cancer. J Clin Pathol 65, 159-163 (2012). 65. Raphael, J., Gong, I.Y., Nofech-Mozes, S., Bartlett, J., Nafisi, H. & Verma, S. Tumour infiltrating lymphocytes and stromal CD68 in early stage HER2 positive breast cancer. J Clin Pathol (2016). 66. Tiainen, S., Tumelius, R., Rilla, K., Hamalainen, K., Tammi, M., Tammi, R., Kosma, V.M., Oikari, S. & Auvinen, P. High numbers of macrophages, especially M2-like (CD163-positive), correlate with hyaluronan accumulation and poor outcome in breast cancer. Histopathology 66, 873-883 (2015). 67. Hollmen, M., Roudnicky, F., Karaman, S. & Detmar, M. Characterization of macrophage--cancer cell crosstalk in estrogen receptor positive and triple-negative breast cancer. Sci Rep 5, 9188 (2015). 133  68. Sznol, M. & Chen, L. Antagonist antibodies to PD-1 and B7-H1 (PD-L1) in the treatment of advanced human cancer. Clin Cancer Res 19, 1021-1034 (2013). 69. Spranger, S., Spaapen, R.M., Zha, Y., Williams, J., Meng, Y., Ha, T.T. & Gajewski, T.F. Up-regulation of PD-L1, IDO, and T(regs) in the melanoma tumor microenvironment is driven by CD8(+) T cells. Sci Transl Med 5, 200ra116 (2013). 70. Stanton, S.E., Adams, S. & Disis, M.L. Variation in the Incidence and Magnitude of Tumor-Infiltrating Lymphocytes in Breast Cancer Subtypes: A Systematic Review. JAMA Oncol 2, 1354-1360 (2016). 71. Denkert, C., Loibl, S., Noske, A., Roller, M., Muller, B.M., Komor, M., Budczies, J., Darb-Esfahani, S., Kronenwett, R., Hanusch, C., von Torne, C., Weichert, W., Engels, K., Solbach, C., Schrader, I., Dietel, M. & von Minckwitz, G. Tumor-associated lymphocytes as an independent predictor of response to neoadjuvant chemotherapy in breast cancer. J Clin Oncol 28, 105-113 (2010). 72. Figenschau, S.L., Fismen, S., Fenton, K.A., Fenton, C. & Mortensen, E.S. Tertiary lymphoid structures are associated with higher tumor grade in primary operable breast cancer patients. BMC Cancer 15, 101 (2015). 73. Kroemer, G., Senovilla, L., Galluzzi, L., Andre, F. & Zitvogel, L. Natural and therapy-induced immunosurveillance in breast cancer. Nat Med 21, 1128-1138 (2015). 74. Salgado, R., Denkert, C., Demaria, S., Sirtaine, N., Klauschen, F., Pruneri, G., Wienert, S., Van den Eynden, G., Baehner, F.L., Penault-Llorca, F., Perez, E.A., Thompson, E.A., Symmans, W.F., Richardson, A.L., Brock, J., Criscitiello, C., Bailey, H., Ignatiadis, M., Floris, G., Sparano, J., Kos, Z., Nielsen, T., Rimm, D.L., Allison, K.H., Reis-Filho, J.S., Loibl, S., Sotiriou, C., Viale, G., Badve, S., Adams, S., Willard-Gallo, K., Loi, S. & International, T.W.G. The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TILs Working Group 2014. Ann Oncol 26, 259-271 (2015). 75. Aaltomaa, S., Lipponen, P., Eskelinen, M., Kosma, V.M., Marin, S., Alhava, E. & Syrjanen, K. Lymphocyte infiltrates as a prognostic variable in female breast cancer. Eur J Cancer 28A, 859-864 (1992). 76. Shaveta Vinayak, R.J.G., Sylvia Adams, Kristin C. Jensen, Judith Manola, Anosheh Afghahi, Lori J. Goldstein, James M. Ford, Sunil S. Badve, Melinda L. Telli. Association of increased tumor-infiltrating lymphocytes (TILs) with immunomodulatory (IM) triple-negative breast cancer (TNBC) subtype and response to neoadjuvant platinum-based therapy in PrECOG0105. J Clin Oncol 32:5s, 2014 (suppl; abstr 1000^) (2014). 77. Loi, S., Drubay, D., Adams, S., Pruneri, G., Francis, P.A., Lacroix-Triki, M., Joensuu, H., Dieci, M.V., Badve, S., Demaria, S., Gray, R., Munzone, E., Lemonnier, J., Sotiriou, C., Piccart, M.J., Kellokumpu-Lehtinen, P.L., Vingiani, A., Gray, K., Andre, F., Denkert, C., Salgado, R. & Michiels, S. Tumor-Infiltrating Lymphocytes and Prognosis: A Pooled Individual Patient Analysis of Early-Stage Triple-Negative Breast Cancers. J Clin Oncol, JCO1801010 (2019). 78. Denkert, C., von Minckwitz, G., Darb-Esfahani, S., Lederer, B., Heppner, B.I., Weber, K.E., Budczies, J., Huober, J., Klauschen, F., Furlanetto, J., Schmitt, W.D., Blohmer, J.U., Karn, T., Pfitzner, B.M., Kummel, S., Engels, K., Schneeweiss, A., Hartmann, A., Noske, A., Fasching, P.A., Jackisch, C., van Mackelenbergh, M., Sinn, P., Schem, C., Hanusch, C., Untch, M. & Loibl, S. Tumour-infiltrating lymphocytes and prognosis in 134  different subtypes of breast cancer: a pooled analysis of 3771 patients treated with neoadjuvant therapy. The Lancet. Oncology 19, 40-50 (2018). 79. Adams, S., Gray, R.J., Demaria, S., Goldstein, L., Perez, E.A., Shulman, L.N., Martino, S., Wang, M., Jones, V.E., Saphner, T.J., Wolff, A.C., Wood, W.C., Davidson, N.E., Sledge, G.W., Sparano, J.A. & Badve, S.S. Prognostic value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1199. J Clin Oncol 32, 2959-2966 (2014). 80. Loi, S., Michiels, S., Salgado, R., Sirtaine, N., Jose, V., Fumagalli, D., Kellokumpu-Lehtinen, P.L., Bono, P., Kataja, V., Desmedt, C., Piccart, M.J., Loibl, S., Denkert, C., Smyth, M.J., Joensuu, H. & Sotiriou, C. Tumor infiltrating lymphocytes are prognostic in triple negative breast cancer and predictive for trastuzumab benefit in early breast cancer: results from the FinHER trial. Ann Oncol 25, 1544-1550 (2014). 81. Loi, S., Sirtaine, N., Piette, F., Salgado, R., Viale, G., Van Eenoo, F., Rouas, G., Francis, P., Crown, J.P., Hitre, E., de Azambuja, E., Quinaux, E., Di Leo, A., Michiels, S., Piccart, M.J. & Sotiriou, C. Prognostic and predictive value of tumor-infiltrating lymphocytes in a phase III randomized adjuvant breast cancer trial in node-positive breast cancer comparing the addition of docetaxel to doxorubicin with doxorubicin-based chemotherapy: BIG 02-98. J Clin Oncol 31, 860-867 (2013). 82. Perez, E.A., Ballman, K.V., Tenner, K.S., Thompson, E.A., Badve, S.S., Bailey, H. & Baehner, F.L. Association of Stromal Tumor-Infiltrating Lymphocytes With Recurrence-Free Survival in the N9831 Adjuvant Trial in Patients With Early-Stage HER2-Positive Breast Cancer. JAMA Oncol 2, 56-64 (2016). 83. Dieci, M.V., Mathieu, M.C., Guarneri, V., Conte, P., Delaloge, S., Andre, F. & Goubar, A. Prognostic and predictive value of tumor-infiltrating lymphocytes in two phase III randomized adjuvant breast cancer trials. Ann Oncol 26, 1698-1704 (2015). 84. Loi S., M.S., Salgado R., Sirtaine N., Jose V., Fumagalli D et al. Tumor infiltrating lymphocytes (TILs) indicate trastuzumab benefit in early-stage HER2-positive breast cancer[abstract]. Cancer Res abstract nr S1-05(2013). 85. Nelson, B.H. CD20+ B cells: the other tumor-infiltrating lymphocytes. J Immunol 185, 4977-4982 (2010). 86. Mahmoud, S.M., Lee, A.H., Paish, E.C., Macmillan, R.D., Ellis, I.O. & Green, A.R. The prognostic significance of B lymphocytes in invasive carcinoma of the breast. Breast Cancer Res Treat 132, 545-553 (2012). 87. Mohammed, Z.M., Going, J.J., Edwards, J., Elsberger, B. & McMillan, D.C. The relationship between lymphocyte subsets and clinico-pathological determinants of survival in patients with primary operable invasive ductal breast cancer. Br J Cancer 109, 1676-1684 (2013). 88. Wouters, M.C.A. & Nelson, B.H. Prognostic Significance of Tumor-Infiltrating B Cells and Plasma Cells in Human Cancer. Clin Cancer Res 24, 6125-6135 (2018). 89. Shen, M., Wang, J. & Ren, X. New Insights into Tumor-Infiltrating B Lymphocytes in Breast Cancer: Clinical Impacts and Regulatory Mechanisms. Front Immunol 9, 470 (2018). 90. Mohammed, Z.M., Going, J.J., Edwards, J., Elsberger, B., Doughty, J.C. & McMillan, D.C. The relationship between components of tumour inflammatory cell infiltrate and 135  clinicopathological factors and survival in patients with primary operable invasive ductal breast cancer. Br J Cancer 107, 864-873 (2012). 91. Chen, Z., Gerhold-Ay, A., Gebhard, S., Boehm, D., Solbach, C., Lebrecht, A., Battista, M., Sicking, I., Cotarelo, C., Cadenas, C., Marchan, R., Stewart, J.D., Gehrmann, M., Koelbl, H., Hengstler, J.G. & Schmidt, M. Immunoglobulin kappa C predicts overall survival in node-negative breast cancer. PLoS One 7, e44741 (2012). 92. Schmidt, M., Hellwig, B., Hammad, S., Othman, A., Lohr, M., Chen, Z., Boehm, D., Gebhard, S., Petry, I., Lebrecht, A., Cadenas, C., Marchan, R., Stewart, J.D., Solbach, C., Holmberg, L., Edlund, K., Kultima, H.G., Rody, A., Berglund, A., Lambe, M., Isaksson, A., Botling, J., Karn, T., Muller, V., Gerhold-Ay, A., Cotarelo, C., Sebastian, M., Kronenwett, R., Bojar, H., Lehr, H.A., Sahin, U., Koelbl, H., Gehrmann, M., Micke, P., Rahnenfuhrer, J. & Hengstler, J.G. A comprehensive analysis of human gene expression profiles identifies stromal immunoglobulin kappa C as a compatible prognostic marker in human solid tumors. Clin Cancer Res 18, 2695-2703 (2012). 93. Olkhanud, P.B., Damdinsuren, B., Bodogai, M., Gress, R.E., Sen, R., Wejksza, K., Malchinkhuu, E., Wersto, R.P. & Biragyn, A. Tumor-evoked regulatory B cells promote breast cancer metastasis by converting resting CD4(+) T cells to T-regulatory cells. Cancer Res 71, 3505-3515 (2011). 94. Iglesia, M.D., Vincent, B.G., Parker, J.S., Hoadley, K.A., Carey, L.A., Perou, C.M. & Serody, J.S. Prognostic B-cell signatures using mRNA-seq in patients with subtype-specific breast and ovarian cancer. Clin Cancer Res 20, 3818-3829 (2014). 95. Mahmoud, S.M., Paish, E.C., Powe, D.G., Macmillan, R.D., Grainge, M.J., Lee, A.H., Ellis, I.O. & Green, A.R. Tumor-infiltrating CD8+ lymphocytes predict clinical outcome in breast cancer. J Clin Oncol 29, 1949-1955 (2011). 96. Liu, S., Lachapelle, J., Leung, S., Gao, D., Foulkes, W.D. & Nielsen, T.O. CD8+ lymphocyte infiltration is an independent favorable prognostic indicator in basal-like breast cancer. Breast Cancer Res 14, R48 (2012). 97. Ali, H.R., Provenzano, E., Dawson, S.J., Blows, F.M., Liu, B., Shah, M., Earl, H.M., Poole, C.J., Hiller, L., Dunn, J.A., Bowden, S.J., Twelves, C., Bartlett, J.M., Mahmoud, S.M., Rakha, E., Ellis, I.O., Liu, S., Gao, D., Nielsen, T.O., Pharoah, P.D. & Caldas, C. Association between CD8+ T-cell infiltration and breast cancer survival in 12,439 patients. Ann Oncol 25, 1536-1543 (2014). 98. Dookeran, K.A., Dignam, J.J., Ferrer, K., Sekosan, M., McCaskill-Stevens, W. & Gehlert, S. p53 as a marker of prognosis in African-American women with breast cancer. Annals of surgical oncology 17, 1398-1405 (2010). 99. Winslow, S., Leandersson, K., Edsjo, A. & Larsson, C. Prognostic stromal gene signatures in breast cancer. Breast Cancer Res 17, 23 (2015). 100. West, N.R., Milne, K., Truong, P.T., Macpherson, N., Nelson, B.H. & Watson, P.H. Tumor-infiltrating lymphocytes predict response to anthracycline-based chemotherapy in estrogen receptor-negative breast cancer. Breast Cancer Res 13, R126 (2011). 101. Liu, S., Chen, B., Burugu, S., Leung, S., Gao, D., Virk, S., Kos, Z., Parulekar, W.R., Shepherd, L., Gelmon, K.A. & Nielsen, T.O. Role of Cytotoxic Tumor-Infiltrating Lymphocytes in Predicting Outcomes in Metastatic HER2-Positive Breast Cancer: A Secondary Analysis of a Randomized Clinical Trial. JAMA Oncol 3, e172085 (2017). 136  102. Lee, H.J., Seo, J.Y., Ahn, J.H., Ahn, S.H. & Gong, G. Tumor-associated lymphocytes predict response to neoadjuvant chemotherapy in breast cancer patients. J Breast Cancer 16, 32-39 (2013). 103. Oda, N., Shimazu, K., Naoi, Y., Morimoto, K., Shimomura, A., Shimoda, M., Kagara, N., Maruyama, N., Kim, S.J. & Noguchi, S. Intratumoral regulatory T cells as an independent predictive factor for pathological complete response to neoadjuvant paclitaxel followed by 5-FU/epirubicin/cyclophosphamide in breast cancer patients. Breast Cancer Res Treat 136, 107-116 (2012). 104. Seo, A.N., Lee, H.J., Kim, E.J., Kim, H.J., Jang, M.H., Lee, H.E., Kim, Y.J., Kim, J.H. & Park, S.Y. Tumour-infiltrating CD8+ lymphocytes as an independent predictive factor for pathological complete response to primary systemic therapy in breast cancer. Br J Cancer 109, 2705-2713 (2013). 105. Zhu, J., Yamane, H. & Paul, W.E. Differentiation of effector CD4 T cell populations (*). Annu Rev Immunol 28, 445-489 (2010). 106. Gu-Trantien, C., Loi, S., Garaud, S., Equeter, C., Libin, M., de Wind, A., Ravoet, M., Le Buanec, H., Sibille, C., Manfouo-Foutsop, G., Veys, I., Haibe-Kains, B., Singhal, S.K., Michiels, S., Rothe, F., Salgado, R., Duvillier, H., Ignatiadis, M., Desmedt, C., Bron, D., Larsimont, D., Piccart, M., Sotiriou, C. & Willard-Gallo, K. CD4(+) follicular helper T cell infiltration predicts breast cancer survival. J Clin Invest 123, 2873-2892 (2013). 107. Liu, S., Foulkes, W.D., Leung, S., Gao, D., Lau, S., Kos, Z. & Nielsen, T.O. Prognostic significance of FOXP3+ tumor-infiltrating lymphocytes in breast cancer depends on estrogen receptor and human epidermal growth factor receptor-2 expression status and concurrent cytotoxic T-cell infiltration. Breast Cancer Res 16, 432 (2014). 108. Droeser, R., Zlobec, I., Kilic, E., Guth, U., Heberer, M., Spagnoli, G., Oertli, D. & Tapia, C. Differential pattern and prognostic significance of CD4+, FOXP3+ and IL-17+ tumor infiltrating lymphocytes in ductal and lobular breast cancers. BMC Cancer 12, 134 (2012). 109. Lee, S., Cho, E.Y., Park, Y.H., Ahn, J.S. & Im, Y.H. Prognostic impact of FOXP3 expression in triple-negative breast cancer. Acta oncologica (Stockholm, Sweden) 52, 73-81 (2013). 110. West, N.R., Kost, S.E., Martin, S.D., Milne, K., Deleeuw, R.J., Nelson, B.H. & Watson, P.H. Tumour-infiltrating FOXP3(+) lymphocytes are associated with cytotoxic immune responses and good clinical outcome in oestrogen receptor-negative breast cancer. Br J Cancer 108, 155-162 (2013). 111. Kim, S.T., Jeong, H., Woo, O.H., Seo, J.H., Kim, A., Lee, E.S., Shin, S.W., Kim, Y.H., Kim, J.S. & Park, K.H. Tumor-infiltrating lymphocytes, tumor characteristics, and recurrence in patients with early breast cancer. Am J Clin Oncol 36, 224-231 (2013). 112. Bindea, G., Mlecnik, B., Tosolini, M., Kirilovsky, A., Waldner, M., Obenauf, A.C., Angell, H., Fredriksen, T., Lafontaine, L., Berger, A., Bruneval, P., Fridman, W.H., Becker, C., Pages, F., Speicher, M.R., Trajanoski, Z. & Galon, J. Spatiotemporal dynamics of intratumoral immune cells reveal the immune landscape in human cancer. Immunity 39, 782-795 (2013). 113. Ma, C.S., Deenick, E.K., Batten, M. & Tangye, S.G. The origins, function, and regulation of T follicular helper cells. J Exp Med 209, 1241-1253 (2012). 137  114. Santos, P.M. & Butterfield, L.H. Dendritic Cell-Based Cancer Vaccines. J Immunol 200, 443-449 (2018). 115. Butterfield, L.H. Cancer vaccines. BMJ (Clinical research ed.) 350, h988 (2015). 116. Kantoff, P.W., Higano, C.S., Shore, N.D., Berger, E.R., Small, E.J., Penson, D.F., Redfern, C.H., Ferrari, A.C., Dreicer, R., Sims, R.B., Xu, Y., Frohlich, M.W. & Schellhammer, P.F. Sipuleucel-T immunotherapy for castration-resistant prostate cancer. N Engl J Med 363, 411-422 (2010). 117. Benedetti, R., Dell'Aversana, C., Giorgio, C., Astorri, R. & Altucci, L. Breast Cancer Vaccines: New Insights. Frontiers in endocrinology 8, 270 (2017). 118. Yang, J.C. & Rosenberg, S.A. Adoptive T-Cell Therapy for Cancer. Advances in immunology 130, 279-294 (2016). 119. Newick, K., O'Brien, S., Moon, E. & Albelda, S.M. CAR T Cell Therapy for Solid Tumors. Annual review of medicine 68, 139-152 (2017). 120. Savas, P., Salgado, R., Denkert, C., Sotiriou, C., Darcy, P.K., Smyth, M.J. & Loi, S. Clinical relevance of host immunity in breast cancer: from TILs to the clinic. Nat Rev Clin Oncol (2015). 121. Zacharakis, N., Chinnasamy, H., Black, M., Xu, H., Lu, Y.C., Zheng, Z., Pasetto, A., Langhan, M., Shelton, T., Prickett, T., Gartner, J., Jia, L., Trebska-McGowan, K., Somerville, R.P., Robbins, P.F., Rosenberg, S.A., Goff, S.L. & Feldman, S.A. Immune recognition of somatic mutations leading to complete durable regression in metastatic breast cancer. Nat Med 24, 724-730 (2018). 122. Kroemer, G., Galluzzi, L., Kepp, O. & Zitvogel, L. Immunogenic cell death in cancer therapy. Annu Rev Immunol 31, 51-72 (2013). 123. Minn, A.J. Interferons and the Immunogenic Effects of Cancer Therapy. Trends in immunology 36, 725-737 (2015). 124. Alizadeh, D. & Larmonier, N. Chemotherapeutic targeting of cancer-induced immunosuppressive cells. Cancer Res 74, 2663-2668 (2014). 125. Brix, N., Tiefenthaller, A., Anders, H., Belka, C. & Lauber, K. Abscopal, immunological effects of radiotherapy: Narrowing the gap between clinical and preclinical experiences. Immunological reviews 280, 249-279 (2017). 126. Derer, A., Deloch, L., Rubner, Y., Fietkau, R., Frey, B. & Gaipl, U.S. Radio-Immunotherapy-Induced Immunogenic Cancer Cells as Basis for Induction of Systemic Anti-Tumor Immune Responses - Pre-Clinical Evidence and Ongoing Clinical Applications. Front Immunol 6, 505 (2015). 127. Reynders, K., Illidge, T., Siva, S., Chang, J.Y. & De Ruysscher, D. The abscopal effect of local radiotherapy: using immunotherapy to make a rare event clinically relevant. Cancer treatment reviews 41, 503-510 (2015). 128. Bang, Y.J., Giaccone, G., Im, S.A., Oh, D.Y., Bauer, T.M., Nordstrom, J.L., Li, H., Chichili, G.R., Moore, P.A., Hong, S., Stewart, S.J., Baughman, J.E., Lechleider, R.J. & Burris, H.A. First-in-human phase 1 study of margetuximab (MGAH22), an Fc-modified chimeric monoclonal antibody, in patients with HER2-positive advanced solid tumors. Ann Oncol 28, 855-861 (2017). 129. Musolino, A., Naldi, N., Bortesi, B., Pezzuolo, D., Capelletti, M., Missale, G., Laccabue, D., Zerbini, A., Camisa, R., Bisagni, G., Neri, T.M. & Ardizzoni, A. Immunoglobulin G fragment C receptor polymorphisms and clinical efficacy of trastuzumab-based therapy in 138  patients with HER-2/neu-positive metastatic breast cancer. J Clin Oncol 26, 1789-1796 (2008). 130. Shi, Y., Fan, X., Deng, H., Brezski, R.J., Rycyzyn, M., Jordan, R.E., Strohl, W.R., Zou, Q., Zhang, N. & An, Z. Trastuzumab triggers phagocytic killing of high HER2 cancer cells in vitro and in vivo by interaction with Fcgamma receptors on macrophages. J Immunol 194, 4379-4386 (2015). 131. Ferris, R.L., Jaffee, E.M. & Ferrone, S. Tumor antigen-targeted, monoclonal antibody-based immunotherapy: clinical response, cellular immunity, and immunoescape. J Clin Oncol 28, 4390-4399 (2010). 132. Weiner, G.J. Building better monoclonal antibody-based therapeutics. Nat Rev Cancer 15, 361-370 (2015). 133. Gul, N. & van Egmond, M. Antibody-Dependent Phagocytosis of Tumor Cells by Macrophages: A Potent Effector Mechanism of Monoclonal Antibody Therapy of Cancer. Cancer Res 75, 5008-5013 (2015). 134. Nguyen, L.T. & Ohashi, P.S. Clinical blockade of PD1 and LAG3--potential mechanisms of action. Nat Rev Immunol 15, 45-56 (2015). 135. Weber, J. Immune checkpoint proteins: a new therapeutic paradigm for cancer--preclinical background: CTLA-4 and PD-1 blockade. Seminars in oncology 37, 430-439 (2010). 136. Gargi D Basu, A.G., Randal Vader, Sandeep Reddy, Karen Anderson, Ann McCullough, and Barbara Pockaj Expression of novel immunotherapeutic targets in luminal breast cancer patients [abstract]. Cancer Res Abstract nr P5-04-08(2015). 137. Schmidt M, v.d.S.L., Heimes A, Battista M, Lebrecht A, Almstedt K, Hoffmann G, Rahnenführer J, Hengstler JG Prognostic significance of immune checkpoint receptors in node-negative breast cancer[abstract]. Cancer res Abstract nr P2-08-07(2015). 138. Larkin, J., Chiarion-Sileni, V., Gonzalez, R., Grob, J.J., Cowey, C.L., Lao, C.D., Schadendorf, D., Dummer, R., Smylie, M., Rutkowski, P., Ferrucci, P.F., Hill, A., Wagstaff, J., Carlino, M.S., Haanen, J.B., Maio, M., Marquez-Rodas, I., McArthur, G.A., Ascierto, P.A., Long, G.V., Callahan, M.K., Postow, M.A., Grossmann, K., Sznol, M., Dreno, B., Bastholt, L., Yang, A., Rollin, L.M., Horak, C., Hodi, F.S. & Wolchok, J.D. Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma. N Engl J Med 373, 23-34 (2015). 139. Robert, C., Thomas, L., Bondarenko, I., O'Day, S., Weber, J., Garbe, C., Lebbe, C., Baurain, J.F., Testori, A., Grob, J.J., Davidson, N., Richards, J., Maio, M., Hauschild, A., Miller, W.H., Jr., Gascon, P., Lotem, M., Harmankaya, K., Ibrahim, R., Francis, S., Chen, T.T., Humphrey, R., Hoos, A. & Wolchok, J.D. Ipilimumab plus dacarbazine for previously untreated metastatic melanoma. N Engl J Med 364, 2517-2526 (2011). 140. Schadendorf, D., Hodi, F.S., Robert, C., Weber, J.S., Margolin, K., Hamid, O., Patt, D., Chen, T.T., Berman, D.M. & Wolchok, J.D. Pooled Analysis of Long-Term Survival Data From Phase II and Phase III Trials of Ipilimumab in Unresectable or Metastatic Melanoma. J Clin Oncol 33, 1889-1894 (2015). 141. Robert, C., Long, G.V., Brady, B., Dutriaux, C., Maio, M., Mortier, L., Hassel, J.C., Rutkowski, P., McNeil, C., Kalinka-Warzocha, E., Savage, K.J., Hernberg, M.M., Lebbe, C., Charles, J., Mihalcioiu, C., Chiarion-Sileni, V., Mauch, C., Cognetti, F., Arance, A., Schmidt, H., Schadendorf, D., Gogas, H., Lundgren-Eriksson, L., Horak, C., Sharkey, B., 139  Waxman, I.M., Atkinson, V. & Ascierto, P.A. Nivolumab in previously untreated melanoma without BRAF mutation. N Engl J Med 372, 320-330 (2015). 142. Vonderheide, R.H., LoRusso, P.M., Khalil, M., Gartner, E.M., Khaira, D., Soulieres, D., Dorazio, P., Trosko, J.A., Ruter, J., Mariani, G.L., Usari, T. & Domchek, S.M. Tremelimumab in combination with exemestane in patients with advanced breast cancer and treatment-associated modulation of inducible costimulator expression on patient T cells. Clin Cancer Res 16, 3485-3494 (2010). 143. Freeman, G.J., Long, A.J., Iwai, Y., Bourque, K., Chernova, T., Nishimura, H., Fitz, L.J., Malenkovich, N., Okazaki, T., Byrne, M.C., Horton, H.F., Fouser, L., Carter, L., Ling, V., Bowman, M.R., Carreno, B.M., Collins, M., Wood, C.R. & Honjo, T. Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads to negative regulation of lymphocyte activation. J Exp Med 192, 1027-1034 (2000). 144. Keir, M.E., Butte, M.J., Freeman, G.J. & Sharpe, A.H. PD-1 and its ligands in tolerance and immunity. Annu Rev Immunol 26, 677-704 (2008). 145. Hui, E., Cheung, J., Zhu, J., Su, X., Taylor, M.J., Wallweber, H.A., Sasmal, D.K., Huang, J., Kim, J.M., Mellman, I. & Vale, R.D. T cell costimulatory receptor CD28 is a primary target for PD-1-mediated inhibition. Science 355, 1428-1433 (2017). 146. Gong, J., Chehrazi-Raffle, A., Reddi, S. & Salgia, R. Development of PD-1 and PD-L1 inhibitors as a form of cancer immunotherapy: a comprehensive review of registration trials and future considerations. J Immunother Cancer 6, 8 (2018). 147. Wein, L., Luen, S.J., Savas, P., Salgado, R. & Loi, S. Checkpoint blockade in the treatment of breast cancer: current status and future directions. Br J Cancer 119, 4-11 (2018). 148. Schmid, P., Adams, S., Rugo, H.S., Schneeweiss, A., Barrios, C.H., Iwata, H., Dieras, V., Hegg, R., Im, S.A., Shaw Wright, G., Henschel, V., Molinero, L., Chui, S.Y., Funke, R., Husain, A., Winer, E.P., Loi, S. & Emens, L.A. Atezolizumab and Nab-Paclitaxel in Advanced Triple-Negative Breast Cancer. N Engl J Med 379, 2108-2121 (2018). 149. Rugo, H.S., Delord, J.P., Im, S.A., Ott, P.A., Piha-Paul, S.A., Bedard, P.L., Sachdev, J., Tourneau, C.L., van Brummelen, E.M.J., Varga, A., Salgado, R., Loi, S., Saraf, S., Pietrangelo, D., Karantza, V. & Tan, A.R. Safety and Antitumor Activity of Pembrolizumab in Patients with Estrogen Receptor-Positive/Human Epidermal Growth Factor Receptor 2-Negative Advanced Breast Cancer. Clin Cancer Res 24, 2804-2811 (2018). 150. Vikas, P., Borcherding, N. & Zhang, W. The clinical promise of immunotherapy in triple-negative breast cancer. Cancer management and research 10, 6823-6833 (2018). 151. Adams, S., Schmid, P., Rugo, H.S., Winer, E.P., Loirat, D., Awada, A., Cescon, D.W., Iwata, H., Campone, M., Nanda, R., Hui, R., Curigliano, G., Toppmeyer, D., O'Shaughnessy, J., Loi, S., Paluch-Shimon, S., Tan, A.R., Card, D., Zhao, J., Karantza, V. & Cortes, J. Pembrolizumab Monotherapy for Previously Treated Metastatic Triple-Negative Breast Cancer: Cohort A of the Phase 2 KEYNOTE-086 Study. Ann Oncol (2018). 152. Adams, S., Loi, S., Toppmeyer, D., Cescon, D.W., De Laurentiis, M., Nanda, R., Winer, E.P., Mukai, H., Tamura, K., Armstrong, A., Liu, M.C., Iwata, H., Ryvo, L., Wimberger, P., Rugo, H.S., Tan, A.R., Jia, L., Ding, Y., Karantza, V. & Schmid, P. Title: Pembrolizumab Monotherapy for Previously Untreated, PD-L1-Positive, Metastatic 140  Triple-Negative Breast Cancer: Cohort B of the Phase 2 KEYNOTE-086 Study. Ann Oncol (2018). 153. Ribas, A. & Hu-Lieskovan, S. What does PD-L1 positive or negative mean? J Exp Med 213, 2835-2840 (2016). 154. Page DB, Y.J., Diab A, Dong Z, Ginsberg A, Wong P et al. Integrated immunologic assessment of tumor infiltrating lymphocytes (TILs) and peripheral blood to assess synergy of cryoablation (cryo) plus ipilimumab (ipi) in early stage breast cancer (ESBC) patients[abstract]. Cancer Res abstract nr P2-15-01(2015). 155. Burugu, S., Dancsok, A.R. & Nielsen, T.O. Emerging targets in cancer immunotherapy. Seminars in cancer biology (2017). 156. Anderson, A.C., Joller, N. & Kuchroo, V.K. Lag-3, Tim-3, and TIGIT: Co-inhibitory Receptors with Specialized Functions in Immune Regulation. Immunity 44, 989-1004 (2016). 157. Scurr, M., Ladell, K., Besneux, M., Christian, A., Hockey, T., Smart, K., Bridgeman, H., Hargest, R., Phillips, S., Davies, M., Price, D., Gallimore, A. & Godkin, A. Highly prevalent colorectal cancer-infiltrating LAP(+) Foxp3(-) T cells exhibit more potent immunosuppressive activity than Foxp3(+) regulatory T cells. Mucosal Immunol 7, 428-439 (2014). 158. Bettini, M., Szymczak-Workman, A.L., Forbes, K., Castellaw, A.H., Selby, M., Pan, X., Drake, C.G., Korman, A.J. & Vignali, D.A. Cutting edge: accelerated autoimmune diabetes in the absence of LAG-3. J Immunol 187, 3493-3498 (2011). 159. Williams, J.B., Horton, B.L., Zheng, Y., Duan, Y., Powell, J.D. & Gajewski, T.F. The EGR2 targets LAG-3 and 4-1BB describe and regulate dysfunctional antigen-specific CD8+ T cells in the tumor microenvironment. J Exp Med 214, 381-400 (2017). 160. Casati, C., Camisaschi, C., Rini, F., Arienti, F., Rivoltini, L., Triebel, F., Parmiani, G. & Castelli, C. Soluble human LAG-3 molecule amplifies the in vitro generation of type 1 tumor-specific immunity. Cancer Res 66, 4450-4460 (2006). 161. Shapiro, M., Herishanu, Y., Katz, B.Z., Dezorella, N., Sun, C., Kay, S., Polliack, A., Avivi, I., Wiestner, A. & Perry, C. Lymphocyte activation gene 3: a novel therapeutic target in chronic lymphocytic leukemia. Haematologica 102, 874-882 (2017). 162. Tassi, E., Grazia, G., Vegetti, C., Bersani, I., Bertolini, G., Molla, A., Baldassari, P., Andriani, F., Roz, L., Sozzi, G., Pastorino, U., Mortarini, R. & Anichini, A. Early Effector T Lymphocytes Coexpress Multiple Inhibitory Receptors in Primary Non-Small Cell Lung Cancer. Cancer Res 77, 851-861 (2017). 163. Bottai, G., Raschioni, C., Losurdo, A., Di Tommaso, L., Tinterri, C., Torrisi, R., Reis-Filho, J.S., Roncalli, M., Sotiriou, C., Santoro, A., Mantovani, A., Loi, S. & Santarpia, L. An immune stratification reveals a subset of PD-1/LAG-3 double-positive triple-negative breast cancers. Breast cancer research : BCR 18, 121 (2016). 164. Deng, W.W., Mao, L., Yu, G.T., Bu, L.L., Ma, S.R., Liu, B., Gutkind, J.S., Kulkarni, A.B., Zhang, W.F. & Sun, Z.J. LAG-3 confers poor prognosis and its blockade reshapes antitumor response in head and neck squamous cell carcinoma. Oncoimmunology 5, e1239005 (2016). 165. Llosa, N.J., Cruise, M., Tam, A., Wicks, E.C., Hechenbleikner, E.M., Taube, J.M., Blosser, R.L., Fan, H., Wang, H., Luber, B.S., Zhang, M., Papadopoulos, N., Kinzler, K.W., Vogelstein, B., Sears, C.L., Anders, R.A., Pardoll, D.M. & Housseau, F. The 141  vigorous immune microenvironment of microsatellite instable colon cancer is balanced by multiple counter-inhibitory checkpoints. Cancer discovery 5, 43-51 (2015). 166. Meng, Q., Liu, Z., Rangelova, E., Poiret, T., Ambati, A., Rane, L., Xie, S., Verbeke, C., Dodoo, E., Del Chiaro, M., Lohr, M., Segersvard, R. & Maeurer, M.J. Expansion of Tumor-reactive T Cells From Patients With Pancreatic Cancer. Journal of immunotherapy (Hagerstown, Md. : 1997) 39, 81-89 (2016). 167. Demeure, C.E., Wolfers, J., Martin-Garcia, N., Gaulard, P. & Triebel, F. T Lymphocytes infiltrating various tumour types express the MHC class II ligand lymphocyte activation gene-3 (LAG-3): role of LAG-3/MHC class II interactions in cell-cell contacts. Eur J Cancer 37, 1709-1718 (2001). 168. Woo, S.R., Turnis, M.E., Goldberg, M.V., Bankoti, J., Selby, M., Nirschl, C.J., Bettini, M.L., Gravano, D.M., Vogel, P., Liu, C.L., Tangsombatvisit, S., Grosso, J.F., Netto, G., Smeltzer, M.P., Chaux, A., Utz, P.J., Workman, C.J., Pardoll, D.M., Korman, A.J., Drake, C.G. & Vignali, D.A. Immune inhibitory molecules LAG-3 and PD-1 synergistically regulate T-cell function to promote tumoral immune escape. Cancer research 72, 917-927 (2012). 169. Huang, R.Y., Eppolito, C., Lele, S., Shrikant, P., Matsuzaki, J. & Odunsi, K. LAG3 and PD1 co-inhibitory molecules collaborate to limit CD8+ T cell signaling and dampen antitumor immunity in a murine ovarian cancer model. Oncotarget 6, 27359-27377 (2015). 170. Huang, R.Y., Francois, A., McGray, A.R., Miliotto, A. & Odunsi, K. Compensatory upregulation of PD-1, LAG-3, and CTLA-4 limits the efficacy of single-agent checkpoint blockade in metastatic ovarian cancer. Oncoimmunology 6, e1249561 (2017). 171. Brignone, C., Escudier, B., Grygar, C., Marcu, M. & Triebel, F. A phase I pharmacokinetic and biological correlative study of IMP321, a novel MHC class II agonist, in patients with advanced renal cell carcinoma. Clin Cancer Res 15, 6225-6231 (2009). 172. Monney, L., Sabatos, C.A., Gaglia, J.L., Ryu, A., Waldner, H., Chernova, T., Manning, S., Greenfield, E.A., Coyle, A.J., Sobel, R.A., Freeman, G.J. & Kuchroo, V.K. Th1-specific cell surface protein Tim-3 regulates macrophage activation and severity of an autoimmune disease. Nature 415, 536-541 (2002). 173. Hastings, W.D., Anderson, D.E., Kassam, N., Koguchi, K., Greenfield, E.A., Kent, S.C., Zheng, X.X., Strom, T.B., Hafler, D.A. & Kuchroo, V.K. TIM-3 is expressed on activated human CD4+ T cells and regulates Th1 and Th17 cytokines. European journal of immunology 39, 2492-2501 (2009). 174. Gao, X., Zhu, Y., Li, G., Huang, H., Zhang, G., Wang, F., Sun, J., Yang, Q., Zhang, X. & Lu, B. TIM-3 expression characterizes regulatory T cells in tumor tissues and is associated with lung cancer progression. PLoS One 7, e30676 (2012). 175. Yan, J., Zhang, Y., Zhang, J.P., Liang, J., Li, L. & Zheng, L. Tim-3 expression defines regulatory T cells in human tumors. PLoS One 8, e58006 (2013). 176. Gleason, M.K., Lenvik, T.R., McCullar, V., Felices, M., O'Brien, M.S., Cooley, S.A., Verneris, M.R., Cichocki, F., Holman, C.J., Panoskaltsis-Mortari, A., Niki, T., Hirashima, M., Blazar, B.R. & Miller, J.S. Tim-3 is an inducible human natural killer cell receptor that enhances interferon gamma production in response to galectin-9. Blood 119, 3064-3072 (2012). 142  177. Ndhlovu, L.C., Lopez-Verges, S., Barbour, J.D., Jones, R.B., Jha, A.R., Long, B.R., Schoeffler, E.C., Fujita, T., Nixon, D.F. & Lanier, L.L. Tim-3 marks human natural killer cell maturation and suppresses cell-mediated cytotoxicity. Blood 119, 3734-3743 (2012). 178. Anderson, A.C., Anderson, D.E., Bregoli, L., Hastings, W.D., Kassam, N., Lei, C., Chandwaskar, R., Karman, J., Su, E.W., Hirashima, M., Bruce, J.N., Kane, L.P., Kuchroo, V.K. & Hafler, D.A. Promotion of tissue inflammation by the immune receptor Tim-3 expressed on innate immune cells. Science 318, 1141-1143 (2007). 179. Wada, J. & Kanwar, Y.S. Identification and characterization of galectin-9, a novel beta-galactoside-binding mammalian lectin. J Biol Chem 272, 6078-6086 (1997). 180. Zhu, C., Anderson, A.C., Schubart, A., Xiong, H., Imitola, J., Khoury, S.J., Zheng, X.X., Strom, T.B. & Kuchroo, V.K. The Tim-3 ligand galectin-9 negatively regulates T helper type 1 immunity. Nat Immunol 6, 1245-1252 (2005). 181. Sabatos, C.A., Chakravarti, S., Cha, E., Schubart, A., Sanchez-Fueyo, A., Zheng, X.X., Coyle, A.J., Strom, T.B., Freeman, G.J. & Kuchroo, V.K. Interaction of Tim-3 and Tim-3 ligand regulates T helper type 1 responses and induction of peripheral tolerance. Nat Immunol 4, 1102-1110 (2003). 182. Sanchez-Fueyo, A., Tian, J., Picarella, D., Domenig, C., Zheng, X.X., Sabatos, C.A., Manlongat, N., Bender, O., Kamradt, T., Kuchroo, V.K., Gutierrez-Ramos, J.C., Coyle, A.J. & Strom, T.B. Tim-3 inhibits T helper type 1-mediated auto- and alloimmune responses and promotes immunological tolerance. Nat Immunol 4, 1093-1101 (2003). 183. Jones, R.B., Ndhlovu, L.C., Barbour, J.D., Sheth, P.M., Jha, A.R., Long, B.R., Wong, J.C., Satkunarajah, M., Schweneker, M., Chapman, J.M., Gyenes, G., Vali, B., Hyrcza, M.D., Yue, F.Y., Kovacs, C., Sassi, A., Loutfy, M., Halpenny, R., Persad, D., Spotts, G., Hecht, F.M., Chun, T.W., McCune, J.M., Kaul, R., Rini, J.M., Nixon, D.F. & Ostrowski, M.A. Tim-3 expression defines a novel population of dysfunctional T cells with highly elevated frequencies in progressive HIV-1 infection. J Exp Med 205, 2763-2779 (2008). 184. Hafler, D.A. & Kuchroo, V. TIMs: central regulators of immune responses. J Exp Med 205, 2699-2701 (2008). 185. Golden-Mason, L., Palmer, B.E., Kassam, N., Townshend-Bulson, L., Livingston, S., McMahon, B.J., Castelblanco, N., Kuchroo, V., Gretch, D.R. & Rosen, H.R. Negative immune regulator Tim-3 is overexpressed on T cells in hepatitis C virus infection and its blockade rescues dysfunctional CD4+ and CD8+ T cells. Journal of virology 83, 9122-9130 (2009). 186. Takamura, S., Tsuji-Kawahara, S., Yagita, H., Akiba, H., Sakamoto, M., Chikaishi, T., Kato, M. & Miyazawa, M. Premature terminal exhaustion of Friend virus-specific effector CD8+ T cells by rapid induction of multiple inhibitory receptors. J Immunol 184, 4696-4707 (2010). 187. Jin, H.T., Anderson, A.C., Tan, W.G., West, E.E., Ha, S.J., Araki, K., Freeman, G.J., Kuchroo, V.K. & Ahmed, R. Cooperation of Tim-3 and PD-1 in CD8 T-cell exhaustion during chronic viral infection. Proc Natl Acad Sci U S A 107, 14733-14738 (2010). 188. Fourcade, J., Sun, Z., Benallaoua, M., Guillaume, P., Luescher, I.F., Sander, C., Kirkwood, J.M., Kuchroo, V. & Zarour, H.M. Upregulation of Tim-3 and PD-1 expression is associated with tumor antigen-specific CD8+ T cell dysfunction in melanoma patients. J Exp Med 207, 2175-2186 (2010). 143  189. Baitsch, L., Baumgaertner, P., Devevre, E., Raghav, S.K., Legat, A., Barba, L., Wieckowski, S., Bouzourene, H., Deplancke, B., Romero, P., Rufer, N. & Speiser, D.E. Exhaustion of tumor-specific CD8(+) T cells in metastases from melanoma patients. J Clin Invest 121, 2350-2360 (2011). 190. Yang, Z.Z., Grote, D.M., Ziesmer, S.C., Niki, T., Hirashima, M., Novak, A.J., Witzig, T.E. & Ansell, S.M. IL-12 upregulates TIM-3 expression and induces T cell exhaustion in patients with follicular B cell non-Hodgkin lymphoma. J Clin Invest 122, 1271-1282 (2012). 191. Lu, X., Yang, L., Yao, D., Wu, X., Li, J., Liu, X., Deng, L., Huang, C., Wang, Y., Li, D. & Liu, J. Tumor antigen-specific CD8(+) T cells are negatively regulated by PD-1 and Tim-3 in human gastric cancer. Cellular immunology 313, 43-51 (2017). 192. Linedale, R., Schmidt, C., King, B.T., Ganko, A.G., Simpson, F., Panizza, B.J. & Leggatt, G.R. Elevated frequencies of CD8 T cells expressing PD-1, CTLA-4 and Tim-3 within tumour from perineural squamous cell carcinoma patients. PLoS One 12, e0175755 (2017). 193. Ceresoli, G.L. & Mantovani, A. Immune checkpoint inhibitors in malignant pleural mesothelioma. The Lancet. Oncology 18, 559-561 (2017). 194. Shayan, G., Srivastava, R., Li, J., Schmitt, N., Kane, L.P. & Ferris, R.L. Adaptive resistance to anti-PD1 therapy by Tim-3 upregulation is mediated by the PI3K-Akt pathway in head and neck cancer. Oncoimmunology 6, e1261779 (2017). 195. Li, Z., Liu, X., Guo, R. & Wang, P. TIM-3 plays a more important role than PD-1 in the functional impairments of cytotoxic T cells of malignant Schwannomas. Tumour Biol 39, 1010428317698352 (2017). 196. Sakuishi, K., Apetoh, L., Sullivan, J.M., Blazar, B.R., Kuchroo, V.K. & Anderson, A.C. Targeting Tim-3 and PD-1 pathways to reverse T cell exhaustion and restore anti-tumor immunity. J Exp Med 207, 2187-2194 (2010). 197. Zhou, Q., Munger, M.E., Veenstra, R.G., Weigel, B.J., Hirashima, M., Munn, D.H., Murphy, W.J., Azuma, M., Anderson, A.C., Kuchroo, V.K. & Blazar, B.R. Coexpression of Tim-3 and PD-1 identifies a CD8+ T-cell exhaustion phenotype in mice with disseminated acute myelogenous leukemia. Blood 117, 4501-4510 (2011). 198. Gautron, A.S., Dominguez-Villar, M., de Marcken, M. & Hafler, D.A. Enhanced suppressor function of TIM-3+ FoxP3+ regulatory T cells. European journal of immunology 44, 2703-2711 (2014). 199. Gupta, S., Thornley, T.B., Gao, W., Larocca, R., Turka, L.A., Kuchroo, V.K. & Strom, T.B. Allograft rejection is restrained by short-lived TIM-3+PD-1+Foxp3+ Tregs. J Clin Invest 122, 2395-2404 (2012). 200. da Silva, I.P., Gallois, A., Jimenez-Baranda, S., Khan, S., Anderson, A.C., Kuchroo, V.K., Osman, I. & Bhardwaj, N. Reversal of NK-cell exhaustion in advanced melanoma by Tim-3 blockade. Cancer Immunol Res 2, 410-422 (2014). 201. Kang, C.W., Dutta, A., Chang, L.Y., Mahalingam, J., Lin, Y.C., Chiang, J.M., Hsu, C.Y., Huang, C.T., Su, W.T., Chu, Y.Y. & Lin, C.Y. Apoptosis of tumor infiltrating effector TIM-3+CD8+ T cells in colon cancer. Sci Rep 5, 15659 (2015). 202. Ngiow, S.F., von Scheidt, B., Akiba, H., Yagita, H., Teng, M.W. & Smyth, M.J. Anti-TIM3 antibody promotes T cell IFN-gamma-mediated antitumor immunity and suppresses established tumors. Cancer research 71, 3540-3551 (2011). 144  203. Koyama, S., Akbay, E.A., Li, Y.Y., Herter-Sprie, G.S., Buczkowski, K.A., Richards, W.G., Gandhi, L., Redig, A.J., Rodig, S.J., Asahina, H., Jones, R.E., Kulkarni, M.M., Kuraguchi, M., Palakurthi, S., Fecci, P.E., Johnson, B.E., Janne, P.A., Engelman, J.A., Gangadharan, S.P., Costa, D.B., Freeman, G.J., Bueno, R., Hodi, F.S., Dranoff, G., Wong, K.K. & Hammerman, P.S. Adaptive resistance to therapeutic PD-1 blockade is associated with upregulation of alternative immune checkpoints. Nat Commun 7, 10501 (2016). 204. Herbst, R.S., Baas, P., Kim, D.W., Felip, E., Pérez-Gracia, J.L., Han, J.Y., Molina, J., Kim, J.H., Arvis, C.D., Ahn, M.J., Majem, M., Fidler, M.J., de Castro, G., Garrido, M., Lubiniecki, G.M., Shentu, Y., Im, E., Dolled-Filhart, M. & Garon, E.B. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet 387, 1540-1550 (2016). 205. Nanda, R., Chow, L.Q., Dees, E.C., Berger, R., Gupta, S., Geva, R., Pusztai, L., Pathiraja, K., Aktan, G., Cheng, J.D., Karantza, V. & Buisseret, L. Pembrolizumab in Patients With Advanced Triple-Negative Breast Cancer: Phase Ib KEYNOTE-012 Study. J Clin Oncol 34, 2460-2467 (2016). 206. Bortnik, S., Choutka, C., Horlings, H.M., Leung, S., Baker, J.H., Lebovitz, C., Dragowska, W.H., Go, N.E., Bally, M.B., Minchinton, A.I., Gelmon, K.A. & Gorski, S.M. Identification of breast cancer cell subtypes sensitive to ATG4B inhibition. Oncotarget 7, 66970-66988 (2016). 207. Cheang, M.C., Treaba, D.O., Speers, C.H., Olivotto, I.A., Bajdik, C.D., Chia, S.K., Goldstein, L.C., Gelmon, K.A., Huntsman, D., Gilks, C.B., Nielsen, T.O. & Gown, A.M. Immunohistochemical detection using the new rabbit monoclonal antibody SP1 of estrogen receptor in breast cancer is superior to mouse monoclonal antibody 1D5 in predicting survival. J Clin Oncol 24, 5637-5644 (2006). 208. Cheang, M.C., Chia, S.K., Voduc, D., Gao, D., Leung, S., Snider, J., Watson, M., Davies, S., Bernard, P.S., Parker, J.S., Perou, C.M., Ellis, M.J. & Nielsen, T.O. Ki67 index, HER2 status, and prognosis of patients with luminal B breast cancer. J Natl Cancer Inst 101, 736-750 (2009). 209. Altman, D.G., McShane, L.M., Sauerbrei, W. & Taube, S.E. Reporting Recommendations for Tumor Marker Prognostic Studies (REMARK): explanation and elaboration. PLoS Med 9, e1001216 (2012). 210. McDermott, D.F., Sosman, J.A., Sznol, M., Massard, C., Gordon, M.S., Hamid, O., Powderly, J.D., Infante, J.R., Fasso, M., Wang, Y.V., Zou, W., Hegde, P.S., Fine, G.D. & Powles, T. Atezolizumab, an Anti-Programmed Death-Ligand 1 Antibody, in Metastatic Renal Cell Carcinoma: Long-Term Safety, Clinical Activity, and Immune Correlates From a Phase Ia Study. J Clin Oncol 34, 833-842 (2016). 211. Triebel, F., Hacene, K. & Pichon, M.F. A soluble lymphocyte activation gene-3 (sLAG-3) protein as a prognostic factor in human breast cancer expressing estrogen or progesterone receptors. Cancer Lett 235, 147-153 (2006). 212. Chen, D.S. & Mellman, I. Elements of cancer immunity and the cancer-immune set point. Nature 541, 321-330 (2017). 213. Ingold Heppner, B., Untch, M., Denkert, C., Pfitzner, B.M., Lederer, B., Schmitt, W., Eidtmann, H., Fasching, P.A., Tesch, H., Solbach, C., Rezai, M., Zahm, D.M., Holms, F., Glados, M., Krabisch, P., Heck, E., Ober, A., Lorenz, P., Diebold, K., Habeck, J.O. & 145  Loibl, S. Tumor-Infiltrating Lymphocytes: A Predictive and Prognostic Biomarker in Neoadjuvant-Treated HER2-Positive Breast Cancer. Clin Cancer Res 22, 5747-5754 (2016). 214. Hamid, O., Schmidt, H., Nissan, A., Ridolfi, L., Aamdal, S., Hansson, J., Guida, M., Hyams, D.M., Gomez, H., Bastholt, L., Chasalow, S.D. & Berman, D. A prospective phase II trial exploring the association between tumor microenvironment biomarkers and clinical activity of ipilimumab in advanced melanoma. J Transl Med 9, 204 (2011). 215. Tumeh, P.C., Harview, C.L., Yearley, J.H., Shintaku, I.P., Taylor, E.J., Robert, L., Chmielowski, B., Spasic, M., Henry, G., Ciobanu, V., West, A.N., Carmona, M., Kivork, C., Seja, E., Cherry, G., Gutierrez, A.J., Grogan, T.R., Mateus, C., Tomasic, G., Glaspy, J.A., Emerson, R.O., Robins, H., Pierce, R.H., Elashoff, D.A., Robert, C. & Ribas, A. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 515, 568-571 (2014). 216. Schmid P, C.C., Braiteh FS et al. Atezolizumab in metastatic TNBC (mTNBC): Long-term clinical outcomes and biomarker analyses [abstract]. in AACR Annual Meeting Abstract nr 2986 (Washington, DC, 2017). 217. Dirix LY Y, T.I., Nikolinakos P et al. Avelumab (MSB0010718C), an anti-PD-L1 antibody, in patients with locally advanced or metastatic breast cancer: A phase Ib JAVELIN solid tumor trial[abstract]. in San Antonio Breast Cancer Symposium Abstract nr S1-04 (San Antonio, Texas, USA, 2015). 218. Salama, A.K. & Moschos, S.J. Next Steps in Immuno-Oncology: Enhancing Antitumor Effects Through Appropriate Patient Selection and Rationally Designed Combination Strategies. Ann Oncol (2016). 219. Krishnamurti, U., Wetherilt, C.S., Yang, J., Peng, L. & Li, X. Tumor-infiltrating lymphocytes are significantly associated with better overall survival and disease-free survival in triple-negative but not estrogen receptor-positive breast cancers. Human pathology 64, 7-12 (2017). 220. Heindl, A., Sestak, I., Naidoo, K., Cuzick, J., Dowsett, M. & Yuan, Y. Relevance of Spatial Heterogeneity of Immune Infiltration for Predicting Risk of Recurrence After Endocrine Therapy of ER+ Breast Cancer. J Natl Cancer Inst 110(2018). 221. Herbst, R.S., Baas, P., Kim, D.W., Felip, E., Perez-Gracia, J.L., Han, J.Y., Molina, J., Kim, J.H., Arvis, C.D., Ahn, M.J., Majem, M., Fidler, M.J., de Castro, G., Jr., Garrido, M., Lubiniecki, G.M., Shentu, Y., Im, E., Dolled-Filhart, M. & Garon, E.B. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet 387, 1540-1550 (2016). 222. Dirix, L.Y., Takacs, I., Jerusalem, G., Nikolinakos, P., Arkenau, H.T., Forero-Torres, A., Boccia, R., Lippman, M.E., Somer, R., Smakal, M., Emens, L.A., Hrinczenko, B., Edenfield, W., Gurtler, J., von Heydebreck, A., Grote, H.J., Chin, K. & Hamilton, E.P. Avelumab, an anti-PD-L1 antibody, in patients with locally advanced or metastatic breast cancer: a phase 1b JAVELIN Solid Tumor study. Breast Cancer Res Treat 167, 671-686 (2018). 223. McArthur, H.L., Diab, A., Page, D.B., Yuan, J., Solomon, S.B., Sacchini, V., Comstock, C., Durack, J.C., Maybody, M., Sung, J., Ginsberg, A., Wong, P., Barlas, A., Dong, Z., Zhao, C., Blum, B., Patil, S., Neville, D., Comen, E.A., Morris, E.A., Kotin, A., Brogi, 146  E., Wen, Y.H., Morrow, M., Lacouture, M.E., Sharma, P., Allison, J.P., Hudis, C.A., Wolchok, J.D. & Norton, L. A Pilot Study of Preoperative Single-Dose Ipilimumab and/or Cryoablation in Women with Early-Stage Breast Cancer with Comprehensive Immune Profiling. Clin Cancer Res 22, 5729-5737 (2016). 224. Emens, L.A. Breast Cancer Immunotherapy: Facts and Hopes. Clin Cancer Res 24, 511-520 (2018). 225. Kwa, M.J. & Adams, S. Checkpoint inhibitors in triple-negative breast cancer (TNBC): Where to go from here. Cancer (2018). 226. Bellmunt, J., de Wit, R., Vaughn, D.J., Fradet, Y., Lee, J.L., Fong, L., Vogelzang, N.J., Climent, M.A., Petrylak, D.P., Choueiri, T.K., Necchi, A., Gerritsen, W., Gurney, H., Quinn, D.I., Culine, S., Sternberg, C.N., Mai, Y., Poehlein, C.H., Perini, R.F., Bajorin, D.F. & Investigators, K.-. Pembrolizumab as Second-Line Therapy for Advanced Urothelial Carcinoma. N Engl J Med 376, 1015-1026 (2017). 227. Balar, A.V., Galsky, M.D., Rosenberg, J.E., Powles, T., Petrylak, D.P., Bellmunt, J., Loriot, Y., Necchi, A., Hoffman-Censits, J., Perez-Gracia, J.L., Dawson, N.A., van der Heijden, M.S., Dreicer, R., Srinivas, S., Retz, M.M., Joseph, R.W., Drakaki, A., Vaishampayan, U.N., Sridhar, S.S., Quinn, D.I., Duran, I., Shaffer, D.R., Eigl, B.J., Grivas, P.D., Yu, E.Y., Li, S., Kadel, E.E., 3rd, Boyd, Z., Bourgon, R., Hegde, P.S., Mariathasan, S., Thastrom, A., Abidoye, O.O., Fine, G.D., Bajorin, D.F. & Group, I.M.S. Atezolizumab as first-line treatment in cisplatin-ineligible patients with locally advanced and metastatic urothelial carcinoma: a single-arm, multicentre, phase 2 trial. Lancet 389, 67-76 (2017). 228. Balar, A.V., Castellano, D., O'Donnell, P.H., Grivas, P., Vuky, J., Powles, T., Plimack, E.R., Hahn, N.M., de Wit, R., Pang, L., Savage, M.J., Perini, R.F., Keefe, S.M., Bajorin, D. & Bellmunt, J. First-line pembrolizumab in cisplatin-ineligible patients with locally advanced and unresectable or metastatic urothelial cancer (KEYNOTE-052): a multicentre, single-arm, phase 2 study. The Lancet. Oncology 18, 1483-1492 (2017). 229. Rittmeyer, A., Barlesi, F., Waterkamp, D., Park, K., Ciardiello, F., von Pawel, J., Gadgeel, S.M., Hida, T., Kowalski, D.M., Dols, M.C., Cortinovis, D.L., Leach, J., Polikoff, J., Barrios, C., Kabbinavar, F., Frontera, O.A., De Marinis, F., Turna, H., Lee, J.S., Ballinger, M., Kowanetz, M., He, P., Chen, D.S., Sandler, A., Gandara, D.R. & Group, O.A.K.S. Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial. Lancet 389, 255-265 (2017). 230. Antonia, S.J., Villegas, A., Daniel, D., Vicente, D., Murakami, S., Hui, R., Yokoi, T., Chiappori, A., Lee, K.H., de Wit, M., Cho, B.C., Bourhaba, M., Quantin, X., Tokito, T., Mekhail, T., Planchard, D., Kim, Y.C., Karapetis, C.S., Hiret, S., Ostoros, G., Kubota, K., Gray, J.E., Paz-Ares, L., de Castro Carpeno, J., Wadsworth, C., Melillo, G., Jiang, H., Huang, Y., Dennis, P.A., Ozguroglu, M. & Investigators, P. Durvalumab after Chemoradiotherapy in Stage III Non-Small-Cell Lung Cancer. N Engl J Med 377, 1919-1929 (2017). 231. de Mingo Pulido, A., Gardner, A., Hiebler, S., Soliman, H., Rugo, H.S., Krummel, M.F., Coussens, L.M. & Ruffell, B. TIM-3 Regulates CD103(+) Dendritic Cell Function and Response to Chemotherapy in Breast Cancer. Cancer Cell 33, 60-74 e66 (2018). 147  232. Yan, W., Liu, X., Ma, H., Zhang, H., Song, X., Gao, L., Liang, X. & Ma, C. Tim-3 fosters HCC development by enhancing TGF-beta-mediated alternative activation of macrophages. Gut 64, 1593-1604 (2015). 233. Sabatos-Peyton, C.A., Nevin, J., Brock, A., Venable, J.D., Tan, D.J., Kassam, N., Xu, F., Taraszka, J., Wesemann, L., Pertel, T., Acharya, N., Klapholz, M., Etminan, Y., Jiang, X., Huang, Y.H., Blumberg, R.S., Kuchroo, V.K. & Anderson, A.C. Blockade of Tim-3 binding to phosphatidylserine and CEACAM1 is a shared feature of anti-Tim-3 antibodies that have functional efficacy. Oncoimmunology 7, e1385690 (2018). 234. Das, M., Zhu, C. & Kuchroo, V.K. Tim-3 and its role in regulating anti-tumor immunity. Immunological reviews 276, 97-111 (2017). 235. Granier, C., Dariane, C., Combe, P., Verkarre, V., Urien, S., Badoual, C., Roussel, H., Mandavit, M., Ravel, P., Sibony, M., Biard, L., Radulescu, C., Vinatier, E., Benhamouda, N., Peyromaure, M., Oudard, S., Mejean, A., Timsit, M.O., Gey, A. & Tartour, E. Tim-3 Expression on Tumor-Infiltrating PD-1(+)CD8(+) T Cells Correlates with Poor Clinical Outcome in Renal Cell Carcinoma. Cancer Res 77, 1075-1082 (2017). 236. Li, H., Wu, K., Tao, K., Chen, L., Zheng, Q., Lu, X., Liu, J., Shi, L., Liu, C., Wang, G. & Zou, W. Tim-3/galectin-9 signaling pathway mediates T-cell dysfunction and predicts poor prognosis in patients with hepatitis B virus-associated hepatocellular carcinoma. Hepatology 56, 1342-1351 (2012). 237. Japp, A.S., Kursunel, M.A., Meier, S., Malzer, J.N., Li, X., Rahman, N.A., Jekabsons, W., Krause, H., Magheli, A., Klopf, C., Thiel, A. & Frentsch, M. Dysfunction of PSA-specific CD8+ T cells in prostate cancer patients correlates with CD38 and Tim-3 expression. Cancer immunology, immunotherapy : CII 64, 1487-1494 (2015). 238. Liu, Z., Meng, Q., Bartek, J., Jr., Poiret, T., Persson, O., Rane, L., Rangelova, E., Illies, C., Peredo, I.H., Luo, X., Rao, M.V., Robertson, R.A., Dodoo, E. & Maeurer, M. Tumor-infiltrating lymphocytes (TILs) from patients with glioma. Oncoimmunology 6, e1252894 (2017). 239. Solinas, C., Garaud, S., De Silva, P., Boisson, A., Van den Eynden, G., de Wind, A., Risso, P., Rodrigues Vitoria, J., Richard, F., Migliori, E., Noel, G., Duvillier, H., Craciun, L., Veys, I., Awada, A., Detours, V., Larsimont, D., Piccart-Gebhart, M. & Willard-Gallo, K. Immune Checkpoint Molecules on Tumor-Infiltrating Lymphocytes and Their Association with Tertiary Lymphoid Structures in Human Breast Cancer. Front Immunol 8, 1412 (2017). 240. Zhang, H., Xiang, R., Wu, B., Li, J. & Luo, G. T-cell immunoglobulin mucin-3 expression in invasive ductal breast carcinoma: Clinicopathological correlations and association with tumor infiltration by cytotoxic lymphocytes. Mol Clin Oncol 7, 557-563 (2017). 241. Cheang, M.C., Voduc, D., Bajdik, C., Leung, S., McKinney, S., Chia, S.K., Perou, C.M. & Nielsen, T.O. Basal-like breast cancer defined by five biomarkers has superior prognostic value than triple-negative phenotype. Clin Cancer Res 14, 1368-1376 (2008). 242. Burugu, S., Gao, D., Leung, S., Chia, S.K. & Nielsen, T.O. LAG-3+ tumor infiltrating lymphocytes in breast cancer: clinical correlates and association with PD-1/PD-L1+ tumors. Ann Oncol 28, 2977-2984 (2017). 243. Chen, T.C., Chen, C.H., Wang, C.P., Lin, P.H., Yang, T.L., Lou, P.J., Ko, J.Y., Wu, C.T. & Chang, Y.L. The immunologic advantage of recurrent nasopharyngeal carcinoma from 148  the viewpoint of Galectin-9/Tim-3-related changes in the tumour microenvironment. Sci Rep 7, 10349 (2017). 244. Liu, J.F., Ma, S.R., Mao, L., Bu, L.L., Yu, G.T., Li, Y.C., Huang, C.F., Deng, W.W., Kulkarni, A.B., Zhang, W.F. & Sun, Z.J. T-cell immunoglobulin mucin 3 blockade drives an antitumor immune response in head and neck cancer. Molecular oncology 11, 235-247 (2017). 245. Gao, J., Ward, J.F., Pettaway, C.A., Shi, L.Z., Subudhi, S.K., Vence, L.M., Zhao, H., Chen, J., Chen, H., Efstathiou, E., Troncoso, P., Allison, J.P., Logothetis, C.J., Wistuba, II, Sepulveda, M.A., Sun, J., Wargo, J., Blando, J. & Sharma, P. VISTA is an inhibitory immune checkpoint that is increased after ipilimumab therapy in patients with prostate cancer. Nature medicine 23, 551-555 (2017). 246. Zhang, Y., Cai, P., Liang, T., Wang, L. & Hu, L. TIM-3 is a potential prognostic marker for patients with solid tumors: A systematic review and meta-analysis. Oncotarget 8, 31705-31713 (2017). 247. Le, D.T., Uram, J.N., Wang, H., Bartlett, B.R., Kemberling, H., Eyring, A.D., Skora, A.D., Luber, B.S., Azad, N.S., Laheru, D., Biedrzycki, B., Donehower, R.C., Zaheer, A., Fisher, G.A., Crocenzi, T.S., Lee, J.J., Duffy, S.M., Goldberg, R.M., de la Chapelle, A., Koshiji, M., Bhaijee, F., Huebner, T., Hruban, R.H., Wood, L.D., Cuka, N., Pardoll, D.M., Papadopoulos, N., Kinzler, K.W., Zhou, S., Cornish, T.C., Taube, J.M., Anders, R.A., Eshleman, J.R., Vogelstein, B. & Diaz, L.A., Jr. PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. N Engl J Med 372, 2509-2520 (2015). 248. Rizvi, N.A., Hellmann, M.D., Snyder, A., Kvistborg, P., Makarov, V., Havel, J.J., Lee, W., Yuan, J., Wong, P., Ho, T.S., Miller, M.L., Rekhtman, N., Moreira, A.L., Ibrahim, F., Bruggeman, C., Gasmi, B., Zappasodi, R., Maeda, Y., Sander, C., Garon, E.B., Merghoub, T., Wolchok, J.D., Schumacher, T.N. & Chan, T.A. Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. Science 348, 124-128 (2015). 249. Davoli, T., Uno, H., Wooten, E.C. & Elledge, S.J. Tumor aneuploidy correlates with markers of immune evasion and with reduced response to immunotherapy. Science 355(2017). 250. Le, D.T., Durham, J.N., Smith, K.N., Wang, H., Bartlett, B.R., Aulakh, L.K., Lu, S., Kemberling, H., Wilt, C., Luber, B.S., Wong, F., Azad, N.S., Rucki, A.A., Laheru, D., Donehower, R., Zaheer, A., Fisher, G.A., Crocenzi, T.S., Lee, J.J., Greten, T.F., Duffy, A.G., Ciombor, K.K., Eyring, A.D., Lam, B.H., Joe, A., Kang, S.P., Holdhoff, M., Danilova, L., Cope, L., Meyer, C., Zhou, S., Goldberg, R.M., Armstrong, D.K., Bever, K.M., Fader, A.N., Taube, J., Housseau, F., Spetzler, D., Xiao, N., Pardoll, D.M., Papadopoulos, N., Kinzler, K.W., Eshleman, J.R., Vogelstein, B., Anders, R.A. & Diaz, L.A., Jr. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science 357, 409-413 (2017). 251. Herbst, R.S., Soria, J.C., Kowanetz, M., Fine, G.D., Hamid, O., Gordon, M.S., Sosman, J.A., McDermott, D.F., Powderly, J.D., Gettinger, S.N., Kohrt, H.E., Horn, L., Lawrence, D.P., Rost, S., Leabman, M., Xiao, Y., Mokatrin, A., Koeppen, H., Hegde, P.S., Mellman, I., Chen, D.S. & Hodi, F.S. Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature 515, 563-567 (2014). 149  252. Allard, B., Aspeslagh, S., Garaud, S., Dupont, F.A., Solinas, C., Kok, M., Routy, B., Sotiriou, C., Stagg, J. & Buisseret, L. Immuno-oncology-101: overview of major concepts and translational perspectives. Seminars in cancer biology 52, 1-11 (2018). 253. Halse, H., Colebatch, A.J., Petrone, P., Henderson, M.A., Mills, J.K., Snow, H., Westwood, J.A., Sandhu, S., Raleigh, J.M., Behren, A., Cebon, J., Darcy, P.K., Kershaw, M.H., McArthur, G.A., Gyorki, D.E. & Neeson, P.J. Multiplex immunohistochemistry accurately defines the immune context of metastatic melanoma. Sci Rep 8, 11158 (2018). 254. Keren, L., Bosse, M., Marquez, D., Angoshtari, R., Jain, S., Varma, S., Yang, S.R., Kurian, A., Van Valen, D., West, R., Bendall, S.C. & Angelo, M. A Structured Tumor-Immune Microenvironment in Triple Negative Breast Cancer Revealed by Multiplexed Ion Beam Imaging. Cell 174, 1373-1387 e1319 (2018). 255. Asleh-Aburaya, K., Sheffield, B.S., Kos, Z., Won, J.R., Wang, X.Q., Gao, D., Wolber, R., Gilks, C.B., Bernard, P.S., Chia, S.K. & Nielsen, T.O. Basal biomarkers nestin and INPP4b identify intrinsic subtypes accurately in breast cancers that are weakly positive for oestrogen receptor. Histopathology 70, 185-194 (2017). 256. Amaria, R.N., Reddy, S.M., Tawbi, H.A., Davies, M.A., Ross, M.I., Glitza, I.C., Cormier, J.N., Lewis, C., Hwu, W.J., Hanna, E., Diab, A., Wong, M.K., Royal, R., Gross, N., Weber, R., Lai, S.Y., Ehlers, R., Blando, J., Milton, D.R., Woodman, S., Kageyama, R., Wells, D.K., Hwu, P., Patel, S.P., Lucci, A., Hessel, A., Lee, J.E., Gershenwald, J., Simpson, L., Burton, E.M., Posada, L., Haydu, L., Wang, L., Zhang, S., Lazar, A.J., Hudgens, C.W., Gopalakrishnan, V., Reuben, A., Andrews, M.C., Spencer, C.N., Prieto, V., Sharma, P., Allison, J., Tetzlaff, M.T. & Wargo, J.A. Neoadjuvant immune checkpoint blockade in high-risk resectable melanoma. Nat Med 24, 1649-1654 (2018). 257. Blank, C.U., Rozeman, E.A., Fanchi, L.F., Sikorska, K., van de Wiel, B., Kvistborg, P., Krijgsman, O., van den Braber, M., Philips, D., Broeks, A., van Thienen, J.V., Mallo, H.A., Adriaansz, S., Ter Meulen, S., Pronk, L.M., Grijpink-Ongering, L.G., Bruining, A., Gittelman, R.M., Warren, S., van Tinteren, H., Peeper, D.S., Haanen, J., van Akkooi, A.C.J. & Schumacher, T.N. Neoadjuvant versus adjuvant ipilimumab plus nivolumab in macroscopic stage III melanoma. Nat Med 24, 1655-1661 (2018). 258. Dieci, M.V., Radosevic-Robin, N., Fineberg, S., van den Eynden, G., Ternes, N., Penault-Llorca, F., Pruneri, G., D'Alfonso, T.M., Demaria, S., Castaneda, C., Sanchez, J., Badve, S., Michiels, S., Bossuyt, V., Rojo, F., Singh, B., Nielsen, T., Viale, G., Kim, S.R., Hewitt, S., Wienert, S., Loibl, S., Rimm, D., Symmans, F., Denkert, C., Adams, S., Loi, S., Salgado, R. & International Immuno-Oncology Biomarker Working Group on Breast, C. Update on tumor-infiltrating lymphocytes (TILs) in breast cancer, including recommendations to assess TILs in residual disease after neoadjuvant therapy and in carcinoma in situ: A report of the International Immuno-Oncology Biomarker Working Group on Breast Cancer. Seminars in cancer biology 52, 16-25 (2018). 259. Picarda, E., Ohaegbulam, K.C. & Zang, X. Molecular Pathways: Targeting B7-H3 (CD276) for Human Cancer Immunotherapy. Clin Cancer Res 22, 3425-3431 (2016). 260. Parra, E.R. Novel Platforms of Multiplexed Immunofluorescence for Study of Paraffin Tumor Tissues. Journal of Cancer treatment and diagnosis 2, 43-53 (2018). 261. Parra, E.R., Uraoka, N., Jiang, M., Cook, P., Gibbons, D., Forget, M.A., Bernatchez, C., Haymaker, C., Wistuba, II & Rodriguez-Canales, J. Validation of multiplex 150  immunofluorescence panels using multispectral microscopy for immune-profiling of formalin-fixed and paraffin-embedded human tumor tissues. Sci Rep 7, 13380 (2017). 262. Gorris, M.A.J., Halilovic, A., Rabold, K., van Duffelen, A., Wickramasinghe, I.N., Verweij, D., Wortel, I.M.N., Textor, J.C., de Vries, I.J.M. & Figdor, C.G. Eight-Color Multiplex Immunohistochemistry for Simultaneous Detection of Multiple Immune Checkpoint Molecules within the Tumor Microenvironment. J Immunol 200, 347-354 (2018). 263. Stack, E.C., Foukas, P.G. & Lee, P.P. Multiplexed tissue biomarker imaging. J Immunother Cancer 4, 9 (2016). 264. Decalf, J., Albert, M.L. & Ziai, J. New tools for pathology: a user's review of a highly multiplexed method for in situ analysis of protein and RNA expression in tissue. The Journal of pathology (2018). 265. Ayers, M., Lunceford, J., Nebozhyn, M., Murphy, E., Loboda, A., Kaufman, D.R., Albright, A., Cheng, J.D., Kang, S.P., Shankaran, V., Piha-Paul, S.A., Yearley, J., Seiwert, T.Y., Ribas, A. & McClanahan, T.K. IFN-gamma-related mRNA profile predicts clinical response to PD-1 blockade. J Clin Invest 127, 2930-2940 (2017). 266. Gibney, G.T., Weiner, L.M. & Atkins, M.B. Predictive biomarkers for checkpoint inhibitor-based immunotherapy. The Lancet. Oncology 17, e542-e551 (2016). 267. Spranger, S., Sivan, A., Corrales, L. & Gajewski, T.F. Tumor and Host Factors Controlling Antitumor Immunity and Efficacy of Cancer Immunotherapy. Advances in immunology 130, 75-93 (2016). 268. Teng, M.W., Ngiow, S.F., Ribas, A. & Smyth, M.J. Classifying Cancers Based on T-cell Infiltration and PD-L1. Cancer Res 75, 2139-2145 (2015). 269. Bruno, T.C., Ebner, P.J., Moore, B.L., Squalls, O.G., Waugh, K.A., Eruslanov, E.B., Singhal, S., Mitchell, J.D., Franklin, W.A., Merrick, D.T., McCarter, M.D., Palmer, B.E., Kern, J.A. & Slansky, J.E. Antigen-Presenting Intratumoral B Cells Affect CD4(+) TIL Phenotypes in Non-Small Cell Lung Cancer Patients. Cancer Immunol Res 5, 898-907 (2017). 270. Coulie, P.G., Van den Eynde, B.J., van der Bruggen, P. & Boon, T. Tumour antigens recognized by T lymphocytes: at the core of cancer immunotherapy. Nat Rev Cancer 14, 135-146 (2014). 271. Cristescu, R., Mogg, R., Ayers, M., Albright, A., Murphy, E., Yearley, J., Sher, X., Liu, X.Q., Lu, H., Nebozhyn, M., Zhang, C., Lunceford, J.K., Joe, A., Cheng, J., Webber, A.L., Ibrahim, N., Plimack, E.R., Ott, P.A., Seiwert, T.Y., Ribas, A., McClanahan, T.K., Tomassini, J.E., Loboda, A. & Kaufman, D. Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy. Science 362(2018). 272. Galluzzi, L., Buque, A., Kepp, O., Zitvogel, L. & Kroemer, G. Immunological Effects of Conventional Chemotherapy and Targeted Anticancer Agents. Cancer Cell 28, 690-714 (2015). 273. Alizadeh, D., Trad, M., Hanke, N.T., Larmonier, C.B., Janikashvili, N., Bonnotte, B., Katsanis, E. & Larmonier, N. Doxorubicin eliminates myeloid-derived suppressor cells and enhances the efficacy of adoptive T-cell transfer in breast cancer. Cancer Res 74, 104-118 (2014). 274. Levine, M.N., Pritchard, K.I., Bramwell, V.H., Shepherd, L.E., Tu, D., Paul, N. & National Cancer Institute of Canada Clinical Trials, G. Randomized trial comparing 151  cyclophosphamide, epirubicin, and fluorouracil with cyclophosphamide, methotrexate, and fluorouracil in premenopausal women with node-positive breast cancer: update of National Cancer Institute of Canada Clinical Trials Group Trial MA5. J Clin Oncol 23, 5166-5170 (2005). 275. Savas, P., Virassamy, B., Ye, C., Salim, A., Mintoff, C.P., Caramia, F., Salgado, R., Byrne, D.J., Teo, Z.L., Dushyanthen, S., Byrne, A., Wein, L., Luen, S.J., Poliness, C., Nightingale, S.S., Skandarajah, A.S., Gyorki, D.E., Thornton, C.M., Beavis, P.A., Fox, S.B., Darcy, P.K., Speed, T.P., Mackay, L.K., Neeson, P.J. & Loi, S. Single-cell profiling of breast cancer T cells reveals a tissue-resident memory subset associated with improved prognosis. Nat Med 24, 986-993 (2018). 276. Andrews, L.P., Marciscano, A.E., Drake, C.G. & Vignali, D.A. LAG3 (CD223) as a cancer immunotherapy target. Immunological reviews 276, 80-96 (2017). 277. Duhoux, F.P., Jager, A., Dirix, L., Huizing, M.T., Jerusalem, G.H.M., Vuylsteke, P., Cuypere, E.D., Breiner, D., Mueller, C., Brignone, C. & Triebel, F. Combination of paclitaxel and LAG3-Ig (IMP321), a novel MHC class II agonist, as a first-line chemoimmunotherapy in patients with metastatic breast carcinoma (MBC): Interim results from the run-in phase of a placebo controlled randomized phase II. Journal of Clinical Oncology 35, 1062-1062 (2017). 278. Hong, D.S., Schoffski, P., Calvo, A., Sarantopoulos, J., Olza, M.O.D., Carvajal, R.D., Prawira, A., Kyi, C., Esaki, T., Akerley, W.L., Braud, F.G.D., Hui, R., Zhang, T., Soo, R.A., Maur, M., Weickhardt, A.J., Chowdhury, N.R., Sabatos-Peyton, C., Kwak, E.L. & Tan, D.S.-W. Phase I/II study of LAG525 ± spartalizumab (PDR001) in patients (pts) with advanced malignancies. Journal of Clinical Oncology 36, 3012-3012 (2018). 279. Gayden, T., Sepulveda, F.E., Khuong-Quang, D.A., Pratt, J., Valera, E.T., Garrigue, A., Kelso, S., Sicheri, F., Mikael, L.G., Hamel, N., Bajic, A., Dali, R., Deshmukh, S., Dervovic, D., Schramek, D., Guerin, F., Taipale, M., Nikbakht, H., Majewski, J., Moshous, D., Charlebois, J., Abish, S., Bole-Feysot, C., Nitschke, P., Bader-Meunier, B., Mitchell, D., Thieblemont, C., Battistella, M., Gravel, S., Nguyen, V.H., Conyers, R., Diana, J.S., McCormack, C., Prince, H.M., Besnard, M., Blanche, S., Ekert, P.G., Fraitag, S., Foulkes, W.D., Fischer, A., Neven, B., Michonneau, D., de Saint Basile, G. & Jabado, N. Germline HAVCR2 mutations altering TIM-3 characterize subcutaneous panniculitis-like T cell lymphomas with hemophagocytic lymphohistiocytic syndrome. Nature genetics 50, 1650-1657 (2018). 280. Pernas, S., Tolaney, S.M., Winer, E.P. & Goel, S. CDK4/6 inhibition in breast cancer: current practice and future directions. Therapeutic advances in medical oncology 10, 1758835918786451 (2018). 281. Goel, S., DeCristo, M.J., Watt, A.C., BrinJones, H., Sceneay, J., Li, B.B., Khan, N., Ubellacker, J.M., Xie, S., Metzger-Filho, O., Hoog, J., Ellis, M.J., Ma, C.X., Ramm, S., Krop, I.E., Winer, E.P., Roberts, T.M., Kim, H.J., McAllister, S.S. & Zhao, J.J. CDK4/6 inhibition triggers anti-tumour immunity. Nature 548, 471-475 (2017). 282. Roulois, D., Yau, H.L. & De Carvalho, D.D. Pharmacological DNA demethylation: Implications for cancer immunotherapy. Oncoimmunology 5, e1090077 (2016).   152  Appendix A Supplemental information on LAG-3 assay Antibody: LAG-3  Clone: 17B4  Reactivity: Human Sensitivity and Specificity LAG-3 mouse clone 17B4 (Abcam) recognizes amino acids 70-99 of human LAG-3 protein. Clone 17B4 was tested by Western blot by the supplier using recombinant human LAG-3 protein. Other tested applications of LAG-3 clone 17B4, including flow cytometry and immunohistochemistry on formalin-fixed paraffin-embedded tissues, were reported in previous publications [PMID: 21441454; 25358689, 27301722, and 27912781]. Full protocol on Ventana LAG-3 immunostains on tissue microarrays were performed on the Ventana Discovery Ultra (Ventana Medical Systems, Tucson, AZ, USA) semi-automated immunostainer. Slides underwent antigen retrieval with Standard Cell Conditioning 1 (Ventana Medical Systems) followed by 120 minutes of primary antibody incubation with no heat and detected using chromoMap DAB Detection Kit (Ventana Medical Systems). LAG-3 antibody (clone 17B4, Abcam, ab40466) was applied at dilution of 1:100. Membranous staining on activated lymphocytes localized in normal tonsil tissue served as a positive control. Analytical validity  To assess the reproducibility of LAG-3 scoring, a second observer independently scored 145 cases from the training set (duplicate 0.6 mm diameter tissue microarray cores, stained by immunohistochemistry per the above protocol). As in Chapter 2, IHC-positive lymphocytes 153  were counted, and cases were categorized as positive if any (>0) were identified. There was a substantial inter-observer agreement between LAG-3 dichotomized scores for both the primary biomarker assessment in the Chapter 2 intraepithelial tumor infiltrating lymphocytes = iTILs (Cohen’s kappa = 0.69 ((95% CI 0.49-0.84) and for the alternative assessment of stromal tumor infiltrating lymphocytes, sTILs= 0.65 (0.46-0.81).                154  Appendix B  Pre-specified statistical plan for Chapter 2 presented at the BC Cancer Agency Breast Outcome Unit List of hypotheses generated on the half of the validation cohort (n= 2,003) and to be tested on the other half (n=1,989): 1) Breast tumors with LAG3+TILs are associated with improved relapse-free survival among ER–.      2) Breast tumors with PD1+TILs are associated with shorter relapse-free survival among ER+ patients. 3) Breast tumors with LAG3+TILs and/or PD-1+TILs or PD-L1 are associated with shorter disease-specific survival in cases that lack CD8+ iTILs and improved outcome in cases with CD8+iTILs among ER– patients.  155  Appendix C   Pre-specified statistical plan for Chapter 3 presented at the BC Cancer Agency Breast Outcome Unit List of hypotheses generated on the half of the validation cohort (n= 2,003) and to be tested on the other half (n=1,989): 1) ER- breast tumors (specifically the basal-like subtype) with H&E sTILs (≥10%) are significantly associated with improved relapse-free survival.  2) Breast tumors with H&E sTILs (≥10%) or with TIM-3+iTILs (≥1) are highly associated with the presence of LAG-3+iTILs, PD-1+iTILs and PD-L1+ tumors.   3) The improved survival of basal-like breast cancer patients with H&E sTILs (≥10%) is dependent on co-infiltration with FOXP3+iTILs. 156  Appendix D  Appendix Table D.1 DSP technical control-normalized counts by cohort  Cohort A     CaseID Beta-Catenin CD8A B7-H3 CD4 FoxP3 MmAb IgG2a CD68 PTEN Rabbit IgG CD14 S6 GZMB Ki67 (8D5) B2M Histone H3 AKT CD3 VISTA PD1 pSTAT3 CD44 STAT3  CD56 PD-L1 CD45 P-AKT CD19 CD20 PanCK CD45RO Bcl-22 14379.170 87.840 13647.300 571.215 63.265 19.390 782.810 158.265 54.370 468.475 4415.310 1667.045 78.380 2716.695 1390.590 3954.365 109.250 157.125 76.145 362.715 12352.625 6165.880 94.580 384.080 252.560 213.255 62.820 77.470 36080.790 114.505 5457.0803 24076.700 182.310 14242.655 1279.680 74.030 23.600 2192.345 83.115 74.165 585.935 8084.240 3753.260 1928.570 7068.230 2324.620 2402.075 663.795 254.305 122.335 355.935 11352.590 3195.420 1593.510 579.365 503.135 378.470 181.665 142.700 68023.420 336.825 483.2054 5238.365 1962.370 3147.055 3012.130 108.550 58.495 1147.160 75.235 116.690 741.050 6781.670 969.270 30.780 8253.065 1427.140 3043.625 2007.125 290.735 411.170 414.085 3965.615 2950.920 185.655 835.030 13206.060 192.185 661.760 36642.470 26044.265 710.005 2955.7005 25301.750 102.280 16300.010 1079.755 91.820 32.305 3941.785 123.015 109.535 1857.195 5344.670 1605.095 388.090 4556.720 963.210 2987.030 322.075 938.365 76.730 283.890 5391.675 3198.610 3640.410 763.800 286.850 303.785 168.080 118.685 14633.850 211.000 206.3456 8511.985 218.940 3937.415 734.710 85.690 35.170 1038.130 50.605 81.070 809.535 9027.705 1246.020 1479.035 17424.240 2305.240 4037.110 302.460 355.900 123.335 327.060 8981.845 3855.420 227.890 566.615 301.940 889.415 87.830 246.510 15288.340 323.335 1102.0957 5683.360 117.880 7631.610 964.200 42.540 11.790 5452.740 112.960 38.290 1156.420 2813.100 1312.220 121.050 1947.820 367.070 1609.310 980.720 565.290 118.160 224.170 12408.020 1580.260 8313.220 320.030 1487.890 148.860 157.330 102.110 2538.380 579.130 591.4608 28408.005 1153.065 5491.535 3433.485 109.450 22.440 2300.100 150.010 75.110 797.450 16430.360 3003.890 527.725 6988.840 2217.800 4441.760 6358.415 674.505 322.915 398.495 5059.075 12362.890 140.730 473.275 7136.085 368.775 248.385 14144.690 10621.425 1577.345 1238.7809 8568.230 245.660 7151.940 722.445 75.430 22.505 1405.845 161.005 100.900 374.380 13431.540 1939.925 575.760 7039.690 2173.540 5963.685 327.175 199.230 68.275 229.855 1851.445 3903.315 59.230 339.690 180.470 308.830 131.215 151.380 41628.800 182.430 9901.32010 3187.950 3928.470 2414.235 10182.060 111.465 17.645 1541.195 123.115 66.460 906.355 5471.615 1042.330 158.690 8349.335 2251.190 2684.895 15672.900 494.010 1058.915 816.180 7429.320 8351.110 278.585 464.585 16720.065 692.955 439.495 30143.300 6709.235 5503.470 2235.99011 18591.585 807.035 3028.345 1736.500 155.495 49.715 4007.385 71.520 222.280 3668.625 6414.330 4173.500 137.025 22351.650 1098.565 1443.645 3080.625 965.410 309.305 692.745 23913.120 3079.210 91.775 1459.040 3535.045 428.075 709.955 251.555 39732.980 1071.410 265.66012 20683.430 298.280 7342.895 718.895 108.575 36.120 2132.130 583.720 101.225 904.070 13650.555 3857.140 683.990 9175.730 4570.535 10107.985 611.005 392.725 150.775 449.505 2892.490 6366.580 177.335 670.045 598.605 491.865 317.965 870.225 149468.975 322.360 13085.04013 1273.050 25.110 2620.100 280.220 51.920 8.050 968.380 12.110 22.680 153.480 475.210 151.040 6.430 2822.400 47.050 168.920 62.490 55.170 60.860 53.550 1202.360 250.160 38.920 122.610 121.790 38.920 6.430 51.920 406.970 81.980 29.99014 14388.940 462.565 4560.970 1189.595 62.730 13.605 869.040 190.150 53.945 693.945 4635.285 1236.845 344.605 1776.125 1541.645 2786.435 1524.565 273.355 120.425 330.675 3864.620 3979.500 116.910 346.470 2436.750 252.110 188.260 3941.125 12746.630 427.505 2243.18515 24201.230 1989.515 7099.425 4106.760 134.035 35.575 8090.570 166.680 93.180 2837.200 16348.265 2612.760 853.870 11227.145 1894.565 5445.130 7066.335 1027.810 489.585 538.720 17274.435 8046.425 169.390 981.930 5222.200 623.105 724.000 17596.405 42547.660 1913.710 1009.74516 1308.635 625.650 1712.530 1443.615 45.760 16.940 1484.360 65.580 49.605 443.005 2700.300 1138.950 51.305 3716.885 547.885 1433.090 1594.945 188.280 109.040 123.880 1449.325 1134.105 52.620 361.205 431.955 126.625 88.950 284.985 19753.310 504.080 1203.76017 5401.610 779.400 17365.690 3133.785 135.870 45.945 5984.420 29.950 96.315 749.575 13121.500 977.325 112.150 17503.630 498.300 1016.380 925.000 183.255 312.320 338.515 2984.145 1279.535 126.190 777.545 1127.290 152.740 114.390 775.250 190165.195 797.325 1129.08018 16779.010 241.065 6098.210 1067.215 80.190 21.485 5083.830 80.745 75.375 3642.925 7565.165 2814.410 555.480 8380.150 738.525 3767.285 1150.845 492.605 147.160 290.260 11403.450 17318.580 110.680 504.370 913.495 907.865 297.090 281.490 175.145 533.750 135.82019 10617.030 1489.355 8201.840 4182.665 153.480 42.860 18310.745 59.745 117.205 3483.925 13692.035 1681.385 629.325 39190.170 1544.615 2543.185 4753.195 1227.995 887.160 441.255 12636.935 6457.760 157.620 1095.640 6961.160 222.810 786.935 5451.170 9584.190 3176.895 1103.24520 3516.845 1153.375 5925.540 858.490 78.580 40.120 1832.040 125.160 118.680 438.495 8013.790 1135.255 61.740 6448.680 858.035 2690.230 563.755 207.370 112.250 233.170 2208.790 3359.760 190.850 463.810 1013.430 191.125 97.095 414.095 120189.145 219.500 959.80021 26861.605 111.310 4354.520 756.815 96.350 16.245 1557.620 280.695 54.615 560.805 17096.170 1551.785 1471.730 5534.615 4616.070 6871.340 269.625 564.875 96.870 464.355 601.035 12629.290 98.005 388.435 150.890 808.635 264.485 151.450 4982.605 348.250 2155.34522 12807.305 2129.030 7536.480 6691.620 135.785 22.415 9611.250 287.935 89.260 3158.540 8622.360 2310.460 271.960 11421.595 1814.070 4989.590 10684.860 1549.590 435.035 429.100 15325.960 5179.475 733.890 797.425 12647.295 338.425 3241.065 42358.680 8471.295 3384.935 2366.16023 7442.640 368.940 36147.320 3600.240 182.105 62.835 15471.870 76.250 163.455 9832.310 7009.910 1844.185 189.230 21113.465 683.280 2196.490 1464.460 492.895 402.055 1388.165 127829.855 1239.370 5126.285 2532.405 937.055 1149.155 195.355 318.215 267.410 937.950 278.05524 18569.600 184.005 6605.380 948.180 79.370 21.580 3436.945 831.255 72.095 378.100 7384.410 1931.025 295.525 2373.475 1513.360 6172.860 552.465 158.860 124.270 351.425 15802.005 5214.335 123.060 397.710 250.450 835.260 121.485 225.860 60184.975 274.060 704.37525 961.765 218.660 2221.625 444.900 56.040 19.935 1431.275 61.115 28.890 358.595 9596.475 715.815 170.695 1199.330 1737.405 2777.120 252.995 69.380 72.110 262.505 721.295 3388.085 36.930 162.405 476.920 677.325 39.590 776.750 12094.805 168.850 1372.57526 14382.315 212.595 2891.215 677.120 101.235 45.975 1747.770 64.690 119.190 409.560 2082.595 1657.760 17.950 3615.270 307.700 1218.270 874.120 321.785 116.415 274.920 2236.045 3474.870 74.825 966.540 594.105 306.625 155.970 381.565 9437.020 205.875 249.90527 11982.890 101.000 7007.860 837.190 72.105 13.265 2275.475 253.450 46.005 1151.160 9571.670 1615.000 407.110 4457.155 1623.740 6566.750 232.895 340.150 63.875 222.545 2225.025 4528.715 5984.325 340.985 103.845 367.065 159.870 111.065 24590.780 221.675 734.11529 5621.860 449.275 8556.685 1139.925 65.575 17.380 1776.155 19.250 79.875 400.110 10042.835 852.625 63.130 9803.700 992.480 2519.950 413.115 132.095 85.650 230.390 12814.115 1758.730 83.280 304.795 596.510 119.265 70.590 161.415 55182.265 195.770 1803.33030 32039.210 1123.595 7166.000 2096.095 122.930 39.425 5454.805 101.910 75.445 3269.795 12094.375 1404.505 1306.750 28485.830 2488.620 2833.200 3210.565 808.080 767.635 433.080 13238.010 11029.645 81.910 909.260 2962.465 391.315 185.135 1371.100 456.340 1365.250 351.67531 1051.510 120.670 1933.340 377.700 74.520 24.110 1660.780 77.580 84.630 287.900 3624.470 2023.600 40.250 1039.770 575.030 1757.970 291.480 262.620 86.270 382.000 18161.450 759.390 61.720 574.580 989.090 216.090 80.320 125.900 51494.670 121.260 108.50033 8917.215 4419.210 2510.455 4284.415 143.540 32.395 13491.585 199.355 80.720 4151.850 7937.145 2167.865 1196.785 14209.455 1692.640 4962.125 12597.165 2487.710 1107.880 535.545 22324.180 6444.295 293.215 1208.490 10354.545 322.010 1794.025 17611.485 6418.000 3401.620 954.31534 11158.325 298.085 7812.080 2818.570 98.375 30.840 7158.670 156.455 132.065 4989.140 4156.380 1236.890 577.895 4782.110 1349.135 3162.205 752.390 1590.845 202.815 359.745 10497.310 5716.545 705.180 1012.945 1895.535 381.310 341.215 333.925 1755.640 1112.200 540.03035 6498.190 166.320 5674.980 757.040 50.895 11.560 2319.255 94.850 50.080 324.460 3783.225 887.555 62.350 1817.655 879.560 2927.775 178.475 108.745 69.090 166.200 1098.945 1388.770 113.535 301.550 261.030 92.515 54.750 125.545 13401.995 108.485 536.85036 25430.530 169.975 2953.925 1835.415 121.655 40.330 3934.095 291.440 120.215 1158.150 4795.225 2121.730 159.095 6586.075 2000.455 5060.365 1283.155 522.455 180.900 429.290 4203.185 6191.695 137.375 602.645 1848.405 5144.255 155.605 1595.955 116257.940 588.710 618.78037 8418.235 640.340 1014.090 1273.745 85.605 33.860 4313.065 337.760 46.440 1299.370 6551.705 1835.990 155.225 16442.440 1041.945 3059.790 1462.130 338.840 197.505 322.910 7168.585 1783.610 92.965 573.650 1120.950 222.750 142.685 1391.850 6020.480 337.315 4034.89038 36975.750 1492.010 14488.990 1207.155 121.950 34.965 2454.540 279.340 82.785 1385.885 21148.355 2991.855 655.355 16794.785 2056.520 6241.435 2423.595 549.235 592.330 572.280 30438.030 8982.390 253.865 1028.720 650.190 605.120 272.440 699.930 61955.570 806.795 701.56539 9440.790 89.805 17001.945 954.000 48.795 21.845 2395.445 76.040 56.905 775.690 8423.040 1032.445 268.045 5192.375 830.195 2456.770 316.865 222.350 86.060 219.655 8037.580 2468.075 918.675 461.400 337.430 244.930 146.140 189.795 12577.300 232.940 256.45540 11224.535 862.900 2124.585 1089.205 109.325 19.465 2628.460 177.250 78.545 1361.405 19019.590 2829.920 450.690 8184.480 1839.530 7934.215 1449.670 476.415 199.315 410.655 11590.560 8115.580 101.740 615.490 443.555 394.670 251.295 552.250 26310.815 303.770 457.585157   Cohort B           CaseID Beta-Catenin CD8A B7-H3 CD4 FoxP3 MmAb IgG2a CD68 PTEN Rabbit IgG CD14 S6 GZMB Ki67 (8D5) B2M Histone H3 AKT CD3 VISTA PD1 p-STAT3 CD44 STAT3  CD56 PD-L1 CD45 P-AKT CD19 CD20 PanCK CD45RO Bcl-21 26200.285 430.670 6987.325 2015.145 144.750 45.980 2146.190 98.850 63.410 1316.395 7821.135 607.055 2081.535 8784.280 2547.680 1374.915 1545.540 284.320 415.005 350.300 8132.210 3213.525 150.935 582.665 2182.680 1964.740 193.355 493.065 1843.775 1020.765 505.9752 77482.975 371.395 8218.300 2683.800 137.530 37.445 2825.665 247.780 72.200 1351.265 18290.545 1086.225 3859.965 10726.035 3488.300 3869.680 1336.510 340.010 393.565 427.615 8807.185 6137.020 95.570 754.465 5502.695 2437.740 498.205 11462.565 5136.460 1017.570 544.4303 38861.460 2989.260 2608.225 3517.260 251.790 86.105 5894.840 171.905 197.875 5057.635 28464.355 3861.295 1086.670 33445.490 1721.235 3206.755 10928.195 1593.435 1102.225 1088.895 59027.730 3160.320 337.260 3368.385 3242.705 704.435 344.885 1067.700 26170.825 1377.160 649.1904 32793.980 2382.340 14527.860 2148.750 161.570 52.780 35716.065 375.510 148.630 18681.415 13070.035 2702.325 382.315 65487.465 2260.915 19358.785 4998.390 1189.770 482.700 1115.465 61780.170 6601.690 273.460 1587.100 2488.030 1511.125 391.245 699.480 9089.905 3581.230 899.9205 29615.690 178.810 2738.690 759.805 41.265 21.145 1050.770 188.030 38.400 1333.440 4017.855 709.510 448.230 1478.100 881.265 850.495 414.570 123.965 89.855 177.185 2058.950 3529.295 30.460 310.695 206.710 740.435 214.560 81.325 5599.645 144.885 281.2506 14995.310 196.310 1352.975 1398.305 97.785 31.260 888.185 218.650 105.220 4010.350 2092.650 1151.530 113.230 4633.300 674.930 2063.440 537.850 578.830 121.505 383.780 12224.775 6702.185 212.045 717.750 140.745 201.710 202.825 279.360 919.190 128.980 375.3857 5970.905 4811.985 1059.045 2630.030 136.860 32.010 6084.835 352.080 87.875 2191.595 19136.515 1927.925 556.555 31313.425 1688.395 6740.830 4743.010 2104.740 960.350 668.375 2836.405 31211.330 186.910 2054.395 2097.310 329.640 485.765 6222.435 963.810 1419.590 1019.9108 18898.025 5164.415 12637.330 8225.530 1919.450 59.290 17126.305 326.655 247.130 10012.485 6627.585 3449.680 268.465 36950.825 1370.205 3431.540 18664.730 2398.460 1510.270 945.150 46515.570 11833.255 257.240 3095.345 19726.330 445.530 3143.995 34437.965 4456.615 7587.715 1649.7009 8497.905 996.500 7424.805 1622.680 63.725 13.180 2627.335 85.300 36.500 541.165 4478.645 971.795 361.780 8747.525 1166.810 2038.335 5070.370 239.430 120.385 199.390 3645.650 1713.465 82.695 498.195 4257.010 157.325 246.960 2377.380 7492.380 938.650 853.34010 26200.140 499.365 8040.920 1341.865 189.275 40.595 4606.995 307.500 114.745 1678.745 14959.895 3508.180 250.405 11948.175 1598.410 11464.655 1443.900 1225.165 427.690 1098.585 91209.150 11924.365 1235.855 1211.040 868.430 2248.665 582.985 831.340 30091.190 672.630 448.65511 5619.890 6652.755 3896.460 4148.500 158.245 25.130 4373.100 235.645 72.320 3566.920 10666.150 1361.510 555.545 28890.015 2158.415 5177.985 7894.115 1086.915 1816.245 440.400 3209.950 14167.630 255.445 1142.635 14700.855 380.255 1324.730 69850.755 877.835 2259.930 3078.32512 24794.840 1082.630 5360.050 5023.355 106.425 32.655 7135.630 353.080 115.380 2597.255 5163.380 2692.190 255.050 8838.140 833.035 5222.680 6911.010 1319.005 455.285 400.000 18793.740 3486.340 113.000 815.065 8889.165 418.945 1353.745 13212.760 2662.680 2418.500 653.26013 2330.440 591.140 10426.705 1004.820 117.400 62.690 2264.620 294.325 119.430 501.925 9381.315 2192.765 259.300 3826.615 1068.760 2457.085 1459.445 259.035 138.000 358.400 3193.605 4888.810 99.915 629.725 729.785 220.670 204.670 658.590 84449.635 609.590 3442.05014 27888.995 235.220 4111.290 765.750 162.810 77.805 10828.030 103.935 261.355 1447.670 6339.565 1088.340 280.025 9450.865 5978.755 446.020 495.325 1212.500 206.975 1022.450 43192.170 2329.045 195.485 2776.555 1568.175 567.925 323.270 184.500 52102.965 4862.870 196.67515 34597.390 12137.670 8466.270 9222.035 250.070 49.985 5859.320 492.165 143.670 3438.210 16281.600 3698.225 487.480 32004.400 2343.775 9520.985 27479.785 2292.295 1070.780 863.885 25319.460 28370.380 555.930 2109.470 36582.370 908.025 2015.540 55170.340 43086.575 5262.495 3651.65016 22803.910 305.180 4930.570 1185.285 91.840 16.690 2798.040 411.810 76.690 3084.640 6960.765 3038.670 476.640 28513.400 827.260 3197.465 753.480 451.985 93.025 452.975 17919.710 5909.400 104.085 639.225 475.115 505.955 280.735 223.470 22319.730 471.560 489.19017 11125.220 1261.480 4106.005 6139.570 90.260 35.680 4205.220 63.330 68.525 1027.010 4576.650 1743.500 790.000 12043.870 590.655 1865.145 8853.875 819.400 420.970 472.575 34275.505 2351.125 181.630 1497.745 5181.985 174.350 339.020 6727.175 1023.000 4859.650 505.20018 18934.935 8255.025 7103.930 9502.465 204.495 65.880 52437.960 197.660 174.565 1294.525 1355.965 1638.920 253.900 30796.615 895.065 2561.740 15137.405 1117.660 2125.985 570.975 5289.850 1743.935 995.060 2325.125 23331.505 325.100 888.350 19135.700 15119.965 7675.760 1537.35019 32886.620 1123.005 8103.615 10772.050 124.900 50.650 45537.770 38.440 92.570 7552.740 2157.180 1504.710 1019.415 12705.530 1334.725 361.295 2184.270 732.465 346.225 823.845 58100.835 732.065 224.315 1760.470 2358.895 527.660 245.580 294.345 15285.710 5101.770 707.54520 5707.080 1929.160 4376.830 3723.090 104.390 25.940 9546.850 215.380 92.860 2700.700 10646.320 2231.910 704.900 17803.090 1220.320 3319.370 3627.700 3920.810 512.580 454.390 18875.190 6489.000 154.110 1695.300 5970.450 227.780 2711.450 6648.620 2845.660 1992.310 915.980158   Appendix Table D.2 DSP signal to noise ratios values in each cohort for each biomarker Cohort A      CoreID Beta-Catenin CD8A B7-H3 CD4 FoxP3 MmAb IgG2aCD68 PTEN Rabbit IgGCD14 S6 GZMB Ki67 (8D5) B2M Histone H3AKT CD3 VISTA PD1 p-STAT3 CD44 STAT3  CD56 PD-L1 CD45 P-AKT CD19 CD20 PanCK CD45RO Bcl-22 442.8588 2.7054 420.3182 17.5926 1.9485 0.5972 24.1095 4.8743 1.6745 14.4284 135.9855 51.3427 2.4140 83.6705 42.8283 121.7890 3.3648 4.8392 2.3452 11.1711 380.4440 189.9007 2.9129 11.8291 7.7785 6.5680 1.9348 2.3860 1111.2390 3.5266 168.07063 575.4949 4.3577 340.4360 30.5876 1.7695 0.5641 52.4027 1.9867 1.7727 14.0053 193.2341 89.7125 46.0978 168.9488 55.5644 57.4158 15.8664 6.0785 2.9241 8.5078 271.3560 76.3787 38.0890 13.8483 12.0262 9.0464 4.3423 3.4109 1625.9342 8.0510 11.54984 63.4044 23.7522 38.0915 36.4584 1.3139 0.7080 13.8851 0.9106 1.4124 8.9696 82.0844 11.7319 0.3726 99.8939 17.2739 36.8396 24.2940 3.5190 4.9767 5.0120 47.9992 35.7175 2.2471 10.1071 159.8443 2.3262 8.0098 443.5152 315.2361 8.5938 35.77545 425.3429 1.7194 274.0164 18.1516 1.5436 0.5431 66.2646 2.0680 1.8414 31.2210 89.8482 26.9829 6.5241 76.6022 16.1923 50.2144 5.4143 15.7747 1.2899 4.7724 90.6384 53.7712 61.1982 12.8401 4.8222 5.1069 2.8256 1.9952 246.0069 3.5471 3.46886 159.4096 4.1002 73.7386 13.7594 1.6048 0.6587 19.4418 0.9477 1.5183 15.1607 169.0679 23.3351 27.6989 326.3154 43.1718 75.6057 5.6644 6.6652 2.3098 6.1251 168.2090 72.2030 4.2678 10.6114 5.6546 16.6567 1.6449 4.6166 286.3150 6.0553 20.63977 267.4888 5.5481 359.1837 45.3803 2.0022 0.5549 256.6346 5.3165 1.8021 54.4272 132.3993 61.7600 5.6972 91.6746 17.2762 75.7426 46.1578 26.6055 5.5612 10.5506 583.9866 74.3753 391.2638 15.0623 70.0279 7.0061 7.4048 4.8058 119.4695 27.2569 27.83728 691.9591 28.0862 133.7622 83.6325 2.6660 0.5466 56.0256 3.6539 1.8295 19.4242 400.2089 73.1684 12.8543 170.2334 54.0209 108.1919 154.8776 16.4295 7.8655 9.7065 123.2284 301.1339 3.4279 11.5280 173.8200 8.9826 6.0501 344.5348 258.7155 38.4208 30.17419 179.8068 5.1552 150.0855 15.1607 1.5829 0.4723 29.5021 3.3787 2.1174 7.8565 281.8647 40.7099 12.0825 147.7299 45.6124 125.1496 6.8659 4.1809 1.4328 4.8236 38.8531 81.9122 1.2430 7.1285 3.7872 6.4809 2.7536 3.1768 873.5923 3.8283 207.782010 93.0937 114.7182 70.4999 297.3341 3.2550 0.5153 45.0056 3.5952 1.9407 26.4672 159.7808 30.4379 4.6340 243.8153 65.7387 78.4037 457.6762 14.4260 30.9222 23.8339 216.9492 243.8671 8.1352 13.5667 488.2553 20.2355 12.8340 880.2374 195.9215 160.7110 65.294811 176.8571 7.6771 28.8079 16.5189 1.4792 0.4729 38.1213 0.6804 2.1145 34.8987 61.0179 39.7015 1.3035 212.6257 10.4504 13.7330 29.3052 9.1837 2.9423 6.5899 227.4795 29.2918 0.8730 13.8795 33.6280 4.0722 6.7536 2.3930 377.9699 10.1921 2.527212 342.0620 4.9329 121.4366 11.8891 1.7956 0.5974 35.2611 9.6535 1.6741 14.9515 225.7525 63.7893 11.3118 151.7480 75.5874 167.1656 10.1048 6.4949 2.4935 7.4339 47.8359 105.2903 2.9328 11.0812 9.8997 8.1344 5.2585 14.3918 2471.9136 5.3312 216.400013 94.2163 1.8583 193.9092 20.7386 3.8425 0.5958 71.6682 0.8962 1.6785 11.3588 35.1695 11.1782 0.4759 208.8811 3.4821 12.5015 4.6248 4.0830 4.5041 3.9631 88.9846 18.5139 2.8804 9.0742 9.0135 2.8804 0.4759 3.8425 30.1192 6.0672 2.219514 531.1338 17.0745 168.3575 43.9111 2.3155 0.5022 32.0786 7.0189 1.9913 25.6153 171.1006 45.6552 12.7203 65.5615 56.9062 102.8547 56.2757 10.0903 4.4452 12.2061 142.6533 146.8939 4.3155 12.7891 89.9469 9.3060 6.9492 145.4773 470.5118 15.7803 82.801915 420.3428 34.5552 123.3075 71.3289 2.3280 0.6179 140.5223 2.8950 1.6184 49.2783 283.9474 45.3801 14.8306 195.0004 32.9060 94.5746 122.7327 17.8517 8.5034 9.3568 300.0337 139.7556 2.9421 17.0548 90.7026 10.8225 12.5749 305.6259 738.9956 33.2386 17.537916 45.1439 21.5830 59.0770 49.8003 1.5786 0.5844 51.2059 2.2623 1.7112 15.2823 93.1521 39.2903 1.7699 128.2212 18.9004 49.4372 55.0207 6.4951 3.7615 4.2735 49.9973 39.1232 1.8152 12.4605 14.9011 4.3682 3.0685 9.8311 681.4287 17.3892 41.526017 81.2002 11.7164 261.0514 47.1089 2.0425 0.6907 89.9614 0.4502 1.4479 11.2681 197.2502 14.6917 1.6859 263.1250 7.4907 15.2788 13.9051 2.7548 4.6950 5.0888 44.8594 19.2347 1.8970 11.6885 16.9461 2.2961 1.7196 11.6540 2858.6760 11.9859 16.973018 416.9510 5.9904 151.5378 26.5198 1.9927 0.5339 126.3309 2.0065 1.8730 90.5251 187.9910 69.9368 13.8034 208.2430 18.3520 93.6154 28.5980 12.2410 3.6569 7.2128 283.3707 430.3590 2.7503 12.5334 22.6999 22.5600 7.3826 6.9949 4.3523 13.2635 3.375119 149.7973 21.0135 115.7210 59.0138 2.1655 0.6047 258.3490 0.8430 1.6537 49.1552 193.1830 23.7229 8.8792 552.9400 21.7932 35.8822 67.0635 17.3260 12.5171 6.2257 178.2964 91.1135 2.2239 15.4585 98.2160 3.1437 11.1030 76.9114 135.2248 44.8233 15.565820 50.9664 16.7148 85.8734 12.4413 1.1388 0.5814 26.5501 1.8138 1.7199 6.3547 116.1365 16.4522 0.8947 93.4548 12.4347 38.9870 8.1700 3.0052 1.6267 3.3791 32.0100 48.6899 2.7658 6.7216 14.6867 2.7698 1.4071 6.0011 1741.7909 3.1810 13.909521 901.8123 3.7370 146.1923 25.4082 3.2347 0.5454 52.2933 9.4236 1.8336 18.8276 573.9618 52.0974 49.4097 185.8111 154.9732 230.6883 9.0520 18.9643 3.2522 15.5896 20.1783 423.9973 3.2903 13.0407 5.0658 27.1479 8.8794 5.0846 167.2787 11.6916 72.360422 286.3254 47.5975 168.4887 149.6006 3.0357 0.5011 214.8731 6.4372 1.9955 70.6136 192.7651 51.6536 6.0801 255.3459 40.5561 111.5494 238.8752 34.6433 9.7258 9.5931 342.6335 115.7945 16.4072 17.8276 282.7482 7.5660 72.4586 946.9883 189.3878 75.6750 52.898923 73.4391 3.6405 356.6779 35.5248 1.7969 0.6200 152.6662 0.7524 1.6129 97.0187 69.1692 18.1972 1.8672 208.3337 6.7422 21.6735 14.4503 4.8636 3.9672 13.6975 1261.3406 12.2293 50.5828 24.9881 9.2462 11.3391 1.9276 3.1399 2.6386 9.2551 2.743724 470.7868 4.6650 167.4632 24.0388 2.0122 0.5471 87.1353 21.0744 1.8278 9.5858 187.2136 48.9564 7.4923 60.1737 38.3675 156.4978 14.0064 4.0275 3.1506 8.9095 400.6212 132.1967 3.1199 10.0830 6.3495 21.1760 3.0800 5.7261 1525.8428 6.9481 17.857725 40.0763 9.1114 92.5740 18.5388 2.3352 0.8307 59.6405 2.5466 1.2038 14.9425 399.8801 29.8276 7.1128 49.9755 72.3968 115.7212 10.5422 2.8910 3.0048 10.9384 30.0560 141.1797 1.5389 6.7673 19.8730 28.2238 1.6497 32.3668 503.9843 7.0359 57.194526 194.2889 2.8719 39.0571 9.1471 1.3676 0.6211 23.6104 0.8739 1.6101 5.5327 28.1335 22.3945 0.2425 48.8382 4.1567 16.4575 11.8084 4.3470 1.5726 3.7139 30.2065 46.9416 1.0108 13.0569 8.0257 4.1422 2.1070 5.1545 127.4835 2.7811 3.375927 485.0711 4.0885 283.6804 33.8897 2.9188 0.5370 92.1119 10.2597 1.8623 46.5993 387.4642 65.3757 16.4799 180.4270 65.7295 265.8241 9.4277 13.7694 2.5857 9.0087 90.0697 183.3238 242.2473 13.8032 4.2037 14.8589 6.4716 4.4959 995.4425 8.9735 29.717229 150.8863 12.0582 229.6546 30.5947 1.7600 0.4665 47.6706 0.5167 2.1438 10.7386 269.5417 22.8838 1.6944 263.1235 26.6374 67.6335 11.0877 3.5453 2.2988 6.1835 343.9207 47.2029 2.2352 8.1805 16.0099 3.2010 1.8946 4.3323 1481.0482 5.2543 48.399930 587.4633 20.6020 131.3941 38.4335 2.2540 0.7229 100.0180 1.8686 1.3833 59.9542 221.7596 25.7527 23.9603 522.3093 45.6307 51.9489 58.8681 14.8168 14.0752 7.9409 242.7290 202.2369 1.5019 16.6720 54.3190 7.1751 3.3946 25.1402 8.3673 25.0329 6.448231 23.2784 2.6714 42.8004 8.3615 1.6497 0.5337 36.7664 1.7175 1.8735 6.3735 80.2387 44.7985 0.8911 23.0185 12.7300 38.9180 6.4528 5.8139 1.9098 8.4567 402.0590 16.8114 1.3664 12.7201 21.8965 4.7838 1.7781 2.7872 1139.9912 2.6845 2.402033 174.3812 86.4202 49.0934 83.7842 2.8070 0.6335 263.8356 3.8985 1.5785 81.1918 155.2154 42.3938 23.4038 277.8740 33.1005 97.0372 246.3447 48.6486 21.6652 10.4729 436.5620 126.0218 5.7340 23.6327 202.4890 6.2971 35.0832 344.4026 125.5076 66.5206 18.662234 174.8429 4.6708 122.4096 44.1650 1.5415 0.4832 112.1712 2.4515 2.0694 78.1762 65.1275 19.3812 9.0552 74.9322 21.1400 49.5495 11.7894 24.9274 3.1780 5.6369 164.4853 89.5741 11.0497 15.8721 29.7017 5.9749 5.3466 5.2324 27.5096 17.4274 8.461935 270.0731 6.9125 235.8594 31.4636 2.1153 0.4804 96.3912 3.9421 2.0814 13.4850 157.2357 36.8879 2.5913 75.5441 36.5556 121.6821 7.4176 4.5196 2.8715 6.9075 45.6736 57.7191 4.7187 12.5328 10.8487 3.8450 2.2755 5.2178 557.0041 4.5088 22.312236 365.2262 2.4411 42.4235 26.3597 1.7472 0.5792 56.5004 4.1856 1.7265 16.6330 68.8677 30.4717 2.2849 94.5874 28.7300 72.6756 18.4283 7.5034 2.5980 6.1653 60.3650 88.9234 1.9729 8.6550 26.5463 73.8804 2.2348 22.9207 1669.6644 8.4549 8.886737 212.2909 16.1481 25.5733 32.1213 2.1588 0.8539 108.7668 8.5176 1.1711 32.7675 165.2208 46.3000 3.9145 414.6452 26.2758 77.1617 36.8720 8.5449 4.9807 8.1431 180.7773 44.9791 2.3444 14.4663 28.2681 5.6173 3.5982 35.0997 151.8244 8.5064 101.751838 687.2653 27.7319 269.3057 22.4373 2.2667 0.6499 45.6223 5.1921 1.5387 25.7593 393.0828 55.6094 12.1810 312.1633 38.2244 116.0090 45.0472 10.2086 11.0096 10.6369 565.7492 166.9550 4.7186 19.1207 12.0850 11.2473 5.0638 13.0095 1151.5632 14.9958 13.039939 267.7670 2.5471 482.2224 27.0581 1.3840 0.6196 67.9415 2.1567 1.6140 22.0007 238.9008 29.2830 7.6025 147.2702 23.5466 69.6808 8.9872 6.3065 2.4409 6.2300 227.9681 70.0015 26.0562 13.0866 9.5705 6.9469 4.1449 5.3831 356.7272 6.6068 7.273840 287.0661 22.0686 54.3360 27.8563 2.7960 0.4978 67.2225 4.5331 2.0088 34.8178 486.4236 72.3749 11.5263 209.3171 47.0458 202.9166 37.0751 12.1843 5.0975 10.5025 296.4271 207.5549 2.6020 15.7411 11.3439 10.0936 6.4268 14.1237 672.8958 7.7689 11.7027159  Cohort B CoreID Beta-Catenin CD8A B7-H3 CD4 FoxP3 MmAb IgG2aCD68 PTEN Rabbit IgGCD14 S6 GZMB Ki67 (8D5) B2M Histone H3AKT CD3 VISTA PD1 p-STAT3 CD44 STAT3  CD56 PD-L1 CD45 P-AKT CD19 CD20 PanCK CD45RO Bcl-21 485.2244265 7.975929 129.404 37.32011 2.680743 0.851541 39.74704 1.830684 1.174341 24.37939 144.846 11.24255 38.54964 162.6832 47.18256 25.46317 28.62311 5.265554 7.685816 6.487491 150.607 59.51389 2.795288 10.79085 40.42283 36.38662 3.580899 9.131472 34.14637 18.90438 9.3705632 1490.187002 7.142834 158.058 51.61603 2.645038 0.720159 54.34444 4.765415 1.388582 25.98813 351.7719 20.89076 74.23656 206.2879 67.08854 74.42341 25.70435 6.539223 7.569217 8.224082 169.3837 118.0299 1.838045 14.51021 105.8303 46.8837 9.5817 220.4531 98.78668 19.57036 10.470723 297.720996 22.90098 19.98184 26.94603 1.928985 0.659658 45.16088 1.316979 1.515937 38.74698 218.0679 29.58171 8.325073 256.2288 13.18653 24.56723 83.72184 12.20744 8.444241 8.342118 452.2165 24.21148 2.583778 25.80549 24.84264 5.396737 2.642194 8.179742 200.4969 10.55054 4.9735014 370.2594875 26.89774 164.0264 24.2604 1.824201 0.595911 403.2512 4.239685 1.678103 210.922 147.5669 30.51052 4.316517 739.3843 25.5268 218.5698 56.43418 13.43306 5.44991 12.59413 697.5272 74.53619 3.087492 17.91911 28.09103 17.06131 4.41734 7.897459 102.6293 40.43377 10.160525 1039.327452 6.275124 96.11107 26.66445 1.448146 0.742059 36.87552 6.598689 1.347602 46.79549 141.0018 24.89941 15.7301 51.87216 30.92695 29.84711 14.54884 4.350404 3.153354 6.218097 72.2564 123.8564 1.068958 10.90347 7.254242 25.98469 7.529728 2.854004 196.5129 5.084567 9.8701356 261.463947 3.422936 23.59099 24.38138 1.705017 0.545061 15.48673 3.812465 1.834656 69.92599 36.48824 20.07852 1.974321 80.78799 11.76834 35.97893 9.378158 10.0927 2.118608 6.691735 213.1558 116.8619 3.697298 12.51496 2.454084 3.517093 3.536534 4.871028 16.02735 2.248944 6.5453567 112.5808842 90.72955 19.9682 49.58898 2.580483 0.603546 114.729 6.638437 1.656875 41.32233 360.8173 36.35085 10.4938 590.4119 31.83454 127.0978 89.42903 39.68469 18.10731 12.60215 53.48016 588.4869 3.524171 38.73544 39.54459 6.215333 9.159056 117.3235 18.17255 26.76624 19.230318 156.1215878 42.6646 104.4003 67.95328 15.85708 0.48981 141.4849 2.698583 2.041606 82.71579 54.75223 28.49872 2.21786 305.2605 11.31963 28.34886 154.1943 19.81431 12.47674 7.808134 384.2774 97.75765 2.125128 25.57146 162.9644 3.680641 25.97338 284.5011 36.81728 62.68412 13.628619 387.4429246 45.43318 338.5173 73.98246 2.905398 0.600913 119.7874 3.889062 1.664136 24.6732 204.1938 44.30681 16.49455 398.8238 53.19809 92.93331 231.1722 10.91627 5.488684 9.09074 166.2152 78.12159 3.770293 22.71408 194.0888 7.172881 11.25959 108.3913 341.5983 42.79564 38.9061210 383.8843912 7.316695 117.8155 19.661 2.773257 0.594798 67.50168 4.505489 1.681243 24.59697 219.1923 51.40185 3.668933 175.0646 23.4199 167.9801 21.15602 17.95111 6.266513 16.09646 1336.396 174.7158 18.10774 17.74416 12.72423 32.94743 8.541895 12.18079 440.896 9.855373 6.57369211 131.8263908 156.0544 91.3997 97.31183 3.711971 0.589477 102.5803 5.527551 1.696418 83.66964 250.1971 31.93709 13.03148 677.6763 50.63018 121.4606 185.1731 25.49589 42.60386 10.33051 75.29616 332.3317 5.992002 26.80292 344.8396 8.919684 31.07434 1638.497 20.59147 53.01143 72.2086212 403.9440345 17.63762 87.32302 81.83776 1.733818 0.531997 116.2498 5.752187 1.879708 42.31306 84.11898 43.85969 4.155136 143.9862 13.57135 85.08506 112.5904 21.48851 7.417255 6.516582 306.1774 56.79755 1.840934 13.2786 144.8174 6.825224 22.05448 215.2551 43.37893 39.40089 10.6425613 26.93283528 6.83179 120.5012 11.61268 1.356789 0.724507 26.17216 3.401506 1.380249 5.800734 108.4196 25.34173 2.996723 44.22409 12.35163 28.39647 16.86677 2.993661 1.594862 4.14202 36.90841 56.49985 1.154715 7.277716 8.434107 2.550278 2.365366 7.611308 975.9823 7.045016 39.7796814 195.5751646 1.649511 28.83095 5.36992 1.141726 0.545618 75.93295 0.728858 1.832786 10.15197 44.45701 7.632124 1.963711 66.27541 41.92679 3.127773 3.47353 8.502812 1.451439 7.170062 302.8906 16.33273 1.370864 19.47095 10.99703 3.982647 2.266972 1.29383 365.3787 34.1015 1.37920915 408.2633954 143.2295 99.90546 108.8238 2.950929 0.589844 69.14238 5.807749 1.695365 40.57229 192.1296 43.64057 5.752464 377.665 27.65751 112.3515 324.2727 27.05002 12.63564 10.1942 298.78 334.7821 6.560202 24.89261 431.687 10.71507 23.7842 651.0326 508.4393 62.0996 43.0909716 637.3995919 8.530187 137.816 33.13029 2.56705 0.466508 78.20894 11.51064 2.143587 86.21979 194.5626 84.93487 13.32272 796.9874 23.12302 89.3734 21.06077 12.63358 2.600172 12.66125 500.8797 165.1756 2.909314 17.86719 13.2801 14.14211 7.846916 6.246284 623.8661 13.18073 13.6735117 224.9943762 25.51194 83.03908 124.1655 1.825401 0.721586 85.04559 1.280774 1.385837 20.77006 92.55732 35.26022 15.97681 243.573 11.9453 37.72034 179.0591 16.57139 8.513619 9.557269 693.1814 47.54871 3.673251 30.29012 104.7995 3.526022 6.856277 136.0491 20.68896 98.28066 10.2170718 176.5665523 76.97736 66.2435 88.60962 1.906897 0.614325 488.9792 1.843162 1.627803 12.07133 12.64425 15.28278 2.367594 287.1756 8.3464 23.88799 141.1549 10.42208 19.82462 5.32429 49.32737 16.26204 9.278844 21.68158 217.5642 3.031528 8.283783 178.4387 140.9923 71.57577 14.3356519 480.2798689 16.40049 118.3461 157.3162 1.824054 0.739698 665.0387 0.561382 1.351903 110.3011 31.5037 21.97495 14.88765 185.553 19.49247 5.276393 31.89932 10.697 5.056309 12.03152 848.5111 10.69116 3.275921 25.7101 34.44957 7.706006 3.586478 4.298647 223.2342 74.50682 10.3330720 116.2825735 39.30691 89.17854 75.85849 2.126961 0.528531 194.5184 4.388398 1.892036 55.02715 216.9203 45.47549 14.36244 362.7405 24.8642 67.63264 73.91491 79.88707 10.44389 9.258261 384.5847 132.2143 3.140013 34.54198 121.6488 4.64105 55.24618 135.4666 57.98073 40.5936 18.66322


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items