Biomarkers for treatment response after immunotherapy

Abstract

The present invention relates to the field of diagnostics for guiding cancer immunotherapy. More specifically, the present invention provides a method and kit for predicting and / or determining a subject's response to treatment with an immunotherapeutic agent based on the expression levels of a set of biomarkers.

Description

[Technical field]

[0001] The present invention relates to the field of diagnostics for guiding cancer immunotherapy. More specifically, the present invention provides methods and kits for predicting and / or determining a subject's response to treatment with an immunotherapeutic agent. [Background technology]

[0002] Cancer is a major public health challenge. In 2015, 1.3 million people died from cancer in the European Union, representing more than a quarter (25.4%) of the total deaths. Many treatments have been designed for different cancers. Immune checkpoint inhibitors (ICIs) are changing the therapeutic landscape for many tumor types, especially in metastatic settings. The development of ICIs targeting cytotoxic T-lymphocyte antigen 4 (CTLA-4) and programmed cell death 1 (PD-1 / PD-L1) has significantly improved the treatment of diverse cancers such as melanoma and non-small cell lung cancer (NSCLC). By blocking prototypic immune checkpoint receptors present on both immune and tumor cells and negatively regulating T-cell function, ICIs enhance antitumor immunity. Although these inhibitors can induce remarkable responses, cancer patients respond differently to ICI treatment. While approximately 25% to 30% of treated patients improve and overcome the disease immediately, 50% of treated patients do not see any lasting benefit. Even more alarming, a growing number of studies have shown that in a significant number of patients, between 4% and 29%, immunotherapy can accelerate tumor progression across multiple tissue structures, leading to so-called hyperprogression. Currently available diagnostics are not efficient enough to prognostically identify which patients will respond to immunotherapy such as ICI therapy, or which patients will experience tumor growth or are non-responders to ICI therapy. In particular, the accuracy of available biomarkers or sets of biomarkers is below 52-58% at best, providing little insight into the biological mechanisms underlying non-responsiveness to ICI. For example, WO2014151006 provides biomarkers for cancer patient selection and prognosis. However, this patent application is limited to predicting the responsiveness of diseased individuals to treatment with PD-L1 axis binding antagonists. WO2019012147 proposes radiomics-based biomarkers for detecting the presence and density of tumor-infiltrating CD8 T cells, predicting survival and / or treatment efficiency of cancer patients treated with immunotherapy such as anti-PD-1 / PD-L1 monotherapy. US20180107786 discloses a method for generating an immune score based on tumor infiltrating lymphocytes, T cell receptor signaling and gene mutation burden. Thus, there is a need for assays that can characterize the immune-tumor microenvironment for companion diagnostic development and treatment decisions. In particular, there is a high need for biomarkers to identify responders and non-responders to immunotherapeutic treatment. Summary of the Invention

[0003] The inventors have addressed the problem of predicting and / or determining the response of a subject diagnosed with cancer to treatment with an immunotherapeutic agent by identifying biomarkers that are able to predict the subject's response to the immunotherapeutic agent based on the RECIST response criteria. Thus, in the present invention, the inventors have identified biomarkers that are able to predict whether a subject will respond, partially respond, or not respond to treatment with an immunotherapeutic agent, more specifically an immune checkpoint inhibitor or a combination of immune checkpoint inhibitors. Thus, a first aspect of the present invention relates to a method for determining and / or predicting the response of a subject diagnosed with cancer to treatment with an immunotherapeutic agent. The method is based on the analysis of the expression levels of one or more biomarkers, more particularly the RNA expression levels of one or more protein-coding genes. Based on said expression levels, it is predicted whether the subject will respond to the immunotherapeutic agent.

[0004] Thus, in an embodiment of the invention, there is provided a method for determining and / or predicting the response of a subject diagnosed with cancer to treatment with an immunotherapeutic agent, comprising analyzing the expression levels of at least three biomarkers from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4; preferably at least four or five, even more preferably at least ten biomarkers. In another embodiment of the invention, there is provided a method for determining and / or predicting the response of a subject diagnosed with cancer to treatment with an immunotherapeutic agent, the method comprising analyzing the expression levels of at least three biomarkers from the list consisting of CD247, LAX1, IKZF3, CD3G, ITGAL, CD3E, TIGIT, PIK3CD, IGHV1.3 and IGHV4.4; preferably at least four or five, even more preferably at least ten biomarkers. In a further embodiment, the expression levels of the biomarkers CD247, LAX1 and IKZF3 are analyzed in a sample from a subject containing or suspected of containing tumor cells, and the expression levels of these biomarkers are used to determine and / or predict the response of the subject to treatment with an immunotherapeutic agent. In yet a further embodiment, the expression levels of the biomarkers CD247, LAX1 and IKZF3 are analyzed in combination with the expression levels of the biomarkers CD3G and ITGAL in a sample from a subject containing or suspected of containing tumor cells. In yet a further embodiment, the expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G and ITGAL are analyzed in combination with the expression levels of at least two; preferably at least four; even more preferably all of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in a sample from a subject containing or suspected of containing tumor cells.

[0005] In some embodiments of the methods, the expression levels of the analyzed biomarkers are compared to corresponding reference values ​​or threshold values ​​that are characteristic of subjects with a known response to treatment with an immunotherapeutic agent. In some embodiments, the methods of the invention further comprise obtaining a risk score based on the expression levels of the analyzed biomarkers, the risk score representing the likelihood of the subject to respond to treatment with an immunotherapeutic agent. In some embodiments, the response is a RECIST 1.1 criteria response, such as early death (ED), progressive disease (PD), stable disease (SD), partial response (PR), and complete response (CR).

[0006] In another embodiment, a kit for determining and / or predicting a subject's response to an immunotherapeutic agent is provided, comprising: determining the expression levels of at least three, preferably at least five, and even more preferably at least ten, biomarkers selected from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4 in a sample from the subject; and optionally means for measuring reference or threshold values ​​for at least three; preferably at least five, and even more preferably at least ten biomarkers selected from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4. Also disclosed is the use of the kit in a method for determining and / or predicting whether a subject diagnosed with cancer is likely to respond to treatment with an immunotherapeutic agent.

[0007] In some embodiments, the method further comprises using a pre-trained machine learning model to obtain and calculate a risk score using the expression levels of the biomarkers as inputs. In particular, the risk score is generated by a machine learning model using the normalized expression levels of the biomarker signatures as inputs. More specifically, the machine learning model uses a pre-trained architecture to calculate a risk score between 0 and 1. In some embodiments, a computer-implemented method is provided for monitoring, predicting or determining the response of a subject diagnosed with cancer to treatment with an immunotherapeutic agent, the method comprising: (a) detecting at least three of the following in a sample from the subject containing or suspected of containing tumor cells: CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4; (b) normalizing the quantified expression levels by comparing with data obtained from corresponding evaluations and expression levels of a reference set; (c) classifying whether the normalized value of step (a) exceeds a pre-defined threshold; and (d) obtaining a risk score of the normalized value calculated using a pre-trained machine learning model, the risk score representing the likelihood of the subject responding to the immunotherapeutic agent. In some embodiments, a computer-implemented method for monitoring, predicting or determining a subject's response to treatment with an immunotherapeutic agent is provided, comprising: (a) providing quantified expression levels of biomarkers CD247, LAX1 and IKZF3 in a sample of a subject containing or suspected of containing tumor cells; (b) optionally providing quantified expression levels of biomarkers CD3G and ITGAL in a sample of a subject containing or suspected of containing tumor cells; and / or providing quantified expression levels of biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4; (c) normalizing the quantified expression levels, such that normalization is performed by comparing with data obtained from corresponding evaluations and expression levels of a reference set; (d) classifying whether the normalized value of step (c) exceeds a predetermined threshold; and (e) obtaining a risk score of the normalized value, calculated using a pre-trained machine learning model, which represents the likelihood of the subject's response to the immunotherapeutic agent.

[0008] In some embodiments, the methods of the present application are characterized in that the immunotherapeutic agent is an immune checkpoint inhibitor, or a combination of several different immune checkpoint inhibitors. In some further preferred embodiments, the immunotherapeutic agent is selected from the group consisting of a PD-1 targeting agent, a PD-L1 targeting agent, and a CTLA-4 targeting agent, or a combination thereof. In some further preferred embodiments, the PD-1 targeting agent or the PD-L1 targeting agent is combined with a CTLA-4 targeting agent as an immunotherapeutic treatment.

[0009] As described above, in the methods of the present application, a sample from a subject diagnosed with cancer is analyzed to quantify the expression level of one or more biomarkers; in particular one or more protein-coding genes. The sample contains or is suspected to contain tumor cells. In some embodiments, the sample contains tumor cells. The sample is preferably a tumor tissue sample or a blood sample containing circulating tumor cells, or a sample suspected to contain tumor cells, such as a biopsy sample or liquid biopsy sample from a subject; preferably a subject diagnosed with cancer. In some embodiments, the sample is a tumor tissue sample. In some embodiments, the sample is a blood sample containing circulating tumor cells. In some embodiments, the sample is an isolated circulating tumor cell sample. In some embodiments, the sample is a cancer patient diagnosed with bladder cancer, kidney cancer, liver cancer, lung cancer, pancreatic cancer, prostate cancer, thyroid cancer, uterine cancer, ovarian cancer, colon cancer, breast cancer, head and neck cancer, skin cancer; even more preferably, a tumor tissue sample from a cancer patient diagnosed with skin cancer or lung cancer, such as melanoma, lung adenocarcinoma, lung squamous cell carcinoma, non-small cell lung cancer (NSCLC). In a preferred embodiment, the cancer is melanoma; preferably stage II, stage III or stage IV melanoma. In some embodiments, the sample is a blood sample containing circulating tumor cells, or a sample of isolated circulating tumor cells from a cancer patient diagnosed with bladder cancer, kidney cancer, liver cancer, lung cancer, pancreatic cancer, prostate cancer, thyroid cancer, uterine cancer, ovarian cancer, colon cancer, breast cancer, head and neck cancer, skin cancer; even more preferably, a sample of isolated circulating tumor cells from a cancer patient diagnosed with skin cancer or lung cancer, such as melanoma, lung adenocarcinoma, lung squamous cell carcinoma or non-small cell lung cancer (NSCLC). In a preferred embodiment, the cancer is melanoma. In some further preferred embodiments, the cancer is stage II, stage III or stage IV melanoma. [Brief description of the drawings]

[0010] [Figure 1](A) PCA analysis of the source dataset (log2 normalized RNA expression values ​​- pre-combat). (B) PCA analysis of the source dataset (log2 normalized RNA expression values ​​- post-combat). [Diagram 2] Normalized RNA expression levels (top 10 biomarkers based on LASSO learner machine learning model) after combat in discovery (DC) and validation cohorts (VC). ED-early death; PD-progression; SD-stable disease; PR-partial response; CR-complete response. [Diagram 3] (A) Discovery cohort: top 10 markers based on the LASSO Learner machine learning model. ED-Early Death; CR-Complete Response; (B) Discovery cohort: top 3 markers based on the LASSO Learner machine learning model. ED-Early Death; CR-Complete Response; (C) Validation cohort: top 10 markers identified in the discovery cohort. ED-Early Death; CR-Complete Response; (D) Validation cohort: top 3 markers identified in the discovery cohort. ED-Early Death; CR-Complete Response. [Figure 4] ROC AUC curves for ED+PD vs PR+CR prediction based on random forest binary classifier. The model is trained on 75% of the discovery cohort and then applied to the validation cohort. There is no mixed data between the cohorts and the datasets are sufficiently independent. The blue solid line shows the ROC AUC of the 3 marker model (CD247, LAX1, IKZF3); the purple dashed line shows the ROC AUC of the 25 marker model, and the green dashed line shows the ROC AUC of the baseline model (CD274 only); ED- early death; CR- complete response. [Diagram 5] (A) Volcano plot of DEG biomarkers between ED vs CR groups in the discovery cohort. Markers highlighted are those identified by the LASSO learner ML model. ED-early death; CR-complete response; (B) Volcano plot of DEG biomarkers between ED vs CR groups in the validation cohort. Markers highlighted are those identified by the LASSO learner ML model. ED-early death; CR-complete response. [Figure 6](A) Scaled heatmap showing log-normalized expression of 25 biomarkers for the discovery cohort. Hierarchical clustering shows prominent clusters of decreased gene expression, overrepresented in ED and PD patients. (B) Scaled heatmap showing log-normalized expression of 25 biomarkers for the validation cohort. Hierarchical clustering shows prominent clusters of decreased gene expression, overrepresented in ED and PD patients. ED-early death; PD-progression; SD-stable disease; PR-partial response; CR-complete response. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Unless otherwise defined, all terms used in disclosing the present invention, including technical and scientific terms, have the meanings commonly understood by one of ordinary skill in the art to which the present invention belongs. For further teaching, definitions of terms are included to better understand the teachings of the present invention. As used herein, the following terms have the following meanings. As used herein, "a," "an," and "the" refer to singular and plural referents unless otherwise clear from the context. By way of example, "a compartment" refers to one or more than one compartment. As used herein, "about" referring to a measurable value, such as a parameter, amount, temporal duration, and the like, is meant to encompass variations of up to ±20%, preferably up to ±10%, more preferably up to ±5%, even more preferably up to ±1%, and even more preferably up to ±0.1% from the specified value, where such variations are appropriate to practice the disclosed invention. However, it should be understood that the value to which the modifier "about" refers is itself specifically disclosed. As used herein, "comprise," "comprising," "comprises," and "comprised of" are synonymous with "include," "including," "includes," or "contain," "containing," "contains," and are inclusive or open-ended terms that specify the presence of, for example, the following components, but do not exclude or preclude the presence of additional, unrecited components, features, elements, materials, or steps that are known in the art or disclosed herein.

[0012] Moreover, the terms first, second, third, etc. in the specification and claims, unless otherwise noted, are used to distinguish between similar elements and do not necessarily describe an order or chronology. It is to be understood that the terms so used are interchangeable under appropriate circumstances, and that the embodiments of the invention described herein are capable of operating in orders other than those described or illustrated herein. The recitation of numerical ranges by endpoints includes all numbers and ratios subsumed within that range and the recited endpoints. Unless otherwise defined, the terms "% by weight", "percent by weight", "% by weight" or "% by weight" here and throughout this specification refer to the relative weight of each component based on the total weight of the formulation. The term "one or more" or "at least one", e.g., one or more of a group of members or at least one member, will be apparent from further illustration, but the term specifically includes reference to any one of the members, or any two or more of the members, e.g., any of ≧3, ≧4, ≧5, ≧6, or ≧7, etc. of the members, and all of the members.

[0013] "Gene" is used broadly herein to encompass polynucleotides that encode a gene product, including nucleic acid sequences defining an open reading frame. As used herein, "diagnosis" generally includes determining whether a subject is susceptible to a disease or disorder, whether a subject currently has a disease or disorder, as well as determining the prognosis of a subject with a disease or disorder. The terms "individual," "subject," "host," and "patient" are used interchangeably herein and refer to any mammalian subject for whom diagnosis, treatment, or therapy is desired, particularly humans. The term "sample" or "biological sample" encompasses a variety of sample types obtained from an organism and can be used in a diagnostic or monitoring assay. The term includes blood and other liquid samples of biological origin, solid tissue samples such as biopsies from living organisms and their progeny, or cultured tissues or cells. The term includes samples that have been manipulated in any way after procurement, such as by treatment with reagents, solubilization, or enrichment for specific components. The term encompasses clinical samples, including cell cultures, cell supernatants, cell lysates, serum, plasma, biological fluids, and cells in tissue samples. The terms "cancer," "neoplasm," "tumor," and "carcinoma" are used interchangeably herein to refer to cells that exhibit relatively autonomous growth and thus exhibit an abnormal proliferative phenotype characterized by a marked loss of cellular proliferation control. Generally, cells of interest for detection herein include immune cells in the tumor microenvironment.

[0014] The term "differentially expressed gene" is intended to encompass polynucleotides that represent or correspond to a differentially expressed gene in a tumor cell when compared to another cell of the same cell type. Such differentially expressed genes may include the open reading frame encoding the gene product (e.g., a polypeptide), as well as introns of such genes and adjacent 5' and 3' non-coding nucleotide sequences involved in regulating expression and up to about 20 kb outside the coding region, or further in either direction. Generally, a difference in expression level associated with a decrease in expression level of at least about 25%, usually at least about 50% to 75%, more usually at least about 90% or more, is indicative of a differentially expressed gene of interest, i.e., a gene that is under-expressed in the test sample relative to the control sample. Further, a difference in expression level associated with an increase in expression of at least about 25%, usually at least about 50% to 75%, more usually at least about 90%, can be at least about 1.5-fold, usually at least about 2-fold to about 10-fold, and can be about a 100-fold to about 1,000-fold increase relative to the control sample, indicating a differentially expressed gene of interest, i.e., an overexpressed or upregulated gene. As used herein, a "variable polynucleotide" refers to a nucleic acid molecule (RNA or DNA) that contains a sequence that represents a variable gene, e.g., the variable polynucleotide contains a sequence that specifically identifies the variable gene (e.g., an open reading frame that encodes a gene product), and detection of the variable polynucleotide in a sample correlates with the presence of the variable gene or a gene product of the variable gene in the sample. "Variable polynucleotide" is also meant to encompass fragments of the disclosed polynucleotides, e.g., fragments that retain biological activity, as well as nucleic acids that are homologous, substantially similar, or substantially identical (e.g., having about 90% sequence identity) to the disclosed polynucleotides.

[0015] The term "microenvironment" as used herein may refer to the tumor microenvironment as a whole or to individual subsets of cells within the microenvironment. Immune cells found within the tumor microenvironment are macrophages, monocytes, mast cells, helper T cells, cytotoxic T cells, regulatory T cells, natural killer cells, myeloid-derived suppressor cells, regulatory B cells, neutrophils, dendritic cells, and fibroblasts. In general, the tumor microenvironment is defined as a complex mixture of cells, soluble factors, signaling molecules, extracellular matrix, and mechanical cues that promote malignant transformation, support tumor growth and invasion, and protect tumors from host immunity. As used herein, the term "marker" or "biomarker" refers to a gene or genes, or a protein, polypeptide, or peptide expressed by this gene or genes, whose expression level alone or in combination with other genes correlates with a response to an immunotherapeutic treatment. The correlation can involve an increase or decrease in gene expression (e.g., an increase or decrease in the level of mRNA or a peptide encoded by the gene). As used herein, the term "in vitro" refers to an artificial environment and processes or reactions that occur within an artificial environment. In vitro environments can consist of, but are not limited to, test tubes and cell cultures. The term "in vivo" refers to a natural environment (e.g., an animal or a cell) and processes or reactions that occur within a natural environment. The term "in silico" refers to an artificial environment in which processing is performed or modeled using a computer system, thereby partially or completely avoiding the need to physically manipulate data (e.g., genes, polynucleotides, proteins, ...).

[0016] All references cited herein are hereby incorporated by reference in their entirety. In particular, the entire teachings of any reference specifically referenced herein are incorporated by reference. Unless otherwise defined, all terms used in disclosing the present invention, including technical and scientific terms, have the meanings commonly understood by those skilled in the art to which the present invention belongs. For further teaching, definitions of terms used in this description are included to better understand the teachings of the present invention. Terms or definitions used herein are provided only to aid in the understanding of the present invention. References throughout this specification to "one embodiment" or "an embodiment" mean that a particular feature, structure, or characteristic described in connection with this embodiment is included in at least one embodiment of the invention. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification do not necessarily all refer to the same embodiment, but may. Furthermore, particular features, structures, or characteristics may be suitably combined in one or more embodiments, as would be apparent to one of ordinary skill in the art from this disclosure. Furthermore, some embodiments described herein may include features included in other embodiments, while others may not, and combinations of features of different embodiments are within the scope of the invention and are meant to form different embodiments, as would be understood by one of ordinary skill in the art. For example, in the following claims, any of the claimed embodiments may be used in any combination.

[0017] In the present invention, the inventors have identified biomarkers that predict and / or determine the response of a subject to treatment with an immunotherapeutic agent. More specifically, the inventors show that it is possible to determine or predict the response of a subject diagnosed with cancer to immunotherapy based on the expression levels of at least three biomarkers selected from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4. More specifically, the inventors have found that it is possible to predict the response to immunotherapy based on the expression levels of a set of only three biomarkers: CD247, LAX1 and IKZF3. This set of three biomarkers can be optionally expanded with the biomarkers CD3G and ITGAL. Furthermore, the set of five biomarkers CD247, LAX1, IKZF3, CD3G and ITGAL can be further expanded with the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4, thereby forming a ten biomarker signature for predicting response to immunotherapy treatment. Immunotherapy using immune checkpoint inhibitors has transformed the therapeutic landscape for many tumor types, especially in metastatic settings. However, while significant and durable responses are achieved in a proportion of patients, the majority of patients respond only partially or not at all, or, even more troubling, a proportion of patients experience tumor growth. Tumor growth is a phenomenon defined clinically as an unexpected acceleration of tumor dynamics measured by imaging with dynamic parameters. High-progression tumor disease (HPD) following treatment with immunotherapies such as immune checkpoint inhibitors (ICIs) often results in the death of patients within 24–65 days after ICI treatment (Ferrara et al., 2020a). To date, the search for biomarkers that predict response to immunotherapy treatment and further predict whether a patient will develop HPD has been explored due to the dynamic interplay between ICIs and the immune microenvironment, as well as the heterogeneity of the immune environment in different tumor types (Davis and Patel, 2019). In this application, the inventors have identified several biomarkers that can predict and / or determine a subject's response to immunotherapy treatment.

[0018] In a first aspect, the present application provides a method for determining and / or predicting the response of a subject diagnosed with cancer to a treatment with an immunotherapeutic agent based on the expression analysis of several biomarkers. More specifically, the expression levels of at least three, preferably at least four or five, even more preferably at least ten or all biomarkers, preferably protein-coding genes, selected from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4 are assessed in a sample of a subject diagnosed with cancer and used in a method for determining and / or predicting the response to a treatment with an immunotherapeutic agent. In another embodiment, the expression levels of at least three biomarkers, preferably at least four or five biomarkers, and even more preferably all ten biomarkers selected from the list consisting of CD247, LAX1, IKZF3, CD3G, ITGAL, CD3E, TIGIT, PIK3CD, IGHV1.3 and IGHV4.4, are assessed in a sample from a subject diagnosed with cancer, and these expression levels are used in a method for determining and / or predicting response to treatment with an immunotherapeutic agent.

[0019] In a preferred embodiment, a method for determining and / or predicting the response or resistance of a subject diagnosed with cancer to treatment with an immunotherapeutic agent is provided. The method comprises analyzing the expression levels of the biomarkers CD247, LAX1 and IKZF3 in a sample from a subject containing or suspected of containing tumor cells. In a further embodiment, the method further comprises analyzing the expression levels of the biomarkers CD3G and ITGAL in a sample from a subject containing or suspected of containing tumor cells. In yet a further embodiment, the method further comprises analyzing the expression levels of at least two, at least three, at least four, or preferably all of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in a sample from a subject containing or suspected of containing tumor cells. Thus, in some embodiments, a method for determining and / or predicting the response of a subject to treatment with an immunotherapeutic agent is provided, the method comprises analyzing the expression levels of the biomarkers CD247, LAX1 and IKZF3 in a sample from a subject containing or suspected of containing tumor cells. In some embodiments, there is provided a method for determining and / or predicting a subject's response to treatment with an immunotherapeutic agent, the method comprising analyzing expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G, and ITGAL in a sample from a subject containing or suspected of containing tumor cells. In some embodiments, there is provided a method for determining and / or predicting a subject's response to treatment with an immunotherapeutic agent, the method comprising analyzing expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G, ITGAL, CD3E, TIGIT, PIK3CD, IGHV1.3, and IGHV4.4 in a sample from a subject containing or suspected of containing tumor cells.In some embodiments, any of these methods may further comprise analyzing the expression level of one or more biomarkers selected from the list consisting of CXCR6, SLAMF7, MYO1G, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, IGLV1.36, TRAV14DV4, TRAJ16, IGHV4.28, IGHV4.31 in the subject sample.

[0020] In one embodiment, a method for determining and / or predicting a subject's response to treatment with such an immunotherapeutic agent is thus provided, the method comprising analyzing the expression levels of biomarkers CD247, LAX1 and IKZF3 in a sample from the subject containing or suspected of containing tumor cells. In another embodiment, there is provided a method for determining and / or predicting the response of a subject diagnosed with cancer to treatment with such an immunotherapeutic agent, the method comprising analyzing the expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G and ITGAL in a sample from the subject containing or suspected of containing tumor cells. In yet another embodiment, there is provided a method for determining and / or predicting the response of a subject diagnosed with cancer to treatment with such an immunotherapeutic agent, the method comprising analyzing the expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G and ITGAL in a sample from the subject containing or suspected of containing tumour cells in combination with the expression levels of one or more of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4; preferably in combination with the expression levels of at least two, at least three or at least four of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4; even more preferably in combination with the expression levels of all of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4.

[0021] In still further embodiments, there is provided a method for determining and / or predicting the response of a subject diagnosed with cancer to treatment with such an immunotherapeutic agent, the method comprising analysing the expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G, ITGAL, CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in a sample from the subject containing or suspected of containing tumour cells. In another embodiment, the present application provides a method for determining and / or predicting a subject's response to treatment with an immunotherapeutic agent, the method comprising analyzing the expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G, ITGAL, CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in combination with one or more biomarkers selected from the list consisting of CXCR6, SLAMF7, MYO1G, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, IGLV1.36, TRAV14DV4, TRAJ16, IGHV4.28, IGHV4.31 in a sample containing or suspected of containing tumor cells.

[0022] In some embodiments, the expression level of a biomarker is compared to a corresponding reference value or threshold value that is characteristic of subjects with a known response to treatment with an immunotherapeutic agent, particularly an immune checkpoint inhibitor. In some embodiments, a risk score is obtained based on an expression analysis, the risk score representing the likelihood of a response to the immunotherapeutic agent. In some further embodiments, the risk score is compared to a threshold score, the threshold score being calculated based on expression analysis of biomarkers in samples obtained from subjects with known responses to treatment with the immunotherapeutic agent.

[0023] The list of biomarkers, preferably protein-coding genes, including CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4 thus comprises predictive biomarkers of one embodiment of the present application. Preferably, at least three; at least five; at least ten, or all of these predictive biomarkers are used to determine whether and how a subject will respond to treatment with an immunotherapeutic agent.

[0024] In a specific embodiment, the method as taught herein may include comparing the expression level of the biomarkers to one or more corresponding reference values ​​or threshold values, which are characteristic of subjects with a known response to treatment with an immunotherapeutic agent, particularly an immune checkpoint inhibitor. The reference values ​​or threshold values ​​may represent the expression level of the one or more biomarkers in subjects with a known prognosis after treatment with an immunotherapeutic agent. The known prognosis may be responding to treatment, having a partial response to treatment, not responding to treatment, or responding to treatment and developing tumor growth. In a specific embodiment, the known prognosis can be any of the Response Evaluation Criteria in Solid Tumors 1.1 (RECIST 1.1; (Eisenhauer et al., 2019) selected from progressive disease (PD), stable disease (SD), partial response (PR), and complete response (CR). Progression (PD) is further divided into early death (ED) and non-early death cohorts. ED is defined as progression-free survival (PFS) < 90 days and overall survival (OS) < 180 days (Ferrara et al., 2020b). In other embodiments, the reference value or threshold is determined by one or more biomarkers in healthy subjects. It may represent the expression level of a marker. The comparison may generally include any means of determining the presence or absence of at least one difference, and optionally the magnitude of the difference between the values ​​or profiles being compared. The comparison may include visual inspection, arithmetic comparison of measurements, or statistical comparison. Such statistical comparison includes, but is not limited to, applying an algorithm. When the value or expression profile of a biomarker includes at least one standard, the comparison to determine the difference in the value or expression profile of a biomarker may also include the measurements of these standards, and the measurements of the biomarker are correlated with the measurements of the internal standard. Reference values ​​or thresholds for the expression levels of any biomarker can be established according to known procedures previously used for other biomarkers. For example, a reference value for the expression level of a biomarker for determining whether a subject is likely to respond to an immunotherapeutic treatment as taught herein can be established by determining the amount or expression of the biomarker in a sample from an individual or a population (e.g., group) of individuals characterized by a known response to treatment with an immunotherapeutic agent. Such a population can include, but is not limited to, ≧2, ≧10, ≧100, or several hundred or more individuals. In certain embodiments, a reference value or threshold as contemplated herein may indicate an absolute amount of a biomarker as contemplated herein. In another embodiment, the amount of a biomarker in a sample from a test subject may be directly determined relative to the reference value (e.g., in terms of increase or decrease, or fold increase or decrease). Advantageously, this may allow the amount or expression level of a biomarker in a sample from a subject to be compared to the reference value (in other words, measuring the relative amount of a biomarker in a sample from a subject to the reference value) without first having to determine the absolute amount of each of the biomarkers.

[0025] As explained, the method, use or product may comprise finding out whether there is a deviation or not between the expression level of one or more biomarkers in a sample of a subject as taught herein and a given reference value or threshold value. A "deviation" of a first value from a second value or a "difference" between a first value and a second value can generally encompass any direction (e.g., increase: first value > second value; or decrease: first value < second value) and any degree of change. For example, a deviation or difference can include, but is not limited to, a decrease in a first value relative to the second value it is compared to of at least about 10% (about 0.9-fold or less), or at least about 20% (about 0.8-fold or less), or at least about 30% (about 0.7-fold or less), or at least about 40% (about 0.6-fold or less), or at least about 50% (about 0.5-fold or less), or at least about 60% (about 0.4-fold or less), or at least about 70% (about 0.3-fold or less), or at least about 80% (about 0.2-fold or less), or at least about 90% (about 0.1-fold or less). For example, a deviation or difference can include, but is not limited to, an increase in a first value relative to a second value being compared of at least about 10% (about 1.1 times or more), or at least about 20% (about 1.2 times or more), or at least about 30% (about 1.3 times or more), or at least about 40% (about 1.4 times or more), or at least about 50% (about 1.5 times or more), or at least about 60% (about 1.6 times or more), or at least about 70% (about 1.7 times or more), or at least about 80% (about 1.8 times or more), or at least about 90% (about 1.9 times or more), or at least about 100% (about 2 times or more), or at least about 150% (about 2.5 times or more), or at least about 200% (about 3 times or more), or at least about 500% (about 6 times or more), or at least about 700% (about 8 times or more), etc. Preferably, deviation or difference can refer to a statistically significant observed change. For example, deviation or difference can refer to an observed change that is outside the margin of error of a reference value in a given population (as described above, for example, the standard deviation or standard error, or a predetermined multiple thereof, such as ±1×SD or ±2×SD or ±3×SD, or ±1×SE or ±2×SE or ±3×SE difference). Deviation or difference can also refer to a value that is outside the reference range defined by the values ​​of a given population (for example, outside the range that includes ≧40%, ≧50%, ≧60%, ≧70%, ≧75%, or ≧80%, or ≧85%, or ≧90%, or ≧95%, or ≧100% of the values ​​in the population).

[0026] In further embodiments, a deviation or difference may be declared if the observed change exceeds a given threshold or cutoff. Such threshold or cutoff may be selected as is generally known in the art to provide a selected accuracy, sensitivity and / or specificity of the prediction method, for example an accuracy, sensitivity and / or specificity of at least 50%, or at least 60%, or at least 70%, or at least 80%, or at least 85%, or at least 90%, or at least 95%. For example, receiver operating characteristic (ROC) curve analysis can be used to select optimal threshold or cut-off values ​​for a given biomarker amount for clinical use of the diagnostic test, based on acceptable overall accuracy, sensitivity and / or specificity or related performance measures known per se, such as positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-), Youden index or similar. For example, as described in Robin X., PanelomiX: A threshold-based algorithm to create panels of biomarkers, 2013, Translational Proteomics, 1(1):57-64, optimal thresholds or cutoff values ​​for each individual biomarker can be selected as the extremes of the receiver operating characteristic (ROC) curve, i.e., the points furthest from the diagonal line. Those skilled in the art will understand that it is not appropriate to indicate exact threshold or cut-off values: relevant threshold or cut-off values ​​can be obtained by correlating the sensitivity and specificity and sensitivity / specificity for any threshold or cut-off value. It is up to the diagnostician to decide what level of positive predictive value / negative predictive value / sensitivity / specificity is desirable and how much of a decrease in positive or negative predictive value is acceptable. The selected threshold or cut-off level may depend on other diagnostic parameters used in combination with the method by the diagnostician.

[0027] In some embodiments, the methods, uses or products as taught herein generate a risk score. The generated risk score represents the likelihood of a subject responding to immunotherapy treatment. In some embodiments, the risk score ranges from 0 to 1, indicating the likelihood of a subject responding to immunotherapy treatment, with 1 being 100% likelihood of the subject responding to treatment with an immunotherapy agent and 0 being 0% likelihood of responding to treatment with an immunotherapy agent. In some embodiments, the likelihood value is generated by a machine learning model, using the normalized expression level of one biomarker signature as input data. In further embodiments, the machine learning model is a pre-trained machine learning model, and a pre-training architecture is used to calculate a risk score between 0 and 1. This pre-training architecture is trained and updated based on a pool of retrospective and prospective samples with previously known annotations regarding response to immunotherapy treatment. In particular, this pre-training architecture is trained and updated based on a pool of reference values ​​or thresholds obtained from samples of subjects with previously known annotations regarding response to immunotherapy treatment.

[0028] In some embodiments, the machine learning model is selected from (1) a linear mixed-effects model with random effects, (2) differential expression analysis, (3) a random forest machine learning model, (4) a gradient boosting machine model, and (5) a pre-training architecture based on a novel deep regression model, such as VGG16 and ResNet50 architectures. In some embodiments, a combination of machine learning models can be applied. Each model is run independently on each dataset and then collectively on the collective data to ensure that the signatures are well replicated in independent trials. In some further embodiments, gene and label bootstrapping and independent cohort comparisons can be applied to rule out the probability of random matching of these models and ensure that one or more biomarkers do not appear alone in the data by chance.

[0029] In certain preferred embodiments, analyses such as, inter alia, biomarker expression analysis and calculation of risk scores are thus trained and updated using machine learning, such as gradient boosting machine (GBM) classifiers, to learn specific biomarker expression signature patterns that are associated with a subject's response to treatment with an immunotherapeutic agent. According to some embodiments, expression levels of one or more biomarkers selected from CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4 are used in a gradient boosting machine (GBM) classifier to learn specific biomarker expression signature patterns associated with the subject's response to one or more immunotherapies. Response variables in these GBM models are further optimized for RECIST 1.1 criteria selected from early death (ED), progressive disease (PD), partial response (PR) and complete response (CR), as well as response to immunotherapies. In a specific embodiment, the expression levels of the biomarkers CD247, LAX1 and IKZF3 are used in a gradient boosting machine (GBM) classifier to learn a specific biomarker expression signature pattern associated with a subject's response to immunotherapy. In another embodiment, the expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G and ITGAL are used in a gradient boosting machine (GBM) classifier to learn a specific biomarker expression signature pattern associated with a subject's response to immunotherapy. In yet a further embodiment, the expression levels of the biomarkers CD247, LAX1, IKZF3, CD3G, ITGAL in combination with at least two, at least three, at least four or preferably all of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 are used in a gradient boosting machine (GBM) classifier to learn a specific biomarker expression signature pattern associated with a subject's response to immunotherapy.

[0030] In some embodiments, the methods of the invention in which a risk score is obtained are further supplemented by the use of other cancer diagnostics known to those of skill in the art. The methods, uses and products are thus intended to predict and / or determine the subject's response to treatment with an immunotherapeutic agent. The subject's response can be determined based on RECIST 1.1 response criteria and can be selected from early death (ED), progressive disease (PD), stable disease (SD), partial response (PR) and complete response (CR) (https: / / recist.eortc.org / recist-1-1-2 / ; Eisenhauer et al., 2009).

[0031] Typically, the methods, kits or uses taught herein feature expression analysis of one or more biomarkers in a sample from a subject containing or suspected of containing tumor cells. As used herein, "biomarker" is a common term in the art and may broadly refer to a biological molecule and / or a detectable portion thereof, the qualitative and / or quantitative assessment in a subject that is predictive or informative (e.g., predictive, diagnostic and / or prognostic) for one or more aspects of the subject's phenotype and / or genotype, such as the subject's status with respect to a given disease or condition. In a preferred embodiment, the biomarkers taught herein are protein-coding genes, and the expression levels of the protein-coding genes are assessed. In an even more preferred embodiment, the RNA-based expression levels of one or more protein-coding genes are assessed. The expression levels of the protein-coding genes disclosed herein may be assessed by any method known in the art that is suitable for determining a specific gene expression level in a sample. Such methods are well known and routinely practiced in the art. Gene expression analysis can be performed, for example, by reverse transcriptase real-time quantitative PCR, gene expression arrays, TruSeq gene expression analysis, in-situ hybridization, Dye-sequencing, pyrosequencing, CRISPR-Cas-based forms of transcriptome sequencing, or other forms (total RNA-Seq, mRNA-Seq, gene expression profiling). Additionally, a customized version of the TruSeq Targeted RNA Expression Kit can be used to measure the expression levels of multiple genes listed here. The TruSeq Targeted RNA Expression Kit enables highly customizable midplex to highplex gene expression profiling studies, with multiplexing up to 384 samples as well as a predefined panel of 12–1,000 assays targeting individual exons, isoforms, splice junctions, coding SNPs (cSNPs), gene fusions, and noncoding RNA transcripts. Additionally, customized versions of the TruSeq Custom Amplicon Kit Dx can be used for qualitative and / or quantitative evaluation of multiple genes listed herein.The TruSeq Custom Amplicon Kit Dx is an FDA cleared, FDA regulated, CE-IVD marked amplicon sequencing kit that enables clinical laboratories to deploy their own next-generation sequencing (NGS) assays for use on the FDA regulated, CE-IVD marked MiSeqDx and NextSeq550dx instruments.

[0032] In certain other embodiments, the biomarkers taught herein may be peptide, polypeptide and / or protein based. Reference to any marker, including any gene, peptide, polypeptide, protein, corresponds to the marker, gene, peptide, polypeptide, protein as commonly known in the art by the respective name. These terms encompass markers, genes, peptides, polypeptides, proteins of any organism, particularly animals, preferably warm-blooded animals, more preferably vertebrates, even more preferably mammals, including human and non-human mammals, and even more preferably human. These terms specifically encompass markers, genes, peptides, polypeptides, proteins having native sequences, i.e., those whose primary sequence is the same as the marker, gene, peptide, polypeptide, protein found or derived in nature. Those skilled in the art will appreciate that native sequences may vary between different species due to genetic divergence between such species. Additionally, native sequences may vary between or within different individuals of the same species due to normal genetic variation (mutation) within a given species. Native sequences may also vary between or within different individuals of the same species due to genetic alterations or post-transcriptional or post-translational modifications. Any such variants or isoforms of the markers, genes, peptides, polypeptides, proteins are contemplated herein. As a result, the entire sequence of a marker, gene, peptide, polypeptide, or protein as found in or derived from nature is considered to be "native". These terms encompass markers, genes, peptides, polypeptides, or proteins when they form part of an organism, organ, tissue, or cell, when they form part of a biological sample, as well as when they are at least partially isolated from such a source. These terms also encompass markers, genes, peptides, polypeptides, or proteins when they are produced by recombinant or synthetic means. In certain embodiments, the biomarkers taught herein may be any of the human biomarkers, such as, in particular, any of the biomarkers selected from CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, and IGHV4.4.

[0033] Reference to "CD247" refers to the CD247 marker, peptide, polypeptide, protein, or nucleic acid, as is commonly known in the art. For further teaching, CD247 is also known as "CD3-zeta," "CD3H," "CD3Q," "CD3Z," "IMD25," "T3Z," or "TCRZ." These terms refer to CD247 nucleic acids, as well as CD247 peptides, polypeptides, and proteins, as is clear from the context. By way of example, human CD247 mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nih.gov / ) accession numbers NM_000734.4 ("isoform 2 precursor"), NM_198053.3 ("isoform 1 precursor"), XM_011510145.2 ("isoform X2"), XM_011510144.3 ("isoform X1"), NM_001378516.1 ("isoform 4 precursor"), NM_001378515.1 ("isoform X3 precursor"). Nucleotides 65 (start codon) to 556 (stop codon) of NM_000734.4 constitute the CD247 coding sequence (CDS). By way of example, nucleotides 65-556 of NM_00734.4 are reproduced below. ATGAAGTGGAAGGCGCTTTTCACCGCGGCCATCCTGCAGGCACAGTTGCCGATTACAGAGGCACAGAGCTTTGGCCTGCTGGATCCCAAACTCTGCTACCTGCTGGATGGAATCCTCTTCATCTA TGGTGTCATTCTCACTGCCTTGTTCCTGAGAGTGAAGTTCAGCAGGAGCGCAGACGCCCCCGCGTACCAGCAGGGCCAGAACCAGCTCTATAACGAGCTCAATCTAGGACGAAGAGAGGAGTACG ATGTTTTGGACAAGAGACGTGGCCGGGACCCTGAGATGGGGGAAAGCCGAGAAGGAAGAACCCTCAGGAAGGCCTGTACAATGAACTGCAGAAAGATAAGATGGCGGAGGCCTACAGTGAGATT GGGATGAAAGGCGAGCGCCGGAGGGGCAAGGGGCACGATGGCCTTTACCAGGGTCTCAGTACAGCCACCAAGGACACCTACGACGCCCTTCACATGCAGGCCCTGCCCCCTCGCTAA (SEQ ID NO: 1)

[0034] By way of example, the human CD247 protein sequence is annotated under NCBI Genbank accession numbers NP_000725.1, NP_932170.1, XP_011508447.1, XP_011508446.1, NP_001365445.1, NP_001365444.1. As an example, the amino acid sequence of NP_000725.1 is reproduced further below. MKWKALFTAAILQAQLPITEAQSFGLLDPKLCYLLDGILFIYGVILTALFLRVKFSRSADAPAYQQGQNQLYNELNLGRREEYDVLDKRRGRDPEMGGKPRRKNPQEGLYNELQKDKMAEAYSEIGMKGERRRGKGHDGLYQGLSTATKDTYDALHMQALPPR (SEQ ID NO: 2) For example, the human CD247 gene is annotated at NCBI Genbank Gene ID 919.

[0035] Reference to "LAX1" ("lymphocyte transmembrane adaptor 1") refers to the LAX1 marker, peptide, polypeptide, protein, or nucleic acid, as such names are commonly known in the art. For further teaching, LAX1 is also known as "LAX." These terms refer to LAX1 nucleic acids, as well as LAX1 peptides, polypeptides, and proteins, as the context will indicate. By way of example, human LAX1 mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nih.gov / ) accession numbers NM_017773.4 ("isoform a"), NM_006711397.4 ("isoform X1"), NM_001136190.2 ("isoform b") or NM_001282878.1 ("isoform c"). Nucleotides 384 (start codon) to 1580 (stop codon) of NM_017773.4 constitute the LAX1 coding sequence (CDS). By way of example, nucleotides 384 to 1580 of NM_017773.4 are reproduced below.

[0036] By way of example, the human LAX1 protein sequence is annotated under NCBI Genbank accession numbers NP_060243.2, NP_00112966.2, NP_001269807. As an example, the amino acid sequence of NP_060243.2 is reproduced further below. MDGVTPTLSTIRGRTLESSTLHVTPRSLDRNKDQITNIFSGFAGLLAILLVVAVFCILWNWNKRKKRQVPYLRVTVMPLLTLPQTRQRAKNIYDILPWRQEDLGRHESRSMRIFSTESLLSRNSESPEHVPSQAGNAFQEHTAHIHATEYAVGIYDNAMVPQMCGNLTPSAHCINVRASRDCASISSEDSHDYVNVPTAEEIAETLASTKSPSRNLFVLPSTQKLEFTEERDEGCGDAGDCTSLYSPGAEDSDSLSNGEGSSQISNDYVNMTGLDLSAIQERQLWVAFQCCRDYENVPAADPSGSQQQAEKDVPSSNIGHVEDKTDDPGTHVQCVKRTFLASGDYADFQPFTQSEDSQMKHREEMSNEDSSDYENVLTAKLGGRDSEQGPGTQLLPDE (SEQ ID NO: 4) For example, the human LAX1 gene is annotated at NCBI Genbank Gene ID 54900.

[0037] Reference to "IKZF3" ("IKAROS family zinc finger 3") refers to an IKZF3 marker, peptide, polypeptide, protein, or nucleic acid, as such names are commonly known in the art. For further teaching, IKZF3 is also known as "AIO," "AIOLOS," "IMD84," and "ZNFN1A3." These terms refer to IKZF3 nucleic acids, as well as IKZF3 peptides, polypeptides, and proteins, as the context will indicate. For example, human IKZF3 mRNA is available under the NCBI Genbank (http: / / www.ncbi.nlm.nih.gov / ) accession numbers XM_047435625.1 (isoform X1), NM_001284516.1 (isoform 14), NM_001257409.2 (isoform 8), NM_001257408.2 (isoform 7), NM_001257413.2 (isoform 12), NM_001257412.2 (isoform 11), NM_001257411.2 (isoform 10), and NM_0012574 NM_001284514.2 (isoform 14), NM_183232.3 (isoform 6), NM_183229.3 (isoform 3), NM_183230.3 (isoform 4), NM_012481.5 (isoform 1), and NM_001284514.2 (isoform 14). Nucleotides 187 (start codon) to 1497 (stop codon) of NM_001257409.2 constitute the IKZF3 coding sequence (CDS). By way of example, nucleotides 187 to 1497 of NM_001257409.2 are reproduced below.

[0038] By way of example, the human IKZF3 protein sequence is annotated under NCBI Genbank accession numbers XP_047291581.1, NP_001271445.1, NP_001244338.1, NP_001244337.1, NP_001244342.1, NP_001244340.1, NP_001244339.1, NP_001244343.1, NP_899054.1, NP_899051.1, NP_001271444.1, NP_899055.1, NP_899052.1, NP_899053.1, NP_036613.2, NP_001271443.1. As an example, the amino acid sequence of NP_001244338.1 is further reproduced below. MEDIQTNAELKSTQEQSVPADDSMKVKDEYSERDENVLKSEPMGNAEEPEIPYSYSREYNEYENIKLERHVVSFDSSRPTSGKMNCDVCGLSCISFNVLMVHKRSHTGERP FQCNQCGASFTQKGNLLRHIKLHTGEKPFKCHLCNYACQRRDALTGHLRTHSASAEARHIKAEMGSERALVLDRLASNVAKRKSSMPQKFIGEKRHCFDVNYNSSYMYEKE SELIQTRMMDQAINNAISYLGAEALRPLVQTPPAPTSEMVPVISSMYPIALTRAEMSNGAPQELEKKSIHLPEKSVPSERGLSPNNSGHDSTDTDSNHEERQNHIYQQNHMVLSRARNGMPLLKEVPRSYELLKPPPICPRDSVKVINKEGEVMDVYRCDHCRVLFLDYVMFTIHMGCHGFRDPFECNMCGYRSHDRYEFSSHIARGEHRALLK (SEQ ID NO: 6) For example, the human IKZF3 gene is annotated at NCBI Genbank Gene ID 22806.

[0039] Reference to "ITGAL" ("integrin subunit alpha L") refers to an ITGAL marker, peptide, polypeptide, protein, or nucleic acid, as commonly known in the art. For further teaching, ITGAL is also known as "CD11A", "LFA-1", "LFA1A". These terms refer to ITGAL nucleic acids, as well as ITGAL peptides, polypeptides, and proteins, as the context may dictate. By way of example, human ITGAL mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_002209.3, XM_005255.13.1, XM_006721044.2, NM_001114380.2, XM_047434073.1, XM_047434072.1. By way of example, human ITGAL proteins are annotated under NCBI Genbank accession numbers NP_002200.2, XP_005255370.1, XP_006721107.1, NP_001107852.1, XP_047290029.1, XP_047290028.1.

[0040] Reference to "CD3G" ("CD3 gamma subunit of the T cell receptor complex") refers to the CD3G marker, peptide, polypeptide, protein or nucleic acid, as commonly known in the art. For further teaching, CD3G is also known as "CD3 gamma", "IMD17" or "T3G". These terms refer to CD3G nucleic acids, as well as CD3G peptides, polypeptides and proteins, as the context will allow. By way of example, human CD3G mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_000073.3, XM_006718941.4 or XM_005271724.5. By way of example, human CD3G protein is annotated under NCBI Genbank accession numbers NP_000064.1, NP_006719004.1 or XP_005271781.1. By way of example, the human CD3G gene is annotated at NCBI Genbank ID:917. Reference to "CD3E" ("CD3 epsilon subunit of the T cell receptor complex") refers to the CD3E marker, peptide, polypeptide, protein or nucleic acid, as commonly known in the art. For further teaching, CD3E is also known as "CD3 epsilon", "IMD18", "T3E" or "TCRE". These terms refer to CD3E nucleic acids, as well as CD3E peptides, polypeptides and proteins, as the context will allow. By way of example, human CD3E mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession number NM_000733.4. By way of example, human CD3E protein is annotated under NCBI Genbank accession number NP_XXX NP_000724.1. By way of example, human CD3E gene is annotated under NCBI Genbank ID:916.

[0041] Reference to "CXCR6" ("CXC motif chemokine receptor 6") refers to a CXCR6 marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, CXCR6 is also known as "BONZO", "CD186", "CDw186", "STRL33", "TYMSTR". These terms refer to CXCR6 nucleic acids, as well as CXCR6 peptides, polypeptides and proteins, as the context will allow. By way of example, human CXCR6 mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_001386436.1, NM_001386437.1, NM_006564.2 or NM_001386435.1. By way of example, the human CXCR6 protein is annotated under NCBI Genbank accession numbers NP_001373365.1, NP_00137336.1, NP_006555.1 or NP_001373364.1. By way of example, the human CXCR6 gene is annotated under NCBI Genbank ID: 10663. Reference to "SLAMF7" ("SLAM family member 7") refers to the SLAMF7 marker, peptide, polypeptide, protein, or nucleic acid, as such names are commonly known in the art. For further teaching, SLAMF7 is also known as "19A," "CD319," "CRACC," and "CS1." These terms refer to SLAMF7 nucleic acids, as well as SLAMF7 peptides, polypeptides, and proteins, as the context indicates. By way of example, human SLAMF7 mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_001282588.2, NM_001282589.2, NM_001282590.2, NM_001282591.2, NM_001282592.2, NM_001282593.2, NM_001282594.2, NM_001282595.1, NM_001282596.2 or NM_021181.5. By way of example, the human SLAMF7 protein is annotated under NCBI Genbank accession numbers NP_001269517.1, NP_001269518.1, NP_001269519.1, NP_001269520.1, NP_001269521.1, NP_001269522.1, NP_001269523.1, NP_001269524.1, NP_001269525.1 or NP_067004.3. NP_XXX. By way of example, the human SLAMF7 gene is annotated under NCBI Genbank ID:57823.

[0042] Reference to "MYO1G" ("myosin IG") refers to MYOIG markers, peptides, polypeptides, proteins or nucleic acids, as commonly known in the art. For further teaching, MYOIG is also known as "HA", "HLA-HA2" or "MHAG". These terms refer to MYO1G nucleic acids, as well as MYO1G peptides, polypeptides and proteins, as the context may dictate. By way of example, human MYO1G mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession number NM_033054.3. By way of example, human MYO1G protein is annotated under NCBI Genbank accession number NP_149043.2. By way of example, human MYO1G gene is annotated under NCBI Genbank ID:64005. Reference to "TIGIT" ("T cell immunoreceptor with Ig and ITIM domains") refers to TIGIT markers, peptides, polypeptides, proteins or nucleic acids, as those names are commonly known in the art. For further teaching, TIGIT is also known as "VSIG9", "VSTM3" or "WUCAM". These terms refer to TIGIT nucleic acids, as well as TIGIT peptides, polypeptides and proteins, as the context will allow. By way of example, human TIGIT mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_173799.4, XM_047447671.1 or XM_047447672.1. By way of example, human TIGIT protein is annotated under NCBI Genbank accession numbers NP_776160.2, XP_047303627.1 or XP_047303628. For example, the TIGIT gene is annotated at NCBI Genbank ID:201633.

[0043] Reference to "ICOS" ("inducible T cell costimulatory molecule") refers to ICOS markers, peptides, polypeptides, proteins or nucleic acids, as they are commonly known in the art. For further teaching, ICOS is also known as "AILIM", "CD278" or "CVID1". These terms refer to ICOS nucleic acids, as well as ICOS peptides, polypeptides and proteins, as the context will allow. By way of example, human ICOS mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_012092.4 or XM_047444022.1. By way of example, human ICOS protein is annotated under NCBI Genbank accession numbers NP_036224.1 or XP_047299978.1. By way of example, ICOS gene is annotated under NCBI Genbank ID:29851. Reference to "CLEC4C" ("C-type lectin domain family 4 member C") refers to a CLEC4C marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, CLEC4C is also known as "BDCA-2", "BDCA2", "CD303", "CLECSF11", "CLECSF7", "DLEC", "HECL" or "PRO34150". These terms refer to CLEC4C nucleic acids, as well as CLEC4 peptides, polypeptides and proteins, as the context will allow. By way of example, human CLEC4C mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_001371390.1, NM_001371391.1, NM_130441.3 or NM_203503.2. By way of example, the human CLEC4C protein is annotated under NCBI Genbank accession numbers NP_001358319.1, NP_001358320.1, NP_569708.1 or NP_987099.1. By way of example, the human CLEC4C gene is annotated under NCBI Genbank ID: 170482.

[0044] Reference to "CXCR3" ("CXC motif chemokine receptor 3") refers to the CXCR3 marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, CXCR3 is also known as "CD182", "CD183", "CKR-L2", "CMKAR3", "GPR9", "IP10-R", "Mig-R" or "MigR". These terms refer to CXCR3 nucleic acids, as well as CXCR3 peptides, polypeptides and proteins, as is clear from the context. For example, human CXCR3 mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_001504.2, XM_047442010.1, NM_001142797.2, XM_005262257.4, XM_005262256.4 or XM_017029435.2. For example, human CXCR3 protein is annotated under NCBI Genbank accession numbers NP_001495.1, XP_047297966.1, NP_001136269.1, XP_005262314.1, XP_00526313.1, XP_016884924.1. By way of example, the human CXCR3 gene is annotated at NCBI Genbank ID:2833. Reference to "TLR8" ("Toll-like receptor 8") refers to a TLR8 marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, TLR8 is also known as "CD288", "IMD98", "hTLR8". These terms refer to TLR8 nucleic acids, as well as TLR8 peptides, polypeptides and proteins, as the context will allow. By way of example, human TLR8 mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_016610.4 or NM_138636.5. By way of example, human TLR8 protein is annotated under NCBI Genbank accession numbers NP_057694.2 or NP_619542.1. By way of example, human TLR8 gene is annotated under NCBI Genbank ID:51311.

[0045] Reference to "FCGR2A" ("Fc gamma receptor IIa") refers to the FCGR2A marker, peptide, polypeptide, protein, or nucleic acid, as is commonly known in the art. For further teaching, FCGR2A is also known as "CD32", "CD32A", "CDw32", "FCG2", "FCGR2", FCGR2A1", "FcGR", "FcgammaRIIa" or "IGFR2". These terms refer to FCGR2A nucleic acids, as well as FCGR2A peptides, polypeptides, and proteins, as the context will allow. By way of example, human FCFR2A mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_001136219.3, NM_001375296.1, NM_001375297.1, or NM_021642.5. By way of example, the human FCGR2A protein is annotated under NCBI Genbank accession numbers NP_001129691.1, NP_001362225.1, NP_001362226.1 or NP_067674.2. By way of example, the human FCGR2A gene is annotated under NCBI Genbank ID:2212. Reference to "IDO1" ("indoleamine 2,3-dioxygenase 1") refers to an IDO1 marker, peptide, polypeptide, protein, or nucleic acid, as is commonly known in the art. For further teaching, IDO1 is also known as "IDO", "IDO-1" or "INDO". These terms refer to IDO nucleic acids, as well as IDO peptides, polypeptides, and proteins, as the context may dictate. By way of example, human IDO mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession number NM_002164.6. By way of example, human IDO protein is annotated under NCBI Genbank accession number NP_002155.1. By way of example, human IDO1 gene is annotated under NCBI Genbank ID:3620.

[0046] Reference to "TRAF3IP3" ("TRAF3 interacting protein 3") refers to a TRAF3IP3 marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, TRAF3IP3 is also known as "T3JAM". These terms refer to TRAF3IP3 nucleic acids, as well as TRAF3IP3 peptides, polypeptides and proteins, as the context may dictate. By way of example, human TRAF3IP3 mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_001287754.2, NM_001320143.2, NM_001320144.2, NM_025228.4. By way of example, the human TRAF3IP3 protein is annotated under NCBI Genbank accession numbers NP_001274683.1, NP_001307072.1, NP_001307073.1 or NP_079504.2. By way of example, the human TRAF3IP3 gene is annotated under NCBI Genbank ID:80342. Reference to "PIK3CD" ("phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit delta") refers to a PIK3CD marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, PIK3CD is also known as "APDS", "IMD14", "IMD14A", "IMD14B", "P110DELTA", "PI3K", "ROCHIS" or "p110D". These terms refer to PIK3CD nucleic acids, as well as PIK3CD peptides, polypeptides and proteins, as the context will allow. By way of example, human PIK3CD mRNA is annotated under NCBI Genbank (http: / / www.ncbi.nlm.nig.gov / ) accession numbers NM_001350234.2, NM_001350235.1 or NM_005026.5. For example, the human PIK3CD protein is annotated under NCBI Genbank accession numbers NP_001337163.1, NP_001337164.1, NP_005017.3. For example, the human PIK3CD gene is annotated under NCBI Genbank ID:5293.

[0047] Reference to "IGLV1.36" ("immunoglobulin lambda chain variable region 1-36" refers to an IGLV1.36 marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, ITGAL is also known as "IGLV136", "V1-11". These terms refer to IGLV1.36 nucleic acids, as well as IGLV1.36 peptides, polypeptides and proteins, as the context will allow. By way of example, the human IGLV1.36 gene is annotated at NCBI Genbank ID:28826. Reference to "TRAV14DV4" ("T cell receptor alpha chain variable region 14 / delta chain variable region 4") refers to the TRAV14DV4 marker, peptide, polypeptide, protein or nucleic acid, as such names are commonly known in the art. For further teaching, TRAV14DV4 is also known as "hADV14S1", "TRAV14 / DV4", "TCRAV6S1-hDV104S1". These terms refer to TRAV14DV4 nucleic acids, as well as TRAV14DV4 peptides, polypeptides and proteins, as the context will allow. By way of example, the human TRAV14DV4 gene is annotated at NCBI Genbank ID:28669. Reference to "TRAJ16" (T cell receptor alpha J chain 16) refers to an ITGAL marker, peptide, polypeptide, protein or nucleic acid, as such terms are commonly known in the art. Further teaching, these terms refer to TRAJ16 nucleic acids, as well as TRAJ16 peptides, polypeptides and proteins, as the context may dictate. By way of example, the human TRAJ16 gene is annotated at NCBI Genbank ID:28739.

[0048] Reference to "IGHV1.3" ("immunoglobulin heavy chain variable region 1-3" refers to an IGHV1.3 marker, peptide, polypeptide, protein, or nucleic acid, as such names are commonly known in the art. For further teaching, IGHV1.3 is also known as "IGHV1-3", "IGHV13" or "VI-3B". These terms refer to IGHV1.3 nucleic acids, as well as IGHV1.3 peptides, polypeptides, and proteins, as the context will indicate. By way of example, the human IGHV1.3 gene is annotated at NCBI Genbank ID:28473. Reference to "IGHV4.28" ("immunoglobulin heavy chain variable region 4-28") refers to an IGHV4.28 marker, peptide, polypeptide, protein, or nucleic acid, as such names are commonly known in the art. For further teaching, IGHV4.28 is also known as "IGHV4-28", "IGHV428" or "VH". These terms refer to IGHV4.28 nucleic acids, as well as IGHV4.28 peptides, polypeptides, and proteins, as the context will allow. By way of example, the human IGHV4.28 gene is annotated at NCBI Genbank ID:28400. Reference to "IGHV4.31" ("immunoglobulin heavy chain variable region 4-31") refers to an IGHV4.31 marker, peptide, polypeptide, protein, or nucleic acid, as such names are commonly known in the art. For further teaching, IGHV4.31 is also known as "IGHV4-31" or "IGHV431." These terms refer to IGHV4.31 nucleic acids, as well as IGHV4.31 peptides, polypeptides, and proteins, as the context will allow. By way of example, the human IGHV4.31 gene is annotated at NCBI Genbank ID:28396. Reference to "IGHV4.4" ("immunoglobulin heavy chain variable region 4-4") refers to an IGHV4.4 marker, peptide, polypeptide, protein, or nucleic acid, as such names are commonly known in the art. For further teaching, IGHV4.4 is also known as "IGHV4-4" or "IGHV44." These terms refer to IGHV4.4 nucleic acids, as well as IGHV4.4 peptides, polypeptides, and proteins, as the context will allow. By way of example, the human IGHV4.4 gene is annotated at NCBI Genbank ID:28401.

[0049] One of skill in the art will understand that any sequence represented in the sequence database or herein may be a precursor of a marker, peptide, polypeptide, protein or nucleic acid and may contain portions that are processed away from the mature molecule. Unless otherwise clear from the context, reference herein to any marker, gene, peptide, polypeptide or protein, or fragment thereof, may also generally encompass modified forms of that marker, gene, peptide, polypeptide or protein, or fragment thereof, including genetic alterations such as mutations, or post-expression modifications including, for example, phosphorylation, glycosylation, lipid modification, methylation, cysteinylation, sulfonation, glutathionylation, acetylation, oxidation, and the like. Reference herein to any marker, gene, peptide, polypeptide or protein also includes fragments thereof. Thus, reference herein to measuring (or measuring the amount of), determining the presence of, or determining the expression level of, any one of the markers, genes, peptides, polypeptides or proteins may include measuring, determining the presence of, or determining the expression level of, that marker, gene, peptide, polypeptide or protein, such as measuring any mature and / or processed soluble / secreted forms thereof (e.g., plasma circulating forms) and / or measuring one or more fragments thereof. For example, any marker, gene, peptide, polypeptide or protein, and / or one or more fragments thereof may be measured together, with the amount measured corresponding to the total amount of the species measured together. In a further example, any marker, gene, peptide, polypeptide or protein, and / or one or more fragments thereof may be measured individually.

[0050] The term "fragment" with respect to a peptide, polypeptide or protein generally refers to an N- and / or C-terminal truncated form of the peptide, polypeptide or protein. Preferably, a fragment may comprise at least about 30%, such as at least about 50% or at least about 70%, preferably at least about 80%, such as at least about 85%, more preferably at least about 90%, even more preferably at least about 95%, or about 99% of the amino acid sequence length of the peptide, polypeptide or protein. For example, so long as the length of the full-length peptide, polypeptide or protein is not exceeded, a fragment may comprise a sequence of ≧5 consecutive amino acids, or ≧10 consecutive amino acids, or ≧20 consecutive amino acids, or ≧30 consecutive amino acids, such as ≧40 consecutive amino acids, such as ≧50 consecutive amino acids, for example ≧60, ≧70, ≧80, ≧90, ≧100, ≧200, ≧300 or ≧400 consecutive amino acids of the corresponding full-length peptide, polypeptide or protein. Reference herein to any protein, polypeptide or peptide may also include variants thereof. The term "variant" of a protein, polypeptide or peptide refers to a protein, polypeptide or peptide whose sequence (i.e., amino acid sequence) is substantially identical (i.e., largely but not completely identical) to the sequence of the described protein or polypeptide, such as at least about 80% identical or at least about 85% identical, such as preferably at least about 90% identical, such as at least 91% identical, 92% identical, more preferably at least about 93% identical, such as at least 94% identical, even more preferably at least about 95% identical, such as at least 96% identical, even more preferably at least about 97% identical, such as at least 98% identical, and most preferably at least 99% identical. Preferably, the variant can display the degree of identity with the described protein, polypeptide or peptide when the entire sequence of the described protein, polypeptide or peptide is queried in a sequence alignment (i.e., overall sequence identity).

[0051] A variant of a protein, polypeptide or peptide may be a homolog (e.g., an ortholog or a paralog) of said protein, polypeptide or peptide. As used herein, the term "homology" generally refers to the structural similarity between macromolecules of the same or different taxa, particularly between two proteins or polypeptides, where this similarity is due to shared descent. When the present specification refers to or encompasses fragments and / or variants of a protein, polypeptide or peptide, this preferably refers to variants and / or fragments that are "functional", i.e. that at least partially retain the biological activity or intended functionality of the respective protein, polypeptide or peptide. Preferably, functional fragments and / or variants may retain at least about 20%, such as at least 30%, or at least about 40%, or at least about 50%, such as at least 60%, more preferably at least about 70%, such as at least 80%, even more preferably at least about 85%, even more preferably at least about 90%, and most preferably at least about 95%, or about 100% or more of the intended biological activity or functionality compared to the corresponding protein, polypeptide or peptide.

[0052] In some embodiments, the risk score obtained by any of the methods herein can be further refined with immunohistochemistry or protein expression data from the subject's tumor. For example, the Cytokine 30-Plex Human Panel and immunohistochemistry staining can be used to further qualitatively and / or quantitatively evaluate subject samples.

[0053] The present invention also provides a computer-implemented method for predicting and / or determining the response of a subject diagnosed with cancer to treatment with an immunotherapeutic agent, analysing a sample obtained from said subject, the method comprising: (a) determining at least three, preferably selected from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31 and IGHV4.4, quantified in a sample from the subject suspected of containing or containing tumor cells; or at least four or five, and even more preferably at least ten, biomarkers; (b) normalizing the quantified expression levels, such that normalization is performed by comparing the normalized values ​​of step (b) with data obtained from corresponding evaluations and expression levels of a reference set; (c) classifying whether the normalized values ​​of step (b) exceed a predefined threshold; and (d) calculating using a predictive algorithm or machine learning model to obtain a risk score of said normalized value indicative of the likelihood of the subject's response to the immunotherapeutic agent. In another embodiment, a computer-implemented method is provided for predicting and / or determining whether a subject diagnosed with cancer will respond to treatment with an immunotherapeutic agent, by analyzing a sample obtained from the subject, the sample containing or suspected to contain tumor cells, the method comprising: (a) providing quantified expression levels of biomarkers CD247, LAX1 and IKZF3, the biomarkers being quantified in the subject sample; (b) optionally providing quantified expression levels of one or more biomarkers selected from the group consisting of biomarkers CD3G and ITGAL, and / or CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in the subject sample; (c) normalizing the quantified expression levels, such that normalization is performed by comparing with data obtained from corresponding evaluations and expression levels of a reference set; (d) classifying whether the normalized value of step (c) exceeds a predetermined threshold; and (e) calculating using a predictive algorithm or machine learning model to obtain a risk score of said normalized value indicative of the likelihood of the subject's response to the immunotherapeutic agent.

[0054] In some embodiments, a kit for predicting and / or determining a subject's response to an immunotherapeutic agent is disclosed, which in some embodiments comprises means for measuring the expression levels of at least three, preferably at least four or five, even more preferably at least ten or all of the biomarkers selected from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4 in a sample from the subject. Optionally, the kit further comprises reference values ​​or threshold values ​​for one or more biomarkers selected from the list consisting of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4. In some other embodiments, a kit for predicting and / or determining a subject's response to an immunotherapeutic agent comprises means for measuring the expression levels of the biomarkers CD247, LAX1 and IKZF3 in a sample from a subject containing or suspected of containing tumor cells; optionally, means for measuring the expression levels of one or more biomarkers selected from the list consisting of the biomarkers CD3G and ITGAL, and / or CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in a sample from a subject containing or suspected of containing tumor cells. Optionally, the kit further comprises reference values ​​or threshold values ​​for the biomarkers CD247, LAX1 and IKZF3. The kit may also further comprise reference values ​​or threshold values ​​for the biomarkers CD3G and ITGAL, and / or one or more biomarkers selected from the list consisting of CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4. In yet some further embodiments, the kits of the present application may include ready-to-use substrate solutions, wash solutions, dilution buffers, and instructions. The kits may also include positive and / or negative control samples. In some embodiments, the kits include instructions. Preferably, the instructions included in the kits are clear, concise, and comprehensive to one of skill in the art. Typically, the instructions provide information regarding the contents of the kit, how to collect samples, methods, experimental readouts and their interpretation, as well as cautions and warnings.

[0055] The terms "kit of parts" and "kit" as used throughout this specification refer to an article of manufacture containing components necessary to perform a particular method (e.g., a method of predicting and / or determining a subject's response to an immunotherapeutic agent, a method of determining whether a subject is suitable for immunotherapeutic treatment, or a method of determining whether a subject is likely to respond to immunotherapeutic treatment) and packaged for transport and storage. Materials suitable for packaging the components included in the kit include glass, plastic (e.g., polyethylene, polypropylene, polycarbonate), bottles, flasks, vials, ampoules, paper, packaging materials, or other types of containers, carriers, or supports. When a kit includes multiple components, at least some of the components (e.g., two or more of the multiple components) or all of the components may be physically separated, e.g., contained in separate containers, carriers, or supports. The components included in the kit may or may not be sufficient to perform a particular method, and external reagents or materials may or may not be required to perform the method. Typically, kits are utilized in conjunction with standard laboratory equipment, such as liquid handling equipment, environmental (e.g., temperature) control equipment, analytical equipment, and the like. The kits may include some or all of the following that are useful in a particular method: solvents, buffers (such as, but not limited to, histidine buffer, citrate buffer, succinate buffer, acetate buffer, phosphate buffer, formate buffer, benzoate buffer, TRIS ((tris(hydroxymethyl)aminomethane) buffer or maleate buffer, or mixtures thereof), enzymes (such as, but not limited to, a thermostable DNA polymerase), detectable labels, detection reagents, and control formulations (positive and / or negative).

[0056] Typically, the kits will include instructions for use, such as a printed insert or instructions on a computer readable medium. These terms may be used interchangeably with the term "article of manufacture," which, when used in this context, broadly encompasses any man-made, tangible product of construction. In a specific embodiment, the kit further comprises a computer readable storage medium having recorded thereon one or more programs for performing the methods taught herein. Further provided is the use of the kit, inter alia, for predicting and / or determining a subject's response to an immunotherapeutic agent, or for determining whether a subject is suitable for immunotherapeutic treatment, or for determining whether a subject is likely to respond to immunotherapeutic treatment.

[0057] Further provided is the use of a kit according to any of the embodiments disclosed herein. In one aspect, the use of a kit comprising means for measuring the expression levels of the biomarkers CD247, LAX1 and IKZF3 in a sample from a subject containing or suspected of containing tumor cells is provided in a method for determining and / or predicting the response or resistance of a subject diagnosed with cancer to treatment with an immunotherapeutic agent as taught herein. In a further embodiment, the use of a kit further comprising means for measuring the expression levels of the biomarkers CD3G and ITGAL in a sample from a subject containing or suspected of containing tumor cells is provided. In another embodiment, the use of a kit further comprising means for measuring the expression levels of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in a sample from a subject containing or suspected of containing tumor cells is provided. In some embodiments, there is provided the use of a kit further comprising reference values ​​or threshold values ​​for the biomarkers CD247, LAX1 and IKZF3, and / or reference values ​​or threshold values ​​for the biomarkers CD3G and ITGAL, and / or reference values ​​or threshold values ​​for the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3 and IGHV4.4.

[0058] The inventors of the present application have identified several biomarkers that predict and / or determine a subject's response to an immunotherapeutic agent or determine whether a subject is suitable for treatment with an immunotherapeutic agent. More specifically, they have identified biomarkers that are particularly suitable for predicting a response after treatment with an immune checkpoint inhibitor. Thus, in some embodiments, the immunotherapeutic agent is an immune checkpoint inhibitor, or a combination of several immune checkpoint inhibitors. Furthermore, the immune checkpoint inhibitor is selected from the group consisting of a PD-1 targeting agent, a PD-L1 targeting agent, a CTLA-4 targeting agent, and combinations thereof. In some further preferred embodiments, the immunotherapeutic treatment is a PD-1 targeting agent or a combination of a PD-L1 targeting agent and a CTLA-4 targeting agent. In some embodiments, the PD-1 targeting agent is a monoclonal antibody against PD-1, such as, for example, pembrolizumab, nivolumab, or cemiplimab. In some embodiments, the PD-L1 targeting agent is a monoclonal antibody against PD-L1, such as, for example, atezolizumab, avelumab, durvalumab, nivolumab, or pembrolizumab. In some embodiments, the CTLA-4 targeting agent is a monoclonal antibody against CTLA-1, such as, for example, ipilimumab. In yet some further embodiments, a PDL-1 targeting agent, such as atezolizumab, avelumab, durvalumab, nivolumab, or pembrolizumab, is combined with a CTLA-4 targeting agent, such as ipilimumab.

[0059] Immunotherapeutic drugs, such as immune checkpoint inhibitors, target specific immune cells that need to be activated or inactivated to initiate an immune response. Monoclonal antibodies targeting either PD-1 or PD-L1 can block these proteins present on T cells and normal (or cancer) cells, respectively, and promote an immune response against cancer cells. These drugs are useful for treating different types of cancer, including bladder cancer, non-small cell lung cancer, and Merkel cell skin cancer. CTLA-4 is another protein on some T cells, and like PD-1, it acts as a kind of switch to suppress the immune system. Anti-CTLA4 treatment is often used to treat melanoma of the skin and some other cancers. The PD-1 targeting agent, PD-L1 targeting agent, or CTLA-4 targeting agent may be any targeting agent known in the art, such as an antibody or fragment thereof that recognizes and blocks PD-1, PD-L1, or CTLA-4. In some embodiments, the PD-1 targeting agent is a monoclonal antibody against PD-1, such as, for example, pembrolizumab, nivolumab, or cemiplimab. In some embodiments, the PD-L1 targeting agent is a monoclonal antibody against PD-L1, such as, for example, atezolizumab, avelumab, durvalumab, nivolumab, or pembrolizumab. In some embodiments, the CTLA-4 targeting agent is a monoclonal antibody against CTLA-1, such as, for example, ipilimumab. In yet some further embodiments, a PLD-1 targeting agent, such as atezolizumab, avelumab, durvalumab, nivolumab, or pembrolizumab, is combined with a CTLA-4 targeting agent, such as ipilimumab.

[0060] In a further aspect of the invention, an immunotherapeutic agent, such as an immune checkpoint inhibitor or a combination of checkpoint inhibitors, is provided for use in treating cancer in a subject who has been found to be responsive to treatment with the immunotherapeutic agent using any of the methods of analyzing the expression level of one or more biomarkers in a subject sample as disclosed herein. In particular, the immunotherapeutic agent is provided for use in the treatment of cancer, preferably melanoma, more preferably stage II, stage III or stage IV melanoma, in a subject that has been found to be responsive to treatment with the immunotherapeutic agent by using any method that analyzes the expression level of one or more, preferably at least three, even more preferably at least four or five biomarkers selected from the list consisting of: CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, PIK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4 in a sample from the subject containing or suspected of containing tumor cells. In another embodiment, an immunotherapeutic agent is provided for use in treating cancer, preferably melanoma, in a subject who has been found to respond to treatment with the immunotherapeutic agent by using any method of analyzing expression levels of the biomarkers CD247, LAX1 and IKZF3, optionally in combination with expression levels of one or more biomarkers selected from the group consisting of the biomarkers CD3G and ITGAL, and / or CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4, in a sample from the subject containing or suspected of containing tumor cells.

[0061] Depending on the type of cancer, different ICI treatments are advised. Combination therapy or monotherapy is possible. The set of predictive biomarkers of the present invention is preferably used to determine whether a subject diagnosed with cancer will respond to treatment with an immunotherapeutic agent, in particular an immune checkpoint inhibitor or a combination of checkpoint inhibitors. More preferably, the set of predictive biomarkers of the present invention is used to determine the response of a subject when receiving an immunotherapeutic treatment, such as treatment with an immune checkpoint inhibitor. Most preferably, the set of predictive biomarkers of the present invention is used to determine the response of a subject when receiving treatment with an immune checkpoint inhibitor selected from the group consisting of a PD-1 targeting agent, a PD-L1 targeting agent, a CTLA-4 targeting agent, and a combination thereof. A further aspect of the invention relates to a method of treating a subject diagnosed with cancer with an immunotherapeutic agent, the method comprising determining whether the subject is likely to respond to treatment with an immunotherapeutic agent by any of the methods disclosed herein, and administering the immunotherapeutic agent to the subject if the subject is found to be responsive to immunotherapeutic treatment. In some preferred embodiments, the immunotherapeutic agent is an immune checkpoint inhibitor, such as a PD-1 targeted agent, a PD-L1 targeted agent, or a CTLA-4 targeted agent. In some further preferred embodiments, the immunotherapeutic agent is a combination of different immune checkpoint inhibitors, such as a combination of a PD-1 targeted agent or a PD-L1 targeted agent and a CTLA-4 targeted agent. Another embodiment of the invention provides a tumor response report comprising the steps of isolating one or more patient samples containing or suspected of containing tumor cells; and detecting from the one or more samples a tumor response profile including CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, P, and / or a combination of CD3G and ITGAL, and / or a combination of CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4, and / or a combination of CD247, LAX1, IKZF3, ITGAL, CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, FCGR2A, IDO1, TRAF3IP3, P The method comprises the steps of generating expression data, preferably gene expression data, even more preferably RNA sequencing data, comprising information on a plurality of biomarkers, preferably protein coding genes, in particular a plurality of protein coding genes, selected from IK3CD, IGLV1.36, TRAV14DV4, TRAJ16, IGHV1.3, IGHV4.28, IGHV4.31, IGHV4.4; and / or the protein coding genes CD247, LAX1 and IKZF3; and assessing the likelihood of the patient's tumour responding to a plurality of possible treatments, preferably treatment with an immunotherapeutic agent. In a further embodiment, immunohistochemistry data or other useful patient or tumour data can be added to the tumour response report.

[0062] In some embodiments, the sample is a sample that contains or is suspected to contain tumor cells. In some embodiments, the sample contains tumor cells. In some embodiments, the sample is a tumor tissue sample or a liquid sample containing tumor cells. In some embodiments, the sample is a tumor tissue sample from a subject diagnosed with cancer. In some embodiments, the sample is a liquid sample, such as a blood sample, that contains tumor cells, in particular circulating tumor cells. In some embodiments, the sample is a sample that contains circulating tumor cells. Thus, the expression level of the biomarker is assessed in a biological sample that contains or is suspected to contain cancer cells. The tissue sample may be, for example, a resected tissue, tissue biopsy or metastatic lesion of a patient suffering from, suspected to suffer from or diagnosed with cancer. Preferably, the sample is a tissue sample, resected sample or biopsy sample of a tumor, a known or suspected metastatic cancer lesion or section. Methods for obtaining biological samples, including excised tissues, biopsies and bodily fluids, such as blood samples containing cancer / tumor cells, are well known in the art. In some embodiments, the sample is a tumor tissue sample, such as a fresh frozen tumor tissue sample, a fresh tumor tissue sample, or a formalin-fixed paraffin-embedded (FFPE) tumor tissue sample. Any technique known in the art can be used to isolate nucleic acid, such as RNA or DNA, from the tumor tissue sample. For example, RNA can be isolated from a tumor tissue sample, such as an FFPE tumor tissue sample, using the QIAGEN RNeasy FFPE kit (standard kit) according to the manufacturer's instructions (Qiagen). In some embodiments, the sample is a liquid sample containing tumor cells, particularly circulating tumor cells. In some embodiments, the liquid sample is a blood sample, such as a peripheral blood sample, known or suspected to contain circulating cancer cells. The sample may contain both cancer cells, i.e., tumor cells and non-cancerous cells, and in certain embodiments, both cancerous and non-cancerous cells. In some embodiments, the sample is a sample of isolated circulating tumor cells. Any technique known in the art can be used to isolate circulating tumor cells from a liquid sample, such as a blood sample. For example, the EasySep® Direct human CTC enrichment kit (Stem Cell Technologies) or Qiagen's CTC ready-to-use CTC AdnaTest (Qiagen) can be used to isolate circulating tumor cells. For example, the AllPrep RNA / mRNA Nano kit (Qiagen) or the RNeasy Tissue / Cells Advances mini kit (Qiagen) can be used to isolate RNA and / or DNA from isolated CTCs.

[0063] In some embodiments, samples obtained from a patient are collected prior to the initiation of any immunotherapy or other treatment regimen or therapy, such as chemotherapy or radiation therapy for the treatment of cancer or management or amelioration of its symptoms. Thus, in some embodiments, samples are collected prior to administration of an immunotherapy or other agent or prior to the initiation of an immunotherapy or other treatment regimen. In some embodiments, the subject is a subject diagnosed with cancer. The cancer can be selected from cancer patients diagnosed with bladder cancer, kidney cancer, liver cancer, lung cancer, pancreatic cancer, prostate cancer, thyroid cancer, uterine cancer, ovarian cancer, colon cancer, breast cancer, head and neck cancer, skin cancer; even more preferably, skin cancer or lung cancer, such as melanoma, lung adenocarcinoma, lung squamous cell carcinoma, non-small cell lung cancer (NSCLC). In a preferred embodiment, the subject is a subject diagnosed with melanoma and / or lung cancer. In a further preferred embodiment, the subject is diagnosed with melanoma. In some embodiments, the subject is diagnosed with stage II, stage III, or stage IV melanoma. The present invention preferably describes patients diagnosed with melanoma, who may have already been treated with anti-PD-1, anti-PD-L1 and / or anti-CTLA-4, are undergoing such treatment or are planned to be treated with such agents, and a set of predictive biomarkers is optimized for the patient to predict with high specificity and sensitivity the response of the patient's tumor to such treatment, or the likelihood of progression or early death. In a preferred embodiment, a diverse set of biomarkers of the present invention is used in the development of novel immunotherapeutic agents.

[0064] One aspect of the present invention relates to a method for developing predictive biomarkers for progression when undergoing immunotherapy treatment, the method comprising the steps of: (a) treating an animal model with the immunotherapy treatment; (b) analyzing the expression levels of biomarkers, preferably protein-coding genes, involved in expressing responsiveness to the immunotherapy treatment; and (c) selecting expression-change biomarkers, preferably protein-coding genes, that correlate with a tendency to respond, partially respond or not respond to the immunotherapy treatment. Another aspect of the present invention provides a method for testing biomarkers predictive of responsiveness to immunotherapy treatment, comprising the steps of: (a) treating an animal model with the immunotherapy treatment; (b) analyzing the expression levels of biomarkers, preferably protein-coding genes, involved in expressing responsiveness to the immunotherapy treatment; and (c) selecting expression-varying biomarkers, preferably protein-coding genes, that correlate with the tendency of the animal model to respond, partially respond or not respond to the immunotherapy treatment, thereby identifying said protein-coding genes. Preferably, the animal model is selected from the group consisting of mouse, rat, rabbit, cat, dog, and frog. These animal models can be used to develop and test predictive biomarkers. The preferred animal model is the mouse model. In another preferred embodiment, the biomarkers of the invention are expressed in an animal model. The predictive biomarkers of the invention are expressed in a mouse model. Preferred mouse models are MISTRG, NSG-SGM3, NOG-EXL, BRG hIL-3 hGM-CSF. In another and further embodiment of the invention, the predictive biomarkers are tested in an animal model, preferably a mouse model.

[0065] Some embodiments of the present application relate to the use of the aforementioned methods for predicting the responsiveness or non-responsiveness of a subject to one or more immunotherapeutic treatments, preferably to the use of the aforementioned methods for predicting the responsiveness or non-responsiveness of a subject to anti-PD-1 / PD-L1 monotherapy or in combination with anti-CTLA-4 therapy. Prognostically identifying which patients will experience progression or become non-responders to immunotherapy treatment is of medical and economic interest so that optimal patient-specific therapies can be determined, ensuring better and safer treatment. The present invention further provides a method for determining the response of a tumor to an immunotherapeutic treatment; the method comprises the steps of quantitatively assessing the expression levels of a plurality of biomarkers, preferably protein-coding genes, involved in responsiveness or resistance to the immunotherapeutic treatment; and classifying the tumor as a responder, partial responder or non-responder to the treatment based on this quantitative assessment.

[0066] In some embodiments, the method further comprises determining immunohistochemistry data regarding the tumor. Preferably, immunohistochemistry data is included in the characterization of the tumor and the tumor microenvironment. This redundant data set allows for more confident and valid identification of responder, partial responder or non-responder tumors as repeat measurements produce similar outputs. In a more preferred embodiment, the method further comprises the step of determining one or more possible immunotherapeutic treatment regimens. Next to targeted ICI treatment, other immunotherapy treatment modalities can be employed in the treatment of cancer patients, leading to positive outcomes in the treatment of cancer patients.

[0067] The present application further provides a method of administering an activating or suppressing immunotherapy treatment to a cancer patient if it is determined that the expression level of one or more gene signatures as disclosed herein in the patient is altered. In one embodiment, the method of the present invention comprises notifying the patient of an increased likelihood of responding to the treatment. In another embodiment, the method of the present invention comprises recommending a particular therapeutic treatment to the patient. In another embodiment, the method of the present invention further comprises administering the treatment to the patient if it is determined that the patient may benefit from the treatment, particularly an immunotherapy. In further embodiments, the immunotherapeutic agent comprises a checkpoint inhibitor, a chimeric antigen receptor T cell therapy, an oncolytic vaccine, a cytokine agonist or cytokine antagonist, or a combination thereof, or other immunotherapy available in the art. The activating immunotherapy may further comprise the use of a checkpoint inhibitor. Checkpoint inhibitors are readily available in the art and include, but are not limited to, PD-1 inhibitors, PD-L1 inhibitors, PD-L2 inhibitors, CTLA-4 inhibitors, or a combination thereof. EXAMPLES

[0068] The invention will now be further described with reference to the following non-limiting examples which further illustrate the invention but are not intended, nor should they be construed, to limit the scope of the invention. method Study design, patient cohorts and response assessment A total of 273 patients with metastatic melanoma treated with PD1 / PD-L1 inhibitors or in combination with CTLA4 inhibitors were selected for analysis, including 210 patients from four interventional clinical trials in the GEO and dbGaP databases (Study356761 [doi:10.1016 / j.cell.2017.09.028], 522397 [doi:10.1016 / j.ccell.2019.01.003], phs000452_01 [doi:10.1038 / s41591-019-0654-5], and GSE78220 [doi:10.1016 / j.cell.2016.02.065]) and 63 patients from an ongoing retrospective observational study in three hospitals in Belgium and Austria (NCT04860076). Patients enrolled in this study were treated with nivolumab or pembrolizumab, or in combination with ipilimumab. We complied with all relevant ethical regulations for research involving human participants, and informed consent was obtained from all patients. The protocol was approved by an independent ethics committee at each participating center. Response was assessed by RECIST v1.1 (PD-progression, SD-stable disease, PR-partial response, CR-complete response). Patients who did not reach an evaluable RECIST score as defined by CR / PR / SD / PD were excluded from the study. Patients with PFS < 90 days and OS < 180 were classified as an early death (ED) subgroup. To ensure the translatability of the results, the dbGaP dataset was used as one discovery cohort, and only data from the NCT04860076 clinical trial were treated as the validation cohort.

[0069] RNA extraction and whole transcriptome sequencing For NCT04860076, total RNA was extracted from FFPE blocks using the QIAGEN RNeasy FFPE kit (standard kit) according to the manufacturer's instructions (Qiagen, Hilden, Germany). All FFPE blocks were obtained from biopsies removed before starting treatment with immune checkpoint inhibitors. On average, 5–7 FFPE slices were used for RNA extraction. RNA samples were quantified using Qubit 2.0 Fluorometer RNA assay (Invitrogen, Carlsbad, CA, USA) and RNA integrity was tested with an Agilent 4200 TapeStation (Agilent Technologies, Palo Alto, CA, USA). Samples with a DV200 score below 30% were excluded from further analysis. RNA samples of satisfactory quality were treated with TURBO DNase (Thermo Fisher Scientific, Waltham, MA, USA) to remove DNA according to the manufacturer's protocol. rRNA was removed using the QIAGEN FastSelect rRNA HMR Kit (Qiagen, Hilden, Germany). RNA-seq libraries were prepared using the NEBNext Ultra II RNA Library Preparation Kit for Illumina (NEB, Ipswich, MA, USA) according to the manufacturer's recommendations. Briefly, enriched RNA was fragmented for 15 min at 94°C. First- and second-strand cDNA was then synthesized. cDNA fragments were end-repaired and adenylated at the 3' end, and universal adaptors were ligated to the cDNA fragments, followed by indexing and library enrichment by limited cycle PCR. Sequencing libraries were validated using an Agilent Tapestation 4200 (Agilent Technologies, Palo Alto, CA, USA) and quantified by quantitative PCR (Applied Biosystems, Carlsbad, CA, USA) using a Qubit 2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA). Sequencing libraries were multiplexed and clustered in flow cell lanes and loaded onto an Illumina NovaSeq S4 instrument according to the manufacturer's instructions. Samples were sequenced against a target of 50 million reads in a 2 × 150 paired-end (PE) format. Image analysis and base calling were performed by HiSeq Control Software (HCS). Raw sequencing data (bcl files) generated from the Illumina HiSeq were converted to FASTQ files and demultiplexed using Illumina bcl2fastq2.17 software. One mismatch was allowed for index sequence identification. Similar procedures were used for RNA extraction and sequencing in Studies356761, 522397, phs000452_01, and GSE78220 in the GEO and dbGaP databases [doi:10.1016 / j.cell.2017.09.028, 10.1016 / j.ccell.2019.01.003, 10.1038 / s41591-019-0654-5, 10.1016 / j.cell.2016.02.065].

[0070] RNA-seq data processing and normalization Whole transcriptome profiles were generated for 63 patients from the NCT04860076 study and combined with the whole transcriptome profiles from the dbGaP dataset. A custom internal pipeline was used to process the resulting FASTQ files. Briefly, raw bulk RNA-seq reads were processed using the nf-core / rnaseq pipeline to convert raw RNA reads and obtain gene expression matrices. The workflow included a series of steps for read quality control and read trimming (FastQC, Trim Galore!), read alignment to the human reference genome GRCh38 (STAR), gene-related counts (featureCounts), and result quality control (RSeQC, Qualimap, Preseq, edgeR, MultiQC). The pipeline was adjusted and fine-tuned to best fit the needs of downstream analysis and available resources. Data were normalized across all datasets using ComBat-seq [doi:10.1093 / nargab / lqaa078].

[0071] Principal Component Analysis (PCA) and Principal Variance Component Analysis (PVCA) PCA was performed on the 24,3769 genes found in the dataset using the prcomp() function. PCA analysis was also performed on the MM cohort of the dbGaP and NCT04860076 studies. PVCA is a hybrid method that combines PCA and VCA to quantify the contribution of known variables to the overall variability observed in the transcription data. PVCA analysis was performed using the pvca package (v1.18.0. Bioconductor). PCA was used to check the comparability of post-combat normalized data and to identify potential confounding variables that may affect the analysis.

[0072] Discovering and validating predictive signatures using machine learning An ensemble model using LASSO-based learners was used in the discovery cohort (dbGaP) to identify predictive markers of response to treatment. Ensemble models are known to be stable and insensitive to the initial seed. This regression analysis method simultaneously performs variable selection and regularization to improve accuracy. To create the base learner, 1000 LASSO models were created with different seeds. For each model, hyperparameter selection was performed by 5-fold cross-validation. The final ensemble model was created by taking the average coefficient of the 1000 LASSO models. More specifically, in each LASSO model, each gene is given a coefficient (if a gene is not selected by this model, its coefficient is 0). As a result, the coefficient of a gene in the final ensemble model is obtained by taking the average of its coefficients in each individual LASSO model. The R package glmnet (v2.0-13) was used. A random forest classifier (RFC) was then trained on the discovery cohort using many of the predictive parameters identified by the LASSO model. For the RFC, 75% of the data from the discovery cohort was used as the training set for model training and 25% of the data from the discovery cohort was used for testing. The predictive performance of the biomarker signatures was compared using the CD274 (PD-L1 only) model as the baseline model. Classifiers were trained as binary outcome models using the following combinations of outcome classes: (ED+PD) vs (PR+CR), (ED+PD+PR) vs CR, (ED+PD) vs CR, ED vs CR. The use of different binary outcomes allows us to explore the specific effect of individual markers for each RECIST1.1 status category and to assess the sensitivity of model performance to the class definition. The trained model with the best performing biomarkers was then validated on the validation cohort, the NCT04860076 study data. As mentioned above, the model training procedure was completely agnostic to the validation cohort. After training the model on 75% of the discovery cohort (dbGaP), it was applied "as-is" to generate predicted calls on the NCT04860076 data (validation dataset). The predicted calls were then compared with the ground truth annotations obtained from the hospitals. The ROC AUC metric was used to constantly evaluate the model performance during model training, testing, and validation.

[0073] Bioinformatics analysis After counting the number of reads mapped to exons of annotated genes (Ensembl release 100 [doi:10.1093 / nar / gkaa942]), the gene hits table was used for downstream differential gene expression analysis. The DESeq2 (v1.26.0) [doi:10.1186 / s13059-014-0550-8] R package was used to compare gene expression between user-defined groups of samples. Wald tests were used to generate p-values ​​and log2 fold change values ​​were calculated. P-values ​​were corrected for multiple testing using the Benjamini-Hochberg method. Genes with an adjusted p-value <0.05 were called differentially expressed genes for each comparison. Functional enrichment analysis was performed on the differentially expressed gene set by running overexpression analysis (ORA) from the clusterProfiler (v3.16.1) [doi:10.1016 / j.xinn.2021.100141] R package. Functional annotation was performed against the Gene Ontology [doi:10.1038 / 75556, 10.1093 / nar / gkaa1113] and KEGG [doi:10.1093 / nar / 28.1.27] databases. Boxplots and violin plots were used to visualize log2-transformed normalized gene expression in groups. Scaled heatmaps were used to visualize log2-transformed normalized expression of predicted significant gene sets.

[0074] Statistics and Reproducibility All two-group comparisons of continuous variables used two-tailed Student's t-test (R function t-test) unless otherwise stated. FDR-corrected p-values ​​are reported unless otherwise stated.

[0075] result Study design, patient cohorts and response assessment Across the cohort, 15% of patients were treated with PD1 / L1 in combination with CTLA4 and 85% were treated with PD1 / L1 as monotherapy (Table 1). The distribution of patients expressing a specific type of treatment response (RECIST) was comparable between the discovery and validation sets, except for the PR category (partial response) (Table 2). The discovery dataset contained almost three times as many patients with PR compared to the validation cohort. For the other RECIST1.1 categories, patient distribution was comparable. [Table 1] [Table 2]

[0076] RNA extraction and RNA quality Overall, many RNA samples from the NCT04860076 clinical trial exceeded the quality threshold (DV200>30%) and the extracted RNA amount was sufficient for downstream whole-transcriptome sequencing. However, we observed high levels of DNA contamination that could not be successfully excluded by a single DNAse treatment. Residual DNA contamination caused higher than expected intronic and intergenic read coverage. Nevertheless, we achieved adequate exonic read coverage representative of transcript abundance.

[0077] RNA-seq data processing, normalization and PCA analysis PCA-based investigation of data distribution between discovery and validation cohorts revealed considerable variability along the PC1 axis and partial variability along the PC2 axis between different studies (Figure 1A). Post-COMBAT normalization values ​​significantly reduced study-specific differences and achieved an acceptable level of normalization between different datasets (Figure 1B). Furthermore, statistical investigation of selected biomarkers (see Results below) confirmed an excellent level of concordance in the distribution of log2-normalized expression values ​​between discovery and validation cohorts (Figure 2).

[0078] Using Machine Learning to Discover and Validate Predictive Signatures The LASSO-based learner identified a 25-gene signature, of which 10 genes (TIGIT, PIK3CD, IKZF3, ITGAL, CD247, LAX1, CD3G, CD3E, IGHV1.3, IGHV4.4) constituted the majority of the explanatory power (Table 3), of which an additional 3 genes (CD247, LAX1, and IKZF3) were most predictive in the machine learning model. These markers provided the highest discriminatory power regardless of the type of statistical analysis and group comparison we performed (Figures 3A and 3B). Importantly, the markers were identified in the discovery cohort alone and then investigated for confirmation in the validation cohort (Figures 3C and 3D). CRF model performance in the discovery (test set) and validation sets achieved high discrimination performance between RECIST 1.1 categories with classification ranging from 73% to 78% in the validation cohort for all 25 marker sets (Figure 4A) and 64% to 77% in the validation cohort for the three marker models (Figure 4B) (Table 4). In all cases, the three marker models (CD247, LAX1, and IKZF3) as well as the full model (25 biomarkers) outperformed the baseline model with CD274 (PD-L1 inhibitor only) marker (Figure 4C). The greatest predictive increase in ROCAUC (+15%) was achieved with the addition of three marker models (CD247, LAX1 and IKZF3) and over 22 biomarkers (CD3G, CD3E, CXCR6, SLAMF7, MYO1G, TIGIT, ICOS, CLEC4C, CXCR3, TLR8, ITGAL, FCGR2A, IDO1, TRAF3IP3, PIKCD, IGLV1.36, TRBJ2.7, TRAV14DV4, TRAJ16, IGH1.3, IGH4.28, IGHV4.31, IGHV4.4), resulting in a model with an additional improvement in ROC AUC of approximately 5%. [Table 3] [Table 4]

[0079] Bioinformatics analysis Differential expression analysis revealed that the majority of DEG genes were downregulated in poor-response subgroups. The top biomarkers identified by the LASSO learner ML model were mostly supported in both the discovery and validation cohorts with FDR-corrected significance thresholds above 0.05 (Figures 5A and 5B). All markers were downregulated in the ED, PD, and SD subgroups compared to the PR and CR subgroups. Heat maps of the biomarker panel showed reasonable clustering of patients to the RECICT group in the discovery cohort (Figure 6A) and even more consistent clustering for the validation cohort (Figure 6B).

[0080] References TIFF2024526977000005.tif75158

Claims

1. A method for determining and / or predicting the response or resistance of a subject diagnosed with cancer to treatment with an immunotherapeutic agent, the method comprising analyzing the expression levels of biomarkers CD247, LAX1, and IKZF3 in a sample from the subject that contains or is suspected of containing tumor cells.

2. The method according to claim 1, wherein the immunotherapeutic agent is an immune checkpoint inhibitor or a combination of immune checkpoint inhibitors.

3. The method according to claim 1 or 2, wherein the immunotherapeutic agent is selected from the group consisting of a PD-1 targeting agent, a PD-L1 targeting agent, a CTLA-1 targeting agent, and combinations thereof.

4. The method according to claim 1 or 2, wherein the sample is a tumor tissue sample or a liquid sample containing tumor cells.

5. The method according to claim 4, wherein the sample is a tumor tissue sample.

6. The method according to claim 4, wherein the sample is a blood sample containing circulating tumor cells.

7. The method according to claim 1, further comprising analyzing the expression levels of biomarkers CD3G and ITGAL in the sample from the subject.

8. The method according to claim 1, wherein the expression levels of at least two, preferably at least four, more preferably all of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, and IGHV4.4 are analyzed in the sample from the subject.

9. The method according to claim 1, wherein the expression levels of the biomarkers are compared to corresponding reference values or thresholds specific to subjects having a known response to treatment with the immunotherapeutic agent.

10. The method according to claim 1, wherein a risk score representing the likelihood of the response to the immunotherapeutic agent is obtained based on the expression levels of the biomarkers.

11. The method according to claim 10, wherein the risk score is obtained and calculated using a pre-trained machine learning model and using the expression levels of the biomarkers as input values.

12. The method according to claim 1, wherein the response is selected from the RECIST 1.1 response criteria.

13. In the method according to claim 1, means for measuring the expression levels of the biomarkers CD247, LAX1, and IKZF3 in the sample from the subject; Optionally, means for measuring the expression levels of the biomarkers CD3G and ITGAL in a sample of the subject; Optionally, means for measuring the expression levels of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4 in a sample of the subject; Optionally, a reference value or threshold value of the biomarkers CD247, LAX1 and IKZF3; Optionally, a reference value or threshold value of the biomarkers CD3G and ITGAL; Optionally, a reference value or threshold value of the biomarkers CD3E, TIGIT, PIK3CD, IGHV1.3, IGHV4.4, Use of a kit comprising. [

14. ] A method of treating a subject diagnosed with cancer using an immunotherapeutic agent, comprising determining or predicting the subject's response to the immunotherapeutic agent by the method according to claim 1, and if it is determined that the subject responds to treatment with the immunotherapeutic agent, administering the immunotherapeutic agent to the subject. [

15. ] A computer-implemented method of monitoring, predicting or determining the response of a subject diagnosed with cancer to treatment with an immunotherapeutic agent, (a) providing the quantified expression levels of the biomarkers CD247, LAX1 and IKZF3 in a sample of the subject that contains or is suspected of containing tumor cells; (b) optionally, providing the quantified expression levels of the biomarkers CD3G and ITGAL in a sample of the subject that contains or is suspected of containing tumor cells; (c) normalizing the quantified expression levels such that normalization is performed by comparison with data obtained from corresponding evaluations and expression levels of a reference set; (d) classifying whether the normalized value of step (c) exceeds a predetermined threshold; (d) obtaining a risk score of the normalized value that represents the likelihood of the subject's response to the immunotherapeutic agent, calculated using a pre-trained machine learning model; A method comprising. [

16. ] The method according to claim 15, wherein the immunotherapeutic agent is an immune checkpoint inhibitor or a combination of immune checkpoint inhibitors. [

17. ] The method according to claim 15 or 16, wherein the immunotherapeutic agent is selected from the group consisting of a PD-1 targeting agent, a PD-L1 targeting agent, a CTLA-4 targeting agent and combinations thereof. [

18. ] The method according to claim 17, wherein the sample is a tumor tissue sample or a liquid sample containing tumor cells.

19. The method according to claim 18, wherein the sample is a tumor tissue sample.

20. The method according to claim 19, wherein the sample is a blood sample containing circulating tumor cells.

21. The method according to claim 1, or the use according to claim 13, wherein the biomarker is a protein-coding gene, in particular, the expression level of the biomarker is an expression level based on RNA of the protein-coding gene.

22. The method according to claim 1, or the use according to claim 13, wherein the subject is diagnosed with melanoma; preferably stage II, stage III or stage IV melanoma.

23. The method according to claim 14 or 15, wherein the biomarker is a protein-coding gene, in particular, the expression level of the biomarker is an expression level based on RNA of the protein-coding gene.

24. The method according to claim 14 or 15, wherein the subject is diagnosed with melanoma; preferably stage II, stage III or stage IV melanoma.

Images