WO2023092119A2

WO2023092119A2 - Methods for predicting responsiveness to a cancer therapy

Info

Publication number: WO2023092119A2
Application number: PCT/US2022/080235
Authority: WO
Inventors: Alexandre Harari; Denarda DANGAJ; George Coukos
Original assignee: Ludwig Institute For Cancer Research Ltd
Priority date: 2021-11-22
Filing date: 2022-11-21
Publication date: 2023-05-25
Also published as: WO2023092119A3

Abstract

This disclosure describes novel methods for predicting responsiveness of a patient to treatment of a cancer therapy (e.g., adoptive cell therapy (ACT)) and novel CD8+ T-cell network-based biomarkers that can improve patient selection and guide the design of adoptive cell therapy clinical trials.

Description

METHODS FOR PREDICTING RESPONSIVENESS TO A CANCER THERAPY

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority under 35 U.S.C. §119(e) to U.S. Provisional Patent Application No. 63/281,979, filed November 22, 2021. The foregoing application is incorporated by reference herein in its entirety.

FIELD OF THE INVENTION

This invention relates generally to methods for predicting responsiveness of a patient a cancer therapy, and more specifically to methods for predicting responsiveness of a patient to adoptive cell therapy (ACT).

BACKGROUND OF THE INVENTION

Adoptive cell therapy (ACT) using ex vzvo-expanded autologous tumor-infiltrating lymphocytes (TILs) is a potent strategy, with objective responses seen in a subset of metastatic patients with melanoma in multiple clinical studies. Clinical benefits with TIL-ACT have also been reported in epithelial cancers, including cervical, lung, colorectal, and breast cancer. These data, along with exciting responses seen in hematologic cancers with chimeric antigen receptor T cells, have accounted for an unprecedented development in the ACT field. However, the benefit of TIL-ACT does not extend to all treated patients for reasons that remain unclear to date.

Understanding the steady state immune contexture could help identify biomarkers for the proper selection of patients for TIL-ACT. Such understanding could elucidate key mechanisms supporting T-cell mediated immune rejection but also resistance and help elaborate future therapeutic strategies. However, to date, no study has investigated the tumor mocroenvironment (TME) dynamics during TIL-ACT therapy in tumors.

Therefore, there remains a need for novel approaches for predicting a patient’s clinical response to a cancer therapy, such as adoptive cell therapy.

SUMMARY OF THE INVENTION

This disclosure addresses the need mentioned above in a number of aspects. In one aspect, this disclosure provides a method of predicting responsiveness to a cancer therapy in a subject. In some embodiments, the method comprises: determining an expression level of each of a set of biomarkers in a sample from the subject; determining a change in the expression level of each of the set of biomarkers as compared to a respective reference expression level for each of the set of biomarkers; determining a distribution of changes of expression levels of the set of biomarkers; and assessing a likelihood of a therapeutic response to the cancer therapy by comparing the distribution of changes of expression levels to a reference distribution of changes of expression levels, wherein the reference distribution of changes of expression levels is correlated positively or negatively with the therapeutic response to the cancer therapy, and wherein the reference distribution of changes of expression levels is associated with one or more characteristics in a tumor and/or tumor microenvironment thereof in the subject.

In some embodiments, the one or more characteristics comprising: increased tumor- intrinsic immunogenicity; increased genomic instability; increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ tumor-infiltrating lymphocytes (TILs); increased activation of macrophages or dendritic cells; increased cell-cell interaction; and/or reprogramed myeloid populations and reconstituted antitumor CD8 TIL-myeloid cell networks.

In some embodiments, the method comprises: (i) determining an expression level of each of a set of biomarkers in a sample from the subject; (ii) determining a change in the expression level of each of the set of biomarkers as compared to a respective reference expression level for each of the set of biomarkers; (iii) determining a distribution of changes of expression levels of the set of biomarkers; and (iv) assessing a likelihood of a therapeutic response to the cancer therapy by comparing the distribution of changes of expression levels to a reference distribution of changes of expression levels, wherein the reference distribution of changes of expression levels is correlated positively or negatively with the therapeutic response to the cancer therapy, and wherein the reference distribution of changes of expression levels is associated with one or more characteristics in a tumor and/or tumor microenvironment thereof in the subject, the one or more characteristics comprising: increased tumor-intrinsic immunogenicity; increased genomic instability; increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ tumor-infiltrating lymphocytes (TILs); increased activation of macrophages or dendritic cells; increased cell-cell interaction (or cellular crosstalk); reprogramed myeloid populations and reconstituted antitumor CD8 TIL-myeloid cell networks. In some embodiments, the step of assessing the likelihood of the therapeutic response comprises identifying the subject as having an increased likelihood of the therapeutic response to the cancer therapy if the distribution of changes of expression levels of the set of biomarkers is identical to the reference distribution of changes of expression levels.

In some embodiments, the reference distribution of changes of expression levels is determined from samples of one or more subjects who have responded positively to the cancer therapy.

In some embodiments, the method comprises determining the reference distribution of changes of expression levels by: (a) determining an expression level of each of a plurality of biomarkers in each of the samples from a plurality of subjects who have responded positively to the cancer therapy; (b) determining whether the determined expression level is increased, decreased, or unchanged as compared to a reference value for each of the plurality of biomarkers to provide a biomarker expression profile of each of the samples; (c) performing an aggregated analysis on the biomarker expression profiles of the samples; (d) identifying a group of biomarkers having increased or decreased expression levels as compared to the reference value; and (e) determining a reference distribution of changes of expression levels of the group of biomarkers.

In some embodiments, the group of biomarkers are associated with the one or more characteristics in tumor microenvironments in the plurality of subjects.

In some embodiments, the plurality of subjects have been administered the cancer therapy.

In some embodiments, the respective reference expression level is determined from samples of one or more subjects who have not been administered the cancer therapy.

In some embodiments, the change in the expression level of each of the set of biomarkers is an increase or decrease in the expression level. In some embodiments, the distribution of changes comprises an increase or decrease in expression levels of the set of biomarkers.

In some embodiments, the increased tumor-intrinsic immunogenicity or genomic instability is characterized by increased copy-number variation.

In some embodiments, the increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ tumor-infiltrating lymphocytes (TILs) is characterized by overexpressed genes of tissue residence and tumor reactivity, exhaustion, activation (HLA class-II genes), DNA repair, recruitment chemokines, adhesion to endothelium, or effector molecules.

In some embodiments, the increased activation of macrophages or dendritic cells is characterized by overexpressed genes and pathways for activation of complement, interferon signaling, IFN-inducible T-cell recruiting chemokines, class-II antigen presentation and processing, or CD28 costimulation.

In some embodiments, the increased cell-cell interaction comprises increased myeloid: T cell interaction, increased B cell: T cell interaction, increased dendritic cell: T cell interaction, or increase CD4 cell: CD8 cell interaction.

In some embodiments, the set of biomarkers comprises a first group of biomarkers expressed in a first population of cells and a second group of biomarkers expressed in a second population of cells, and wherein the first population of cells interact with the second population of cells.

In some embodiments, the first population of cells comprises myeloid cells, B cells, CD4 cells, or dendritic cells. In some embodiments, the first population of cells comprises myeloid cells. In some embodiments, the second population of cells comprises T cells or CD8 cells. In some embodiments, the second population of cells comprises T cells. In some embodiments, the T cells comprise progenitor exhausted T cells or CD8+ tumor infiltrating lymphocytes.

In some embodiments, the first population of cells comprises myeloid cells, and the second population of cells comprises T cells. In some embodiments, the first population of cells comprises dendritic cells, and the second population of cells comprises T cells. In some embodiments, the first population of cells comprises B cells, and the second population of cells comprises T cells. In some embodiments, the first population of cells comprises CD4 cells, and the second population of cells comprises CD8 cells.

In some embodiments, the set of biomarkers comprises a first group of biomarkers associated with a first signaling pathway and a second group of biomarkers associated with a second signaling pathway.

In some embodiments, the reprogramed myeloid populations and reconstituted antitumor CD8 TIL-myeloid cell networks are characterized by increased number of progenitor exhausted T cells, increased number of CD8+ TILs, increased number of CD4 CXCL13 TILs, increased number of CXCL9+ macrophages, increased number of type-I IFN macrophages, or maintained number of CD8+/PD1+/GZMB+/- tumor reactive and polyfunctional TILs.

In some embodiments, the likelihood of the therapeutic response in the subject comprises complete or partial response as defined by response evaluation criteria in solid tumors (RECIST), stable disease as defined by RECIST, or long-term survival in spite of disease progression or response as defined by immune-related response criteria (irRC).

In some embodiments, the cancer cell therapy comprises a cancer immunotherapy. In some embodiments, the cancer therapy comprises an immune cell therapy. In some embodiments, the immune cell therapy comprises a T cell. In some embodiments, the immune cell therapy comprises a tumor infiltrating lymphocyte. In some embodiments, the cancer therapy comprises an adoptive cell therapy (ACT). In some embodiments, the adoptive cell therapy comprises a T-cell receptor (TCR) T cell therapy or a chimeric antigen receptor (CAR) T cell therapy.

In some embodiments, the set of biomarkers comprise one or more biomarkers set forth in Tables 1-5.

In some embodiments, the set of biomarkers comprises:

(a) one or more differentially expressed genes in malignant cells selected from: B2M, SERAC1, HLA-C, OLA1, PSMB9, IFIT3, NCSTN, GBP3, TRIM69, ARSA, TAPI, HLA-A, SEPTIN8, HLA-E, MAN1C1, ANK2, Clorfl98, AL136295.2, EPAS1, APOL1, HTRA2, PSMB8, TMEM62, SEC63, LGALS3BP, TSEN54, and AC009228.1;

(b) one or more differentially expressed genes in CD8 T cells selected from: CXCL13, DUSP4, RGS1, CD8A, VCAM1, NKG7, LYST, TNFAIP3, CTLA4, MT-ATP8, CD7, TNFRSF9, HLA-DRB5, HLA-DPA1, CST7, CCL4L2, CD74, HLA-DRB1, TTN, HAVCR2, HLA-DQA1, CBLB, PMAIP1, PRF1, RNF19A, HLA-DRA, JUN, CD8B, BHLHE40, CD27, BRD2, CMC1, HLA-DPB1, CCL4, CCL5, and MTRNR2L12;

(c) one or more differentially expressed genes in macrophages selected from: IFI27, C1QB, C1QA, CCL4L2, C1QC, IFITM3, FCGR3A, STAT1, CCL3L1, HLA-C, SERPING1, LY6E, IFI6, GBP1, HLA-DQA2, PSAP, B2M, HLA-DQA1, CXCL10, VAMP5, IFITM1, PLAAT4, CTSC, LGALS3BP, CXCL9, APOCI, PSME2, APOE, HLA-DRB5, HSPA8, HLA-B, WARS, GBP4, C3, NCF1, RPS4Y1, IER2, FN1, RPS21, RPS29, YBX1, and RPS2; or

(d) one or more differentially expressed genes in dendritic cells selected from: AREG, CXCR4, ARL4C, JUNB, FOSB, IRF1, LDLRAD4, STAT1, TSPYL2, IRF7, FAM118A, ISG20, MX1, FOS, AKAP13, TXN, TCL1A, PLAC8, RGS1, GZMB, IRF4, NEAT1, NR4A3, GPR183, JCHAIN, ITM2C, ZC3HAV1, PLD4, RANBP2, LILRA4, KLF6, JUN, PDE4B, AC004687.1, SELL, ICAM1, HLA-DQB1, UCP2, WARS, HLA-B, HLA-C, HLA-E, NBPF14, PLEK, HLA- DQA2, HLA-DQA1, SNHG5, SNX3, HLA-DPB1, RPL36A, CYBA, FGL2, ITGB2, RPS20, LYZ, and CST3.

In some embodiments, the set of biomarkers comprises one or more biomarkers selected from:

(i) TOX, PKM, PRF1, LYST, TNFRSF9, ITM2A, GAPDH, PARK7, HAVCR2, CTLA4, PDCD1, SLA, CBLB, RGS1, KLRC2, STAT3, PHLDA1, GNLY, PTPN6, SH2D2A, GZMB, CD7, IFNG, CYTOR, SUB1, VCAM1, RBPJ, NPM1, APOBEC3C, EIF4A1, TPI1, MIF, LAG3, SAMSN1, DUSP4, CXCL13, ARPC1B, DYNLL1, ATP5MC2, CSNK2B, RPL12, SRGN, S100A4, CTSD, and FXYD5;

(ii) CD8A, PRF1, HLA-DQA1, CD7, CXCL13, TNFRSF9, HAVCR2, CST7, LYST, NKG7, BHLHE40, CD8B, PMAIP1, CTLA4, CD27, HLA-DPA1, TTN, VC AMI, HLA-DRA, RGS1, CBLB, HLA-DRB5, DUSP4, HLA-DPB1, CD74, CCL4L2, HLA-DRB1, BRD2, CMC1, MT-ATP8, RNF19A, TNFAIP3, JUN, CCL4, RPS4Y1, CCL5, and RPS26;

(iii) GZMK, AHNAK, IL32, CCL4, FOS, TSC22D3, CD52, GZMM, TXNIP, SEPTIN9, DNAJB1, ANXA1, LTB, SPOCK2, CD48, WIPF1, EMP3, ITM2C, CCNH, KLRG1, THEMIS, AO AH, PTPRC, TC2N, VIM, KLF6, ZFP36L2, CNN2, CYBA, CD69, SELPLG, LIME1, BIN2, SLC2A3, TRAT1, MBP, LCP1, KLRK1, TAPBP, ITGAL, LINC02446, TUBA4A, GZMH, KRT86, DDIT4, and SKAP1; or

(iv) MTRNR2L8, CD52, ANXA1, ZFP36L2, S100A10, VIM, BTG1, DUSP2, RPS29, GPR183, RPS2, LTB, EMP3, PLP2, RPL38, S100A4, IL7R, MTRNR2L12, SLC2A3, AHNAK, TAGLN2, CD44, RPL17, RPS21, TXNIP, FXYD5, TC2N, RPL27A, RPL39, AL138963.4, S100A11, EML4, ANXA2, HSPA1A, HSPA1B, DUSP1, DNAJB1, HSPE1, and DDIT4 In some embodiments, the sample is obtained from neoplasia tissue, tumor microenvironment, or tumor-infiltrating immune cells. In some embodiments, the sample comprises a biological sample that comprises a plasma sample, a blood sample, or a tissue sample. In some embodiments, the sample is obtained from a primary tumor or a metastasis. In some embodiments, the sample comprises immune cells. In some embodiments, the immune cells are selected from T cells, macrophages, dendritic cells, fibroblasts, NK cells, NKT cells, and NK-DC cells. In some embodiments, the sample comprises protein, DNA, or RNA.

In some embodiments, the expression level comprises a mRNA or protein level.

In some embodiments, the mRNA level is determined by at least one technique selected from reverse transcription polymerase chain reaction (RT-PCR), competitive RT-PCR, real-time RT-PCR, ribonucleic acid sequencing (RNA-seq), immunohistochemistry (IHC), immunofluorescence, RNase protection assay (RPA), northern blotting, and DNA chip.

In some embodiments, the protein level is determined by at least one technique selected from western blot, enzyme linked immunosorbent assay (ELISA), radioimmunoassay (RIA), radioimmunodiffusion, Ouchterlony immunodiffusion, rocket immunoelectrophoresis, immunohistochemical staining, immunoprecipitation assay, complement fixation assay, fluorescence activated cell sorter (FACS), and protein chip.

In some embodiments, the subject has a cancer. In some embodiments, the cancer is a carcinoma, a sarcoma, a lymphoma, a melanoma, a pediatric tumor, or a leukemia.

In some embodiments, the cancer is selected from adrenal gland tumors, biliary cancer, bladder cancer, brain cancer, breast cancer, carcinoma, central or peripheral nervous system tissue cancer, cervical cancer, colon cancer, endocrine or neuroendocrine cancer or hematopoietic cancer, esophageal cancer, fibroma, gastrointestinal cancer, glioma, head and neck cancer, Li-Fraumeni tumors, liver cancer, lung cancer, lymphoma, melanoma, meningioma, multiple neuroendocrine type I and type II tumors, nasopharyngeal cancer, oral cancer, oropharyngeal cancer, osteogenic sarcoma tumors, ovarian cancer, pancreatic cancer, pancreatic islet cell cancer, parathyroid cancer, pheochromocytoma, pituitary tumors, prostate cancer, rectal cancer, renal cancer, respiratory cancer, sarcoma, skin cancer, stomach cancer, testicular cancer, thyroid cancer, tracheal cancer, urogenital cancer, and uterine cancer. In another aspect, this disclosure also provides a method of treating cancer in a patient in need thereof with a cancer therapy. In some embodiments, the method comprises: selecting a patient who is likely responsive to treatment of the cancer therapy according to a method described herein; and administering to the patient the cancer therapy.

In some embodiments, the cancer therapy comprises an adoptive cell therapy (ACT). In some embodiments, the adoptive cell therapy comprises a T-cell receptor (TCR) T cell therapy or a chimeric antigen receptor (CAR) T cell therapy.

In some embodiments, the cancer is a carcinoma, a sarcoma, a lymphoma, a melanoma, a pediatric tumor, or a leukemia.

The foregoing summary is not intended to define every aspect of the disclosure, and additional aspects are described in other sections, such as the following detailed description. The entire document is intended to be related as a unified disclosure, and it should be understood that all combinations of features described herein are contemplated, even if the combinations of features are not found together in the same sentence, or paragraph, or section of this document. Other features and advantages of the invention will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples, while indicating specific embodiments of the disclosure, are given by way of illustration only, because various changes and modifications within the spirit and scope of the disclosure will become apparent to those skilled in the art from this detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS

Figs. 1A and IB show differential gene expression (Fig. 1A) and transcription factor/regulon (Fig. IB) analysis between CD8 T cells from responders (Rs) versus non-responders (NRs).

Fig. 2 shows differential gene expression analysis between macrophages from Rs versus NRs. The x-axis displays the log fold-change as computed at single-cell level while the y-axis shows the log fold-change as computed in patient-averaged data.

Fig. 3 shows differential gene expression analysis between dendritic cells from Rs versus NRs. The x-axis displays the log fold-change as computed at single-cell level while the y-axis shows the log fold-change as computed in patient-averaged data. Figs. 4A and 4B show that the TME of TIL- ACT responders is characterized by high levels of myeloid:T cell interaction in contrast to non-responders. Fig. 4A shows a heatmap displaying the number of significant ligand-receptor interactions. Fig. 4B shows a heatmap displaying selected significant pairs of ligand-receptor.

Fig. 5 shows characterization of the TME characteristics and progression post TIL-ACT. Number of significant (unadjusted - value < 0.05) ligand-receptor pair interaction according to the indicated main cell types categorized into five different pathways and split by clinical response and by time when the biopsies were taken (TO and T30).

DETAILED DESCRIPTION OF THE INVENTION

This disclosure describes novel methods for predicting responsiveness of a patient to treatment of a cancer therapy (e.g., adoptive cell therapy (ACT)) and novel CD8⁺ T-cell networkbased biomarkers that can improve patient selection and guide the design of adoptive cell therapy clinical trials. This disclosure is based, at least in part, on an unexpected discovery that responders of an adoptive cell therapy have higher tumor cell-intrinsic immunogenicity, endogenous CD8⁺ TILs and myeloid cells characterized by increased cytotoxicity, exhaustion, and costimulation, and type-I IFN signaling, rich baseline intratumoral and stromal tumor-reactive T-cell networks with activated myeloid populations, and/or reprogrammed myeloid compartments and increased TIL- myeloid networks.

Methods of Predicting Responsiveness to a Cancer Therapy

Accordingly, in one aspect, this disclosure provides a method for predicting or determining responsiveness (or sensitivity or susceptibility) of a subject to a cancer therapy e.g., an adoptive cell therapy) or a clinical outcome of a cancer therapy in a subject, and/or for assessing a prognosis of a patient with a malignant disease.

In some embodiments, the method may include: (i) determining an expression level of each of a set of biomarkers in a sample from the subject; (ii) determining a change in the expression level of each of the set of biomarkers as compared to a respective reference expression level for each of the set of biomarkers; (iii) determining a distribution of changes of expression levels of the set of biomarkers; and (iv) assessing a likelihood of a therapeutic response to the cancer therapy by comparing the distribution of changes of expression levels to a reference distribution of changes of expression levels. In some embodiments, the reference distribution of changes of expression levels is correlated positively or negatively with the therapeutic response to the cancer therapy.

In some embodiments, the reference distribution of changes of expression levels is associated with one or more characteristics in a tumor and/or tumor microenvironment thereof in the subject. In some embodiments, the one or more characteristics may include: increased tumor- intrinsic immunogenicity; increased genomic instability; increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ tumor-infiltrating lymphocytes (TILs); increased activation of macrophages or dendritic cells; increased cell-cell interaction (or cellular crosstalk); and/or reprogramed myeloid populations and reconstituted antitumor CD8 TIL-myeloid cell networks.

In some embodiments, the step of assessing the likelihood of the therapeutic response may include identifying the subject as having an increased likelihood of the therapeutic response to the cancer therapy if the distribution of changes of expression levels of the set of biomarkers is identical to the reference distribution of changes of expression levels.

In some embodiments, the reference distribution of changes of expression levels is determined from samples of one or more subjects who have responded positively to the cancer therapy. In some embodiments, the reference distribution of changes of expression levels is determined from samples of one or more subjects who have responded positively to the cancer therapy, wherein such samples are obtained from the subjects prior to or after treatment of the cancer therapy. In some embodiments, the reference distribution of changes of expression levels is determined from samples of one or more subjects who have responded positively to the cancer therapy, wherein such samples are obtained from the subjects prior to treatment of the cancer therapy.

In some embodiments, the method may include determining the reference distribution of changes of expression levels by: (a) determining an expression level of each of a plurality of biomarkers in each of the samples from a plurality of subjects who have responded positively to the cancer therapy; (b) determining whether the determined expression level is increased, decreased, or unchanged as compared to a reference value for each of the plurality of biomarkers to provide a biomarker expression profile of each of the samples; (c) performing an aggregated analysis on the biomarker expression profiles of the samples; (d) identifying a group of biomarkers having increased or decreased expression levels as compared to the reference value; and (e) determining a reference distribution of changes of expression levels of the group of biomarkers.

In some embodiments, the change in the expression level of each of the set of biomarkers comprises an increase or decrease in the expression level. In some embodiments, the distribution of changes may include an increase or decrease in expression levels of the set of biomarkers.

The terms “predicting,” “prediction,” or “predictive,” as used herein, refers to an advance declaration, indication, or foretelling of a response or reaction to a therapy (e.g., adoptive cell therapy) in a subject not (yet) having been treated with the therapy. For example, a prediction of responsiveness (or sensitivity or susceptibility) to a cancer therapy in a subject may indicate that the subject will respond or react to the cancer therapy, for example, within a certain time period, e.g., so that the subject will have a clinical benefit from the cancer therapy. A prediction of unresponsiveness (or insensitivity or insusceptibility) to a cancer therapy in a subject may indicate that the subject will minimally or not respond or react to the cancer therapy, for example, within a certain time period, e.g., so that the subject will have no clinical benefit from the cancer therapy.

The terms “responsiveness,” “sensitivity,” or “susceptibility” may be used interchangeably herein and refer to the quality that predisposes a subject having a neoplastic disease to be responsive or reactive to a cancer therapy (e.g., adoptive cell therapy). A subject is “responsive,” “sensitive,” or “susceptible” (which terms are used interchangeably) to immunotherapy (i.e., treatment with an adoptive cell therapy), in particular a subject “responds positively,” if the subject will have a clinical benefit from the treatment. A neoplastic tissue, including a tumor, is “responsive,” “sensitive,” or “susceptible” to a cancer therapy if the proliferation rate of the neoplastic tissue is inhibited as a result of contact with the cancer therapy, compared to the proliferation rate of the neoplastic tissue in the absence of contact with the cancer therapy, e.g., treatment with an adoptive cell therapy. The terms “unresponsiveness,” “insensitivity,” “insusceptibility,” or “resistance” may be used interchangeably herein and refer to the quality that predisposes a subject having a neoplastic disease (e.g., cancer) to a minimal (e.g., insignificant) or no response to a cancer therapy (e.g., adoptive cell therapy). A subject is “unresponsive,” “insensitive,” “unsusceptible,” or “resistant” (which terms are used interchangeably) to a cancer therapy (i.e., treatment with a cancer therapy), in particular a subject “responds negatively,” if the subject will have no clinical benefit from the treatment. A neoplastic tissue, including a tumor, is “unresponsive,” “insensitive,” “unsusceptible,” or “resistant” to a cancer therapy if the proliferation rate of the neoplastic tissue is not inhibited or inhibited to a very low e.g., therapeutically insignificant) degree, as a result of treatment of the cancer therapy, compared to the proliferation rate of the neoplastic tissue in the absence of treatment the cancer therapy. The methods as disclosed herein may allow making a prediction that a subject having a neoplastic disease will be responsive to a cancer therapy or will be unresponsive to the cancer therapy. This may, in some embodiments, include predicting that a subject having a neoplastic disease will have a comparatively low probability (e.g, less than 50%, less than 40%, less than 30%, less than 20% or less than 10%>) of being responsive to a cancer therapy; or that a subject having a neoplastic disease will have a comparatively high probability (e.g, at least 50%, at least 60%, at least 70%, at least 80%) or at least 90%) of being responsive to the cancer therapy.

The terms “determining responsiveness,” “predicting responsiveness,” and “assessing a likelihood of a therapeutic response” may be used interchangeably herein.

The term “prognosis,” as used herein, refers to anticipation of progression of a disease (e.g. , cancer) or condition and prospect (e.g., the probability, duration, and/or extent) of recovery. A good prognosis of the diseases or conditions may generally encompass anticipation of a satisfactory partial or complete recovery from the diseases or conditions, such as within an acceptable time period. A good prognosis of such may more commonly encompass anticipation of not further worsening or aggravating within a given time period. A poor prognosis of the diseases or conditions as taught herein may generally encompass anticipation of a substandard recovery and/or unsatisfactorily slow recovery, or to substantially no recovery or even further worsening of such.

In some embodiments, a therapeutic response may include an anti-tumor response when referring to a cancer patient treated with a cancer therapy, such as an adoptive cell therapy (e.g., TIL- ACT). For example, an anti -turn or response may include at least one positive therapeutic effect, such as a reduced number of cancer cells, reduced tumor size, reduced rate of cancer cell infiltration into peripheral organs, reduced rate of tumor metastasis or tumor growth, or progression-free survival. Positive therapeutic effects in cancer can be measured in a number of ways (see, e.g., W. A. Weber, J. Null. Med. 5O: 1S-1OS (2009); Eisenhauer et al., 2009 European Journal of Cancer, 45: 228-247). In some embodiments, an anti -tumor response to a cancer therapy is assessed using RECIST 1.1 criteria, bidimensional irRC, or unidimensional irRC. In some embodiments, an antitumor response is any of stable disease (SD), partial response (PR), complete response (CR), progression-free survival (PFS), and disease-free survival (DFS). In some embodiments, one or more biomarkers of this disclosure predict whether a subject with a solid tumor is likely to achieve a complete response or a partial response.

The disclosed method may be used to predict or determine the likelihood of a complete response or partial response, or whether a response is likely to be a complete response or a partial response. As used herein, a “complete response” or “CR” to a therapy refers to disappearance of all detectable signs of cancer in response to treatment. As used herein, a “partial response” to a therapy refers to a decrease in tumor load in an individual, for example, in terms of tumor number, size, and growth rate, and or an increase in the time of disease progression.

In some embodiments, the likelihood of the therapeutic response in the subject may include complete or partial response as defined by response evaluation criteria in solid tumors (RECIST 1 .0 criteria (Therasse P. et al., 2000 J. Natl Cancer Inst 92:2015-16)), stable disease (SD) as defined by RECIST, or long-term survival in spite of disease progression or response as defined by immune-related response criteria (irRC).

The term “biomarker,” as used herein, refers to an indicator, e.g., predictive, diagnostic, and/or prognostic, which can be detected in a sample. The biomarker may serve as an indicator of a particular subtype of a disease or disorder (e.g., cancer) characterized by certain molecular, pathological, histological, and/or clinical features, and/or may serve as an indicator of a particular cell type or state (e.g., epithelial, mesenchymal, etc.) and/or response to therapy. Biomarkers include, but are not limited to, polynucleotides (e.g., DNA and/or RNA), polynucleotide copy number alterations (e.g., DNA copy numbers), polypeptides, polypeptide and polynucleotide modifications (e.g., posttranslational modifications), carbohydrates, and/or glycolipid-based molecular markers. A biomarker may be present in a sample obtained from a subject before the onset of a physiological or pathophysiological state (e.g., primary cancer, metastatic cancer, etc.), including a symptom thereof (e.g., response to therapy). Thus, the presence of the biomarker in a sample obtained from the subject can be indicative of an increased risk that the subject will develop the physiological or pathophysiological state or symptom thereof. Alternatively and/or additionally, the biomarker may be normally expressed in an individual, but its expression may change (i.e., it is increased (upregulated; over-expressed) or decreased (downregulated; underexpressed)) before the onset of a physiological or pathophysiological state, including a symptom thereof. Thus, a change in the level of the biomarker may be indicative of an increased risk that the subject will develop the physiological or pathophysiological state or symptom thereof. Alternatively, or in addition, a change in the level of a biomarker may reflect a change in a particular physiological or pathophysiological state, or symptom thereof, in a subject, thereby allowing the nature (e.g., severity) of the physiological or pathophysiological state, or symptom thereof, to be tracked over a period of time.

In some embodiments, a level of a biomarker may include the concentration of the biomarker, the expression level of the biomarker, or the activity of the biomarker.

The terms “expression level” and “level of expression” are used interchangeably and generally refer to the amount of a biomarker in a sample. “Expression” refers to the process by which information (e.g., gene-encoded and/or epigenetic) is converted into the structures present and operating in the cell. Therefore, as used herein, “expression” may refer to transcription into a polynucleotide, translation into a polypeptide, or even polynucleotide and/or polypeptide modifications (e.g., posttranslational modification of a polypeptide). Fragments of a transcribed polynucleotide, a translated polypeptide, or polynucleotide and/or polypeptide modifications (e.g. , post-translational modification of a polypeptide) shall also be regarded as expressed whether they originate from a transcript generated by alternative splicing or a degraded transcript, or from a post-translational processing of the polypeptide, e.g, by proteolysis. The term “expression” with respect to a gene sequence refers to transcription of the gene to produce a RNA transcript (e.g, mRNA, antisense RNA, siRNA, shRNA, miRNA, etc.) and, in some embodiments, translation of a resulting mRNA transcript to a protein. Thus, expression of a coding sequence may result from transcription and translation of the coding sequence. Conversely, expression of a non-coding sequence results from the transcription of the non-coding sequence. The terms “biomarker signature,” “signature,” “biomarker expression signature,” or “expression signature” are used interchangeably herein and refer to one or a combination of biomarkers whose expression is an indicator, e.g, predictive, diagnostic, and/or prognostic. The biomarker signature may serve as an indicator of a particular subtype of a disease or disorder (e.g, primary cancer, metastatic cancer, etc.) or symptom thereof (e.g., response to therapy, drug resistance, and/or disease burden) characterized by certain molecular, pathological, histological, and/or clinical features. In some embodiments, the biomarker signature is a “gene signature.” The term “gene signature” is used interchangeably with “gene expression signature” and refers to one or a combination of polynucleotides whose expression is an indicator, e.g., predictive, diagnostic, and/or prognostic. In some embodiments, the biomarker signature is a “protein signature.” The term “protein signature” is used interchangeably with “protein expression signature” and refers to one or a combination of polypeptides whose expression is an indicator, e.g., predictive, diagnostic, and/or prognostic.

In some embodiments, a biomarker signature may include at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, or 100 or more biomarkers.

The term “distribution of changes” or “pattern of changes,” as used herein with reference to biomarkers, refers to an aggregated analysis of changes in biomarker levels (e.g., expression level, concentration, amount, activity) of each biomarker in a set of selected biomarkers relative to a corresponding reference biomarker level. The set of biomarkers may include at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, or 100 or more biomarkers. As will be understood by a person skilled in the art, a reference level or reference value for each biomarker may be the same or different. Such a distribution of changes or a pattern of changes in biomarker levels may include increased levels (e.g., expression levels) for a first group of biomarkers and/or decreased levels (e.g., expression levels) for a second group of biomarkers. The changes in biomarkers levels for the first and second groups of biomarkers collectively constitute a distribution of changes or a pattern of changes. In some embodiments, a distribution of changes or a pattern of changes in biomarker levels may include ratios of levels between two or more biomarkers. As used herein, the terms “increase,” “elevate,” “elevated,” “enhance,” and “activate” all generally refer to an increase by a statically significant amount as compared to a reference level (e.g., a reference expression level). For the avoidance of any doubt, these terms mean an increase of at least 5% (e.g., 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 100%) as compared to a reference level, for example, an increase of at least about 10%, at least about 20%, at least about 30%, at least about 40%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90% or at least about 100%, as compared to a reference level. In some embodiments, these terms may refer to an increase of 10-20%, 10-30%, 10-40%, 10-50%, 10-60%, 10-70%, 10-80%, 10-90%, 10-100%, 10-110%, 10-120%, 10-130%, 10-140%, 10-150%, 10-160%, 10-170%, 10-180%, 10-190%, 10-200%, 10- 210%, 10-220%, 10-230%, 10-240%, 10-250%, 10-260%, 10-270%, 10-280%, 10-290%, or 10- 300%, as compared to a reference level. In some embodiments, these terms may refer to an increase of 10-300%, 20-300%, 30-300%, 40-300%, 50-300%, 60-300%, 70-300%, 80-300%, 90-300%, 100-300%, 110-300%, 120-300%, 130-300%, 140-300%, 150-300%, 160-300%, 170-300%, 180- 300%, 190-300%, 200-300%, 210-300%, 220-300%, 230-300%, 240-300%, 250-300%, 260- 300%, 270-300%, 280-300%, or 290-300% as compared to a reference level. In some embodiments, these terms may refer to an increase of at least 2-fold, at least 3 -fold, at least 4-fold, at least 5-fold, at least 10-fold or greater, as compared to a reference level.

An increased level (e.g., expression level) of a biomarker as compared to a predetermined reference value can be an increase of at least 5% (e.g., 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 100%) as compared to a predetermined reference value, for example, an increase of at least about 10%, at least about 20%, at least about 30%, at least about 40%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, at least about 100%, or any increase from 10% to 100%, as compared to a predetermined reference value; or at least a 2-fold, at least a 3 -fold, at least a 4-fold, at least a 5-fold or at least a 10-fold increase, or any increase from 2-fold to 10-fold or greater, as compared to a predetermined reference value.

In some embodiments, an increased level of a biomarker as compared to a predetermined reference value can be an increase of 10-20%, 10-30%, 10-40%, 10-50%, 10-60%, 10-70%, 10- 80%, 10-90%, 10-100%, 10-110%, 10-120%, 10-130%, 10-140%, 10-150%, 10-160%, 10-170%, 10-180%, 10-190%, 10-200%, 10-210%, 10-220%, 10-230%, 10-240%, 10-250%, 10-260%, 10- 270%, 10-280%, 10-290%, or 10-300% as compared to a predetermined reference value. In some embodiments, an increased level of a biomarker as compared to a predetermined reference value can be an increase of 10-300%, 20-300%, 30-300%, 40-300%, 50-300%, 60-300%, 70-300%, 80- 300%, 90-300%, 100-300%, 110-300%, 120-300%, 130-300%, 140-300%, 150-300%, 160-300%, 170-300%, 180-300%, 190-300%, 200-300%, 210-300%, 220-300%, 230-300%, 240-300%, 250- 300%, 260-300%, 270-300%, 280-300%, or 290-300% as compared to a predetermined reference value. In some embodiments, an increased level of a biomarker as compared to a predetermined reference value can be an increase of 10-20%, 20-30%, 30-40%, 40-50%, 50-60%, 60-70%, 70- 80%, 80-90%, 90-100%, 100-110%, 110-120%, 120-130%, 130-140%, 140-150%, 150-160%, 160-170%, 170-180%, 180-190%, 190-200%, 200-210%, 210-220%, 220-230%, 230-240%, 240- 250%, 250-260%, 260-270%, 270-280%, 280-290%, or 290-300% as compared to a predetermined reference value.

As used herein, the terms “decrease,” “reduce,” and “inhibit” all generally refer to a decrease by a statistically significant amount. However, for avoidance of doubt, the term “reduced,” “decrease,” “reduce,” or “inhibit” means a decrease by at least 5% (e.g., 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 100%) as compared to a reference level, for example, a decrease by at least about 10%, a decrease by at least about 20%, or at least about 30%, or at least about 40%, or at least about 50%, or at least about 60%, or at least about 70%, or at least about 80%, or at least about 90% or up to and including a 100% decrease (e.g., absent level as compared to a reference sample), or any decrease of 10-100% as compared to a reference level. In some embodiments, these terms refer to a decrease of 10-20%, 10-30%, 10-40%, 10-50%, 10-60%, 10-70%, 10-80%, 10-90%, 10-100%, 10-110%, 10-120%, 10- 130%, 10-140%, 10-150%, 10-160%, 10-170%, 10-180%, 10-190%, 10-200%, 10-210%, 10- 220%, 10-230%, 10-240%, 10-250%, 10-260%, 10-270%, 10-280%, 10-290%, or 10-300%, as compared to a reference level. In some embodiments, these terms refer to a decrease of 10-300%, 20-300%, 30-300%, 40-300%, 50-300%, 60-300%, 70-300%, 80-300%, 90-300%, 100-300%, 110-300%, 120-300%, 130-300%, 140-300%, 150-300%, 160-300%, 170-300%, 180-300%, 190- 300%, 200-300%, 210-300%, 220-300%, 230-300%, 240-300%, 250-300%, 260-300%, 270- 300%, 280-300%, or 290-300%, as compared to a reference level. In some embodiments, these terms refer to a decrease of 10-20%, 20-30%, 30-40%, 40-50%, 50-60%, 60-70%, 70-80%, 80- 90%, 90-100%, 100-110%, 110-120%, 120-130%, 130-140%, 140-150%, 150-160%, 160-170%, 170-180%, 180-190%, 190-200%, 200-210%, 210-220%, 220-230%, 230-240%, 240-250%, 250- 260%, 260-270%, 270-280%, 280-290%, or 290-300%, as compared to a reference level.

As used herein, the term “higher” with reference to a biomarker measurement refers to a statistically significant and measurable difference in the level of a biomarker measurement compared to the level of another biomarker or to a control level where the biomarker measurement is greater than the level of the other biomarker or the control level. The difference may be of at least about 10%, at least about 20%, at least about 30%, at least about 40%, or at least about 50%.

As used herein, the term “lower” with reference to a biomarker measurement refers to a statistically significant and measurable difference in the level of a biomarker measurement compared to the level of another biomarker or to a control level where the biomarker measurement is less than the level of the other biomarker or the control level. The difference may be of at least about 10%, at least about 20%, at least about 30%, at least about 40%, or at least about 50%.

The terms “reference level,” “reference value, “control level,” “control value,” “predetermined value,” and “predetermined level” are used interchangeably herein. The terms “reference sample,” “reference cell,” “reference tissue,” “control sample,” “control cell,” and “control tissue are used interchangeably herein.

In some embodiments, a reference level or a control level of biomarkers may be determined from a reference sample, reference cell, reference tissue, control sample, control cell, or control tissue that is obtained from a healthy and/or non-diseased part of the body (e.g., tissue or cells) of the same subject or individual, but at different time-points, e.g., before and after therapy. In some embodiments, a reference level or a control level of biomarkers may be determined from a reference sample, reference cell, reference tissue, control sample, control cell, or control tissue that is obtained from a healthy individual who is not the subject or individual being assessed. In some examples, a reference sample, reference cell, reference tissue, control sample, control cell, or control tissue is or may include a functional T-cell, a dysfunctional T-cell (e.g., an exhausted T- cell), T-cells from a subject that is responsive or sensitive to therapy or T-cells from a subject that is non-responsive or resistant to therapy. In some embodiments, T-cells may include CD8+ T-cells. In some embodiments, T-cells may include CD8+ T-cells from a subject that is non-responsive or resistant to a cancer therapy.

In some embodiments, the expression level may include a mRNA or protein level. As used herein, the “amount” or “level” of a biomarker is a detectable level or amount in a sample. These can be measured by methods known to one skilled in the art. These terms encompass a quantitative amount or level (e.g., weight or moles), a semi -quantitative amount or level, a relative amount or level (e.g., weight % or mole % within class), a concentration, and the like. Thus, these terms encompass absolute or relative amounts or levels or concentrations of a biomarker in a sample. The expression level or amount of biomarker assessed can be used to determine the response to treatment.

In some embodiments, the mRNA level is determined by at least one technique selected from reverse transcription polymerase chain reaction (RT-PCR), competitive RT-PCR, real-time RT-PCR, ribonucleic acid sequencing (RNA-seq), immunohistochemistry (IHC), immunofluorescence, RNaseprotection assay (RPA), northern blotting, and DNA chip.

In some embodiments, the protein level is determined by at least one technique selected from western blot, enzyme-linked immunosorbent assay (ELISA), radioimmunoassay (RIA), radioimmunodiffusion, Ouchterlony immunodiffusion, rocket immunoelectrophoresis, immunohistochemical staining, immunoprecipitation assay, complement fixation assay, fluorescence-activated cell sorter (FACS), and protein chip.

The terms “sample” or “biological sample,” as used herein, include any biological specimen obtained (isolated, removed) from a subject. Samples may include, without limitation, organ tissue (e.g., primary or metastatic tumor tissue), whole blood, plasma, serum, whole blood cells, red blood cells, white blood cells (e.g., peripheral blood mononuclear cells), saliva, urine, stool (feces), tears, sweat, sebum, nipple aspirate, ductal lavage, tumor exudates, synovial fluid, cerebrospinal fluid, lymph, fine needle aspirate, amniotic fluid, any other bodily fluid, exudate or secretory fluid, cell lysates, cellular secretion products, inflammation fluid, semen, and vaginal secretions. In some embodiments, a sample may be readily obtainable by non-invasive or minimally invasive methods, such as blood collection (“liquid biopsy”), urine collection, feces collection, tissue (e.g., tumor tissue) biopsy or fine-needle aspiration, allowing the provision/removal/isolation of the sample from a subject. The term “tissue,” as used herein, encompasses all types of cells of the body, including cells of organs but also including blood and other body fluids recited above. The tissue may be healthy or affected by pathological alterations, e.g., tumor tissue. The tissue may be from a living subject or may be cadaveric tissue. In some embodiments, useful samples are those known to comprise, expected or predicted to comprise, known to potentially comprise, or expected or predicted to potentially comprise tumor cells.

The biological sample may be any sample in which the methylation level of the relevant gene(s) can be determined. In some embodiments, the biological sample is a neoplastic tissue sample, such as a tumor sample, e.g., a primary or metastatic tumor sample. The biological sample may also be derived from a biological fluid or body fluid, for example, whole blood, blood, urine, lymph fluid, serum, plasma, nipple aspirate, ductal fluid, and tumor exudate. It has been shown in the literature that cancer or tumor cells often release genomic DNA in circulating or other bodily fluids. Since said genomic DNA has the same methylation profile as the DNA inside the tumor or cancer cell, said methylation profile can be detected in the circulating or other bodily fluid sample. This has, for example, been reviewed by Qureshi et al., 2010 (Int. J. Surgery 2010, 8: 194-198), hereby incorporated by reference in its entirety. In some embodiments, the sample is a body fluid comprising neoplastic cells.

A sample can be obtained from a subject in any way typically used in clinical settings for obtaining a sample comprising the required cells or nucleic acid, including RNA, genomic DNA, mitochondrial DNA, and protein-associated nucleic acids. For example, the sample can be obtained from fresh, frozen, or paraffin-embedded surgical samples or biopsies of an organ or tissue comprising the suitable cells or nucleic acid to be tested. If desired, the sample can be mixed with a fluid or purified or amplified or otherwise treated. For examples, samples may be treated in one or more purification steps in order to increase the purity of the desired cells or nucleic acid in the sample, or they may be examined without any purification steps. Any nucleic acid specimen in purified or non-purified form obtained from such sample can be utilized in the methods as taught herein.

In some embodiments, the sample is obtained from neoplasia tissue, tumor microenvironment, or tumor-infiltrating immune cells. In some embodiments, the sample may include a biological sample that may include a plasma sample, a blood sample, or a tissue sample. In some embodiments, the sample is obtained from a primary tumor or a metastasis. In some embodiments, the sample may include immune cells. In some embodiments, the immune cells are selected from T cells, macrophages, dendritic cells, fibroblasts, NK cells, NKT cells, and NK-DC cells. In some embodiments, the sample may include protein, DNA, or RNA. In some embodiments, a subject having an increased likelihood of being responsive to a cancer therapy (e.g., adoptive cell therapy) may have increased tumor-intrinsic immunogenicity or genomic instability. In some embodiments, increased tumor-intrinsic immunogenicity or genomic instability is characterized by increased copy-number variation. As used herein, the term “copy number variation” refers to a variation in the number of copies of a nucleic acid sequence present in a test sample as compared to the number of copies of the nucleic acid sequence present in a reference sample.

In some embodiments, a subject having an increased likelihood of being responsive to a cancer therapy (e.g., adoptive cell therapy) may have increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ tumor-infiltrating lymphocytes (TILs). In some embodiments, increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ tumor-infiltrating lymphocytes (TILs) is characterized by overexpressed genes of tissue residence and tumor reactivity, exhaustion, activation (HLA class-II genes), DNA repair, recruitment chemokines, adhesion to endothelium, or effector molecules.

In some embodiments, a subject having an increased likelihood of being responsive to a cancer therapy (e.g., adoptive cell therapy) may have increased activation of macrophages or dendritic cells. In some embodiments, increased activation of macrophages or dendritic cells is characterized by overexpressed genes and pathways for activation of complement, interferon signaling, IFN-inducible T-cell recruiting chemokines, class-II antigen presentation and processing, or CD28 costimulation.

In some embodiments, a subject having an increased likelihood of being responsive to a cancer therapy (e.g., adoptive cell therapy) may have increased cell-cell interaction (e.g., cellular crosstalk). For example, a subject having an increased likelihood of being responsive to a cancer therapy (e.g., adoptive cell therapy) may have myeloid: T cell doublets, dendritic cell (DC): T cell doublets, B cell: T cell doublets, or CD4: CD8 doublets.

In some embodiments, increased cell-cell interaction comprises increased myeloid: T cell interaction, increased B cell: T cell interaction, increased dendritic cell: T cell interaction, or increase CD4 cell: CD8 cell interaction.

In some embodiments, the set of biomarkers may include a first group of biomarkers expressed in a first population of cells and a second group of biomarkers expressed in a second population of cells, and wherein the first population of cells interact with the second population of cells.

In some embodiments, the first population of cells may include myeloid cells, B cells, CD4 cells, or dendritic cells. In some embodiments, the first population of cells may include myeloid cells. In some embodiments, the second population of cells may include T cells or CD8 cells. In some embodiments, the second population of cells may include T cells. In some embodiments, the T cells may include progenitor-exhausted T cells or CD8+ TILs.

In some embodiments, the first population of cells may include myeloid cells, and the second population of cells may include T cells. In some embodiments, the first population of cells comprise dendritic cells, and the second population of cells comprise T cells. In some embodiments, the first population of cells may include B cells, and the second population of cells may include T cells. In some embodiments, the first population of cells may include CD4 cells, and the second population of cells may include CD8 cells.

In some embodiments, a subject having an increased likelihood of being responsive to a cancer therapy (e.g., adoptive cell therapy) may have increased ligand-receptor interactions. In some embodiments, a ligand and a receptor that interact may be involved in and be part of separate signaling pathways. In some embodiments, the set of biomarkers may include a first group of biomarkers associated with a first signaling pathway and a second group of biomarkers associated with a second signaling pathway.

In some embodiments, the set of biomarkers may include: one or more differentially expressed genes in malignant cells selected from: B2M, SERAC1, HLA-C, OLA1, PSMB9, IFIT3, NCSTN, GBP3, TRIM69, ARSA, TAPI, HLA-A, SEPTIN8, HLA-E, MAN1C1, ANK2, Clorfl98, AL136295.2, EPAS1, APOL1, HTRA2, PSMB8, TMEM62, SEC63, LGALS3BP, TSEN54, and AC009228.1.

In some embodiments, the set of biomarkers may include: one or more differentially expressed genes in CD8 T cells selected from: CXCL13, DUSP4, RGS1, CD8A, VCAM1, NKG7, LYST, TNFAIP3, CTLA4, MT-ATP8, CD7, TNFRSF9, HLA-DRB5, HLA-DPA1, CST7, CCL4L2, CD74, HLA-DRB1, TTN, HAVCR2, HLA-DQA1, CBLB, PMAIP1, PRF1, RNF19A, HLA-DRA, JUN, CD8B, BHLHE40, CD27, BRD2, CMC1, HLA-DPB1, CCL4, CCL5, and MTRNR2L12. In some embodiments, the set of biomarkers may include: one or more differentially expressed genes in macrophages selected from: IFI27, C1QB, C1QA, CCL4L2, C1QC, IFITM3, FCGR3A, STAT1, CCL3L1, HLA-C, SERPING1, LY6E, IFI6, GBP1, HLA-DQA2, PSAP, B2M, HLA-DQA1, CXCL10, VAMP5, IFITM1, PLAAT4, CTSC, LGALS3BP, CXCL9, APOCI, PSME2, APOE, HLA-DRB5, HSPA8, HLA-B, WARS, GBP4, C3, NCF1, RPS4Y1, IER2, FN1, RPS21, RPS29, YBX1, and RPS2.

In some embodiments, the set of biomarkers may include: one or more differentially expressed genes in dendritic cells selected from: AREG, CXCR4, ARL4C, JUNB, FOSB, IRF1, LDLRAD4, STAT1, TSPYL2, IRF7, FAM118A, ISG20, MX1, FOS, AKAP13, TXN, TCL1A, PLAC8, RGS1, GZMB, IRF4, NEAT1, NR4A3, GPR183, JCHAIN, ITM2C, ZC3HAV1, PLD4, RANBP2, LILRA4, KLF6, JUN, PDE4B, AC004687.1, SELL, ICAM1, HLA-DQB1, UCP2, WARS, HLA-B, HLA-C, HLA-E, NBPF14, PLEK, HLA-DQA2, HLA-DQA1, SNHG5, SNX3, HLA-DPB1, RPL36A, CYBA, FGL2, ITGB2, RPS20, LYZ, and CST3.

In some embodiments, the reprogramed myeloid populations and reconstituted antitumor CD8 TIL-myeloid cell networks are characterized by increased number of progenitor exhausted T cells, increased number of CD8+ TILs, increased number of CD4 CXCL13 TILs, increased number of CXCL9+ macrophages, increased number of type-I IFN macrophages, and/or maintained number of CD8+/PD1+/GZMB+/- tumor-reactive and polyfunctional TILs.

In some embodiments, the set of biomarkers may include one or more biomarkers set forth in Tables 1-5.

In some embodiments, the set of biomarkers may include one or more biomarkers selected from: TOX, PKM, PRF1, LYST, TNFRSF9, ITM2A, GAPDH, PARK7, HAVCR2, CTLA4, PDCD1, SLA, CBLB, RGS1, KLRC2, STAT3, PHLDA1, GNLY, PTPN6, SH2D2A, GZMB, CD7, IFNG, CYTOR, SUB1, VCAM1, RBPJ, NPM1, APOBEC3C, EIF4A1, TPI1, MIF, LAG3, SAMSN1, DUSP4, CXCL13, ARPC1B, DYNLL1, ATP5MC2, CSNK2B, RPL12, SRGN, S100A4, CTSD, and FXYD5.

In some embodiments, the set of biomarkers may include one or more biomarkers selected from: CD8A, PRF1, HLA-DQA1, CD7, CXCL13, TNFRSF9, HAVCR2, CST7, LYST, NKG7, BHLHE40, CD8B, PMAIP1, CTLA4, CD27, HLA-DPA1, TTN, VCAM1, HLA-DRA, RGS1, CBLB, HLA-DRB5, DUSP4, HLA-DPB1, CD74, CCL4L2, HLA-DRB1, BRD2, CMC1, MT- ATP8, RNF19A, TNFAIP3, JUN, CCL4, RPS4Y1, CCL5, and RPS26.

In some embodiments, the set of biomarkers may include one or more biomarkers selected from: GZMK, AHNAK, IL32, CCL4, FOS, TSC22D3, CD52, GZMM, TXNIP, SEPTIN9, DNAJB1, ANXA1, LTB, SPOCK2, CD48, WIPF1, EMP3, ITM2C, CCNH, KLRG1, THEMIS, AO AH, PTPRC, TC2N, VIM, KLF6, ZFP36L2, CNN2, CYBA, CD69, SELPLG, LIME1, BIN2, SLC2A3, TRAT1, MBP, LCP1, KLRK1, TAPBP, ITGAL, LINC02446, TUBA4A, GZMH, KRT86, DDIT4, and SKAP1.

In some embodiments, the set of biomarkers may include one or more biomarkers selected from: MTRNR2L8, CD52, ANXA1, ZFP36L2, S100A10, VIM, BTG1, DUSP2, RPS29, GPR183, RPS2, LTB, EMP3, PLP2, RPL38, S100A4, IL7R, MTRNR2L12, SLC2A3, AHNAK, TAGLN2, CD44, RPL17, RPS21, TXNIP, FXYD5, TC2N, RPL27A, RPL39, AL138963.4, S100A11, EML4, ANXA2, HSPA1A, HSPA1B, DUSP1, DNAJB1, HSPE1, and DDIT4.

The terms “patient,” “subject,” “host,” and “individual” are used interchangeably herein and refer to any subject, particularly a vertebrate subject, and even more particularly a mammalian subject, for whom therapy or prophylaxis is desired. Suitable vertebrate animals that fall within the scope of the invention include, but are not restricted to, any member of the subphylum Chordata including primates (e.g., humans, monkeys and apes, and includes species of monkeys such from the genus Macaca (e.g., cynomologus monkeys such as Macaca fascicularis, and/or rhesus monkeys ( Macaca mulatta )) and baboon ( Papio ursinus), as well as marmosets (species from the genus Callithrix), squirrel monkeys (species from the genus Saimiri ) and tamarins (species from the genus Saguinus), as well as species of apes such as chimpanzees ( Pan troglodytes)), rodents (e.g., mice rats, guinea pigs), lagomorphs (e.g., rabbits, hares), bovines (e.g., cattle), ovines (e.g., sheep), caprines (e.g, goats), porcines (e.g, pigs), equines (e.g., horses), canines (e.g., dogs), felines (e.g., cats), avians (e.g., chickens, turkeys, ducks, geese, companion birds such as canaries, budgerigars etc.), marine mammals (e.g., dolphins, whales), reptiles (snakes, frogs, lizards etc.), and fish. A preferred subject is a human with cancer.

In some embodiments, the subject has a cancer. In some embodiments, the cancer is a carcinoma, a sarcoma, a lymphoma, a melanoma, a pediatric tumor, or a leukemia. In some embodiments, the cancer is selected from adrenal gland tumors, biliary cancer, bladder cancer, brain cancer, breast cancer, carcinoma, central or peripheral nervous system tissue cancer, cervical cancer, colon cancer, endocrine or neuroendocrine cancer or hematopoietic cancer, esophageal cancer, fibroma, gastrointestinal cancer, glioma, head and neck cancer, Li-Fraumeni tumors, liver cancer, lung cancer, lymphoma, melanoma, meningioma, multiple neuroendocrine type I and type II tumors, nasopharyngeal cancer, oral cancer, oropharyngeal cancer, osteogenic sarcoma tumors, ovarian cancer, pancreatic cancer, pancreatic islet cell cancer, parathyroid cancer, pheochromocytoma, pituitary tumors, prostate cancer, rectal cancer, renal cancer, respiratory cancer, sarcoma, skin cancer, stomach cancer, testicular cancer, thyroid cancer, tracheal cancer, urogenital cancer, and uterine cancer.

Methods of Treatment

In another aspect, this disclosure also provides a method of treating cancer in a patient in need thereof with a cancer therapy. In some embodiments, the method may include: selecting a patient who is likely responsive to treatment of the cancer therapy according to the method described herein; and administering to the patient the cancer therapy.

As used herein, the term “treatment” refers to a clinical intervention designed to alter the natural course of the individual or cell being treated during the course of clinical pathology. Desirable effects of treatment include decreasing the rate of disease progression, ameliorating or palliating the disease state, and remission or improved prognosis. For example, an individual is successfully “treated” if one or more symptoms associated with a cancer are mitigated or eliminated, including, but are not limited to, reducing the proliferation of (or destroying) cancerous cells, reducing pathogen infection, decreasing symptoms resulting from the disease, increasing the quality of life of those suffering from the disease, decreasing the dose of other medications required to treat the disease, and/or prolonging survival of individuals. The phrase “treatment with a therapy,” “treating with a therapy,” “treatment with an agent,” “treating with an agent,” and the like refers to the administration of an effective amount of a therapy or agent, including a cancer therapy and optionally an agent, (e.g., a cytotoxic agent or an immunotherapeutic agent) to a patient, or the concurrent administration of two or more therapies or agents, including cancer therapies or agents, e.g., two or more agents selected from cytotoxic agents and immunotherapeutic agents) in effective amounts to a patient. In some embodiments, the cancer cell therapy may include a cancer immunotherapy. In some embodiments, the cancer therapy may include an immune cell therapy. In some embodiments, the immune cell therapy may include a T cell. In some embodiments, the immune cell therapy may include a tumor infiltrating lymphocyte. In some embodiments, the cancer therapy may include an adoptive cell therapy. In some embodiments, the adoptive cell therapy may include a T-cell receptor (TCR) T cell therapy or a chimeric antigen receptor (CAR) T cell therapy.

In some embodiments, the disclosed methods include administration of an adoptive cell therapy. As used herein, the term “adoptive cell therapy,” “ACT,” or “adoptive immunotherapy” are used interchangeably and refer to the administration of a modified immune cell to a subject with cancer. An “immune cell” (also interchangeably referred to herein as an “immune effector cell”) refers to a cell that is part of a subject’s immune system and helps to fight cancer in the body of a subject. Non-limiting examples of immune cells for use in the disclosed methods include T cells, tumor-infiltrating lymphocytes, and natural killer (NK) T cells. The immune cells may be autologous or heterologous to the subject undergoing therapy.

As used herein, the terms “T cell” and “T lymphocyte” are used interchangeably. T cells include thymocytes, naive T lymphocytes, immature T lymphocytes, mature T lymphocytes, resting T lymphocytes, or activated T lymphocytes. A T cell can be a T helper (Th) cell, for example, a T helper 1 (Thl) or a T helper 2 (Th2) cell. The T cell can be a helper T cell (HTL; CD4⁺ T cell) CD4⁺ T cell, a cytotoxic T cell (CTL; CD8⁺ T cell), a tumor-infiltrating cytotoxic T cell (TIL; CD8⁺ T cell), CD4⁺CD8⁺ T cell, or any other subset of T cells. Other illustrative populations of T cells suitable for use in particular embodiments include naive T cells and memory T cells. Also included are “natural killer T (NKT) cells” or “NKT cells,” which refer to a specialized population of T cells that express a semi-invariant ab T cell receptor, but also express a variety of molecular markers that are typically associated with NK cells, such as NK1.1. NKT cells include NK1.1⁺ and NK1. G, as well as CD4⁺, CD4, CD8⁺, and CD8 cells.

The TCR on NKT cells is unique in that it recognizes glycolipid antigens presented by the MHC Llike molecule CD Id. NKT cells can have either protective or deleterious effects due to their ability to produce cytokines that promote either inflammation or immune tolerance. Also included are”gamma-delta T cells (y5 T cells),” which refer to a specialized population that to a small subset of T cells possessing a distinct TCR on their surface, and unlike the majority of T cells in which the TCR is composed of two glycoprotein chains designated a- and b-TCR chains, the TCR in y6 T cells is made up of a g- chain and a d-chain. y6 T cells can play a role in immunosurveillance and immunoregulation and were found to be an important source of IL- 17 and to induce robust CD8⁺ cytotoxic T cell response. Also included are “regulatory T cells” or “Tregs,” which refer to T cells that suppress an abnormal or excessive immune response and play a role in immune tolerance. Tregs are typically transcription factor Foxp3 -positive CD4⁺ T cells and can also include transcription factor Foxp3 -negative regulatory T cells that are IL-10- producing CD4⁺ T cells.

T cells can be obtained from a number of sources, including peripheral blood mononuclear cells, bone marrow, lymph nodes tissue, cord blood, thymus issue, tissue from a site of infection, ascites, pleural effusion, spleen tissue, and tumors. In some embodiments, T cells can be obtained from a unit of blood collected from the subject using any number of techniques known to the skilled person, such as FICOLL separation. In one embodiment, T cells from the circulating blood of an individual are obtained by apheresis. The apheresis product typically contains lymphocytes, including T cells, monocytes, granulocyte, B cells, other nucleated white blood cells, red blood cells, and platelets.

The disclosed immune effector cells, such as T cells, can be genetically modified (forming modified immune cells) following isolation using known methods, or the immune cells can be activated and expanded, or differentiated in the case of progenitors, in vitro prior to being genetically modified. In some embodiments, immune effector cells, such as T cells, are genetically modified with the TCRs or CARs described herein (e.g., transduced with a viral vector comprising a nucleic acid encoding a TCR or a CAR) and then may be activated and expanded in vitro. Techniques for activating and expanding T cells are known in the art and suitable for use with the disclosed technology. See, e.g., US 6,905,874; US 6,867,041; US 6,797,514; WO 2012079000; US 2016/0175358. TCR-expressing or CAR-expressing immune effector cells suitable for use in the disclosed methods may be prepared according to known techniques described in the art.

For use in the disclosed methods, the immune cells may be modified with a TCR or a CAR against a TAA. In other words, non-limiting examples of adoptive cell therapy for use in the disclosed methods include a modified TCR against a tumor-associated antigen (TAA), or a chimeric antigen receptor (CAR) against a TAA.

The TAA may be from any cancer including, but not limited to, adrenal gland tumors, biliary cancer, bladder cancer, brain cancer, breast cancer, carcinoma, central or peripheral nervous system tissue cancer, cervical cancer, colon cancer, endocrine or neuroendocrine cancer or hematopoietic cancer, esophageal cancer, fibroma, gastrointestinal cancer, glioma, head and neck cancer, Li-Fraumeni tumors, liver cancer, lung cancer, lymphoma, melanoma, meningioma, neuroendocrine type I or type II tumors, multiple myeloma, myelodysplastic syndromes, myeloproliferative diseases, nasopharyngeal cancer, oral cancer, oropharyngeal cancer, osteogenic sarcoma tumors, ovarian cancer, pancreatic cancer, pancreatic islet cell cancer, parathyroid cancer, pheochromocytoma, pituitary tumor, prostate cancer, rectal cancer, renal cancer, respiratory cancer, sarcoma, skin cancer, stomach cancer, testicular cancer, thyroid cancer, tracheal cancer, urogenital cancer, or uterine cancer.

In some embodiments, the TAA is selected from AFP, ALK, BAGE proteins, BCMA, BIRC5 (survivin), BIRC7, P-catenin, brc-abl, BRCA1, BORIS, CA9, carbonic anhydrase IX, caspase-8, CALR, CCR5, CD19, CD20 (MS4A1), CD22, CD30, CD40, CDK4, CEA, CTLA4, cyclin-Bl, CYP1B1, EGFR, EGFRvIII, ErbB2/Her2, ErbB3, ErbB4, ETV6-AML, EpCAM, EphA2, Fra-1, FOLR1, GAGE proteins (e.g., GAGE-1, -2), GD2, GD3, GloboH, glypican-3, GM3, gplOO, Her2, HLA/B-raf, HLA/k-ras, HLA/MAGE-A3, hTERT, LMP2, MAGE proteins (e.g., MAGE-1, -2, -3, -4, -6, and -12), MART-1, mesothelin, ML-IAP, Mucl, Muc2, Muc3, Muc4, Muc5, Mucl6 (CA-125), MUM1, NA17, NY-BR1, NY-BR62, NY-BR85, NY-ESO1, 0X40, p 15, p53, PAP, PAX3, PAX5, PCTA-1, PLAC1, PRLR, PRAME, PSMA (FOLH1), RAGE proteins, Ras, RGS5, Rho, SART-1, SART-3, STEAP1, STEAP2, TAG-72, TGF-p, TMPRSS2, Thompson- nouvelle antigen (Tn), TRP-1, TRP-2, tyrosinase, or uroplakin-3.

As used herein, a “T cell receptor” refers to an isolated TCR polypeptide that binds specifically to a TAA, or a TCR expressed on an isolated immune cell e.g., a T cell). TCRs bind to epitopes on small antigenic determinants (for example, comprised in a tumor associated antigen) on the surface of antigen-presenting cells that are associated with a major histocompatibility complex (MHC; in mice) or human leukocyte antigen (HLA; in humans) complex. TCR also refers to an immunoglobulin superfamily member having a variable binding domain, a constant domain, a transmembrane region, and a short cytoplasmic tail (see, e.g., Janeway el al., Immunobiology: The Immune System in Health and Disease, 3rd Ed., Current Biology Publications, 1997) capable of specifically binding to an antigen peptide bound to a MHC receptor.

A TCR can be found on the surface of a cell and generally is comprised of a heterodimer having a and P chains (also known as TCRa and TCRP, respectively), or y and 6 chains (also known as TCRy and TCR6, respectively). Like immunoglobulins, the extracellular portions of TCR chains e.g, a-chain, P-chain) contain two immunoglobulin regions, a variable region (e.g., TCR variable a region or Va and TCR variable P region or VP; typically amino acids 1 to 116 based on Kabat numbering at the N-terminus), and one constant region (e.g., TCR constant domain a or Ca and typically amino acids 117 to 259 based on Kabat, TCR constant domain p or CP, typically amino acids 117 to 295 based on Kabat) adjacent to the cell membrane. Also, like immunoglobulins, the variable domains contain CDRs separated by framework regions (FRs). In some embodiments, a TCR is found on the surface of T cells (or T lymphocytes) and associates with the CD3 complex. The source of a TCR of the present disclosure may be from various animal species, such as a human, mouse, rat, rabbit or other mammal. In some embodiments, the source of a TCR of the present disclosure is a mouse genetically engineered to produce TCRs comprising human alpha and beta chains (see, e.g., WO 2016/164492).

TCRa and TCRP polypeptides (and similarly, TCRy and TCRS polypeptides) are linked to each other via a disulfide bond. Each of the two polypeptides that make up the TCR contains an extracellular domain comprising constant and variable regions, a transmembrane domain, and a cytoplasmic tail (the transmembrane domain and the cytoplasmic tail also being a part of the constant region). The variable region of the TCR determines its antigen specificity, and, similar to immunoglobulins, comprises three CDRs. The TCR is expressed on most T cells in the body and is known to be involved in recognition of MHC-restricted antigens. The TCR a chain includes a covalently linked Va and Ca region, whereas the P chain includes a VP region covalently linked to a CP region. The Va and VP regions form a pocket or cleft that can bind an antigen in the context of a major histocompatibility complex (MHC) (or HLA in humans).

As used herein, a “chimeric antigen receptor” or “CAR” refers to an antigen-binding protein that includes an immunoglobulin antigen-binding domain (e.g., an immunoglobulin variable domain) and a TCR constant domain or a portion thereof, which can be administered to a subject as chimeric antigen receptor T-cell (CAR-T) therapy. As used herein, a “constant domain” of a TCR polypeptide includes a membrane-proximal TCR constant domain, and may also include a TCR transmembrane domain and/or a TCR cytoplasmic tail. For example, in some embodiments, the CAR is a dimer that includes a first polypeptide comprising an immunoglobulin heavy chain variable domain linked to a TCRP constant domain and a second polypeptide comprising an immunoglobulin light chain variable domain (e.g., a K or variable domain) linked to a TCRa constant domain. In some embodiments, the CAR is a dimer that includes a first polypeptide comprising an immunoglobulin heavy chain variable domain linked to a TCRa constant domain and a second polypeptide comprising an immunoglobulin light chain variable domain (e.g., a K or X variable domain) linked to a TCRP constant domain.

CARs are typically artificial, constructed hybrid proteins or polypeptides containing the antigen-binding domain of an scFv or other antibody agent linked to a T cell signaling domain. In the context of this disclosure, the CAR is directed to a tumor-associated antigen. Features of the CAR include its ability to redirect T cell specificity and reactivity against selected targets in a non- MHC-restricted manner using the antigen-binding properties of monoclonal antibodies. Non- MHC-restricted antigen recognition provides CAR-expressing T cells with the ability to recognize antigens independent of antigen processing, thereby bypassing the major mechanism of tumor escape. As used in the adoptive cell therapy disclosed herein, immune cells can be manipulated to express the CAR in any known manner, including, for example, by transfection using RNA and DNA, both techniques being known in the art.

In some embodiments, TCR- or CAR-expressing immune effector cells are formulated by first harvesting them from their culture medium, and then washing and concentrating the cells in a medium and container system suitable for administration (a “pharmaceutically acceptable” carrier) in a treatment-effective amount. A suitable infusion medium can be any isotonic medium formulation, typically normal saline, Normosol R (Abbott) or Plasma-Lyte A (Baxter), but also 5% dextrose in water or Ringer’s lactate can be utilized. The infusion medium may be supplemented with human serum albumin.

A therapeutically effective number of immune cells to be administered in the disclosed methods is typically greater than 10² cells, such as up to and including 10⁶, up to and including 10⁸, up to and including 10⁹ cells, or more than IO¹⁰ cells. The number and/or type of cells to be administered to a subject will depend upon the ultimate use for which the therapy is intended.

TCRs and CARs of the present disclosure may be recombinant, meaning that they may be created, expressed, isolated or obtained by technologies or methods known in the art as recombinant DNA technology, which include, e.g., DNA splicing and transgenic expression. Recombinant TCRs or CARs may be expressed in a non-human mammal (including transgenic non-human mammals, e.g., transgenic mice), or a cell (e.g., CHO cells) expression system or isolated from a recombinant combinatorial human antibody library.

In some embodiments, immune cells, e.g., antigen-specific lymphocytes (e.g., neoantigen- specific lymphocytes), as described herein, may be administered to a subject at a dose ranging from about 10⁷ to about 10¹². A more accurate dose can also depend on the subject to which it is being administered. For example, a lower dose may be required if the subject is juvenile, and a higher dose may be required if the subject is an adult human subject. In some embodiments, a more accurate dose can depend on the weight of the subject.

In some embodiments, administration of the T cell therapy may be carried out in any convenient way, including infusion or injection (/.<?., intravenous, intrathecal, intramuscular, intraluminal, intratracheal, intraperitoneal, or subcutaneous), transdermal administration, or other methods known in the art. Administration can be once every two weeks, once a week, or more often, but the frequency may be decreased during a maintenance phase of the disease or disorder.

In some embodiments, immune cells, e.g., antigen-specific lymphocytes, may be activated and expanded using the methods described herein or other methods known in the art. In some embodiments, the immune cells may be expanded to therapeutic levels, before administering to a patient together with (e.g., before, simultaneously or after) any number of relevant treatment modalities.

In some embodiments, the disclosed methods lead to increased efficacy and duration of anti-tumor response. Methods according to this aspect of the disclosure may include selecting a subject with cancer and administering to the subject a therapeutically effective amount of a cancer therapy (e.g., adoptive cell therapy). In some embodiments, the methods provide for increased tumor inhibition, e.g, by about 20%, more than 20%, more than 30%, more than 40%, more than 50%, more than 60%, more than 70%, or more than 80% as compared to an untreated subject. In some embodiments, the methods provide for increased duration of the anti-tumor response, e.g., by about 20%, more than 20%, more than 30%, more than 40%, more than 50%, more than 60%, more than 70% or more than 80% as compared to an untreated subject. In some embodiments, administration of a cancer therapy (e.g., adoptive cell therapy) increases response and duration of response in a subject, e.g., by more than 2%, more than 3%, more than 4%, more than 5%, more than 6%, more than 7%, more than 8%, more than 9%, more than 10%, more than 20%, more than 30%, more than 40% or more than 50% more than an untreated subject.

In some embodiments, the disclosed methods lead to a delay in tumor growth and development, e.g., tumor growth may be delayed by about 3 days, more than 3 days, about 7 days, more than 7 days, more than 15 days, more than 1 month, more than 3 months, more than 6 months, more than 1 year, more than 2 years, or more than 3 years as compared to an untreated subject.

In some embodiments, administration of any of the combinations disclosed herein prevents tumor recurrence and/or increases duration of survival of the subject, e.g., increases duration of survival by 1-5 days, by 5 days, by 10 days, by 15 days, more than 15 days, more than 1 month, more than 3 months, more than 6 months, more than 12 months, more than 18 months, more than 24 months, more than 36 months, or more than 48 months more than the survival of an untreated subject.

In some embodiments, administration of a cancer therapy (e.g., adoptive cell therapy) to a subject with a cancer leads to complete disappearance of all evidence of tumor cells (“complete response”). In some embodiments, administration of a cancer therapy (e.g., adoptive cell therapy) to a subject with a cancer leads to at least 30% or more decrease in tumor cells or tumor size (“partial response”). In some embodiments, administration of a cancer therapy (e.g., adoptive cell therapy) to a subject with a cancer leads to complete or partial disappearance of tumor cells/lesions, including new measurable lesions. Tumor reduction can be measured by any methods known in the art, e.g., X-rays, positron emission tomography (PET), computed tomography (CT), magnetic resonance imaging (MRI), cytology, histology, or molecular genetic analyses.

In some embodiments, administration of a cancer therapy (e.g., adoptive cell therapy) to a subject with a cancer leads to an improved overall response rate, as compared to an untreated subject. In some embodiments, administration of a cancer therapy (e.g., adoptive cell therapy) to a subject with a cancer leads to increased overall survival (OS) or progression-free survival (PFS) of the subject as compared to a untreated subject.

In some embodiments, the PFS is increased by at least one month, at least 2 months, at least 3 months, at least 4 months, at least 5 months, at least 6 months, at least 7 months, at least 8 months, at least 9 months, at least 10 months, at least 11 months, at least 1 year, at least 2 years, or at least 3 years as compared to a untreated subject.

In some embodiments, the OS is increased by at least one month, at least 2 months, at least 3 months, at least 4 months, at least 5 months, at least 6 months, at least 7 months, at least 8 months, at least 9 months, at least 10 months, at least 11 months, at least 1 year, at least 2 years, or at least 3 years as compared to a untreated subject.

As used herein, “cancer,” “tumor,” and “malignancy” all relate equivalently to hyperplasia of a tissue or organ. If the tissue is a part of the lymphatic or immune system, malignant cells may include non-solid tumors of circulating cells. Malignancies of other tissues or organs may produce solid tumors. The methods described herein can be used in the treatment of lymphatic cells, circulating immune cells, and solid tumors.

As used herein, the term “cancer” refers to a malignant neoplasm characterized by deregulated or unregulated cell growth. The term “cancer” includes primary malignant cells or tumors (e.g., those whose cells have not migrated to sites in the subject’s body other than the site of the original malignancy or tumor) and secondary malignant cells or tumors (e.g., those arising from metastasis, the migration of malignant cells or tumor cells to secondary sites that are different from the site of the original tumor). The term “metastatic” or “metastasis” generally refers to the spread of a cancer from one organ or tissue to another non-adj acent organ or tissue. The occurrence of the neoplastic disease in the other non-adj acent organ or tissue is referred to as metastasis.

Examples of cancer include but are not limited to carcinoma, lymphoma, blastoma, sarcoma, and leukemia or lymphoid malignancies. More particular examples of such cancers include without limitation: squamous cell cancer (e.g., epithelial squamous cell cancer), lung cancer including small-cell lung cancer, non-small cell lung cancer, adenocarcinoma of the lung, squamous carcinoma of the lung and large cell carcinoma of the lung, cancer of the peritoneum, hepatocellular cancer, gastric or stomach cancer including gastrointestinal cancer, pancreatic cancer, glioma, glioblastoma, cervical cancer, ovarian cancer, liver cancer, bladder cancer, hepatoma, breast cancer, colon cancer, rectal cancer, colorectal cancer, endometrial cancer or uterine carcinoma, salivary gland carcinoma, kidney or renal cancer, prostate cancer, vulvar cancer, thyroid cancer, hepatic carcinoma, anal carcinoma, penile carcinoma, as well as CNS cancer, melanoma, head and neck cancer, bone cancer, bone marrow cancer, duodenum cancer, esophageal cancer, thyroid cancer, or hematological cancer.

Other examples of cancers or malignancies include, but are not limited to: Acute Childhood Lymphoblastic Leukemia, Acute Lymphoblastic Leukemia, Acute Lymphocytic Leukemia, Acute Myeloid Leukemia, Adrenocortical Carcinoma, Adult (Primary) Hepatocellular Cancer, Adult (Primary) Liver Cancer, Adult Acute Lymphocytic Leukemia, Adult Acute Myeloid Leukemia, Adult Hodgkin’s Disease, Adult Hodgkin’s Lymphoma, Adult Lymphocytic Leukemia, Adult Non- Hodgkin’s Lymphoma, Adult Primary Liver Cancer, Adult Soft Tissue Sarcoma, AIDS- Related Lymphoma, AIDS-Related Malignancies, Anal Cancer, Astrocytoma, Bile Duct Cancer, Bladder Cancer, Bone Cancer, Brain Stem Glioma, Brain Tumors, Breast Cancer, Cancer of the Renal Pelvis and Urethra, Central Nervous System (Primary) Lymphoma, Central Nervous System Lymphoma, Cerebellar Astrocytoma, Cerebral Astrocytoma, Cervical Cancer, Childhood (Primary) Hepatocellular Cancer, Childhood (Primary) Liver Cancer, Childhood Acute Lymphoblastic Leukemia, Childhood Acute Myeloid Leukemia, Childhood Brain Stem Glioma, Glioblastoma, Childhood Cerebellar Astrocytoma, Childhood Cerebral Astrocytoma, Childhood Extracranial Germ Cell Tumors, Childhood Hodgkin’s Disease, Childhood Hodgkin’s Lymphoma, Childhood Hypothalamic and Visual Pathway Glioma, Childhood Lymphoblastic Leukemia, Childhood Medulloblastoma, Childhood Non-Hodgkin’s Lymphoma, Childhood Pineal and Supratentorial Primitive Neuroectodermal Tumors, Childhood Primary Liver Cancer, Childhood Rhabdomyosarcoma, Childhood Soft Tissue Sarcoma, Childhood Visual Pathway and Hypothalamic Glioma, Chronic Lymphocytic Leukemia, Chronic Myelogenous Leukemia, Colon Cancer, Cutaneous T-Cell Lymphoma, Endocrine Pancreas Islet Cell Carcinoma, Endometrial Cancer, Ependymoma, Epithelial Cancer, Esophageal Cancer, Ewing’s Sarcoma and Related Tumors, Exocrine Pancreatic Cancer, Extracranial Germ Cell Tumor, Extragonadal Germ Cell Tumor, Extrahepatic Bile Duct Cancer, Eye Cancer, Female Breast Cancer, Gallbladder Cancer, Gastric Cancer, Gastrointestinal Carcinoid Tumor, Gastrointestinal Tumors, Germ Cell Tumors, Gestational Trophoblastic Tumor, Hairy Cell Leukemia, Head and Neck Cancer, Hepatocellular Cancer, Hodgkin’s Disease, Hodgkin’s Lymphoma, Hypergammaglobulinemia, Hypopharyngeal Cancer, Intestinal Cancers, Intraocular Melanoma, Islet Cell Carcinoma, Islet Cell Pancreatic Cancer, Kaposi’s Sarcoma, Kidney Cancer, Laryngeal Cancer, Lip and Oral Cavity Cancer, Liver Cancer, Lung Cancer, Lymphoproliferative Disorders, Macroglobulinemia, Male Breast Cancer, Malignant Mesothelioma, Malignant Thymoma, Medulloblastoma, Melanoma, Mesothelioma, Metastatic Occult Primary Squamous Neck Cancer, Metastatic Primary Squamous Neck Cancer, Metastatic Squamous Neck Cancer, Multiple Myeloma, Multiple Myeloma/Plasma Cell Neoplasm, Myelodysplasia Syndrome, Myelogenous Leukemia, Myeloid Leukemia, Myeloproliferative Disorders, Nasal Cavity and Paranasal Sinus Cancer, Nasopharyngeal Cancer, Neuroblastoma, Non-Hodgkin’s Lymphoma During Pregnancy, Non-melanoma Skin Cancer, Non-Small Cell Lung Cancer, Occult Primary Metastatic Squamous Neck Cancer, Oropharyngeal Cancer, Osteo- /Malignant Fibrous Sarcoma, Osteosarcoma/Malignant Fibrous Histiocytoma, Osteosarcoma/Malignant Fibrous Histiocytoma of Bone, Ovarian Epithelial Cancer, Ovarian Germ Cell Tumour, Ovarian Low Malignant Potential Tumor, Pancreatic Cancer, Paraganglioma, Paraproteinemias, Purpura, Parathyroid Cancer, Penile Cancer, Pheochromocytoma, Pituitary Tumor, Plasma Cell Neoplasm/Multiple Myeloma, Primary Central Nervous System Lymphoma, Primary Liver Cancer, Prostate Cancer, Rectal Cancer, Renal Cell Cancer, Renal Pelvis and Urethra Cancer, Retinoblastoma, Rhabdomyosarcoma, Salivary Gland Cancer, Sarcoidosis Sarcomas, Sezary Syndrome, Skin Cancer, Small Cell Lung Cancer, Small Intestine Cancer, Soft Tissue Sarcoma, Squamous Neck Cancer, Stomach Cancer, Supratentorial Primitive Neuroectodermal and Pineal Tumors, T-Cell Lymphoma, Testicular Cancer, Thymoma, Thyroid Cancer, Transitional Cell Cancer of the Renal Pelvis and Urethra, Transitional Renal Pelvis and Urethra Cancer, Trophoblastic Tumours, Urethra and Renal Pelvis Cell Cancer, Urethral Cancer, Uterine Cancer, Uterine Sarcoma, Vaginal Cancer, Visual Pathway and Hypothalamic Glioma, Vulvar Cancer, Waldenstrom’s Macroglobulinemia, or Wilms’ Tumour.

In some embodiments, the tumor, including any metastases of the tumor, may be of epithelial or melanocyte origin. In some embodiments, the tumor, including any metastases of the tumor, may originate from chromaffin cells, ganglia of the sympathetic nervous system, follicular thyroid cells or parafollicular thyroid cells.

Tumors of epithelial origin include any tumors originated from epithelial tissue in any of several sites, such as without limitation skin, lung, intestine, colon, breast, bladder, head and neck (including lips, oral cavity, salivary glands, nasal cavity, nasopharynx, paranasal sinuses, pharynx, throat, larynx, and associated structures), esophagus, thyroid, kidney, liver, pancreas, bladder, penis, testes, prostate, vagina, cervix, or anus.

In some embodiments, the tumor may be a carcinoma, including any malignant neoplasm originated from epithelial tissue in any of several sites, such as without limitation skin, lung, intestine, colon, breast, bladder, head and neck (including lips, oral cavity, salivary glands, nasal cavity, nasopharynx, paranasal sinuses, pharynx, throat, larynx, and associated structures), esophagus, thyroid, kidney, liver, pancreas, bladder, penis, testes, prostate, vagina, cervix, or anus. In some embodiments, the tumor may be thyroid carcinoma.

In some embodiments, the tumor may be a squamous cell carcinoma (SCC). SCC may include, without limitation, SCC originated from skin, head and neck (including lips, oral cavity, salivary glands, nasal cavity, nasopharynx, paranasal sinuses, pharynx, throat, larynx, and associated structures), thyroid, esophagus, lung, penis, prostate, vagina, cervix, anus, or bladder. In some embodiments, the tumor may be lung squamous cell carcinoma, or head and neck squamous cell carcinoma.

Tumors of melanocyte origin include any tumors originated from melanocytes in any of several sites, such as without limitation skin, mouth, eyes, or small intestine.

In some embodiments, the tumor may be a melanoma, including any malignant neoplasm originated from melanocytes in any of several sites, such as without limitation skin, mouth, eyes, or small intestine. In some embodiments, the tumor may be skin cutaneous melanoma.

Tumors originating from chromaffin cells include pheochromocytoma. Tumors originating from ganglia of the sympathetic nervous system include paraganglioma. Tumors originating from follicular or parafollicular thyroid cells include thymoma.

In some embodiments, the tumor may be pheochromocytoma, paraganglioma, or thymoma. Also described herein, the T cell therapy can be used in combination with chemotherapy, radiation, immunosuppressive agents, such as cyclosporin, azathioprine, methotrexate, mycophenolate, and FK506, antibodies, or other immunoablating agents such as CAMPATH, anticancer antibodies. CD3 or other antibody therapies, cytoxine, fludarabine, cyclosporine, FK506, rapamycin, mycophenolic acid, steroids, FR901228, cytokines, and irradiation.

Additional Definitions

To aid in understanding the detailed description of the compositions and methods according to the disclosure, a few express definitions are provided to facilitate an unambiguous disclosure of the various aspects of the disclosure. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

Unless defined otherwise, all technical and scientific terms used herein have the meaning commonly understood by a person skilled in the art to which this invention belongs. The following references provide one of skill with a general definition of many of the terms used in this invention: Singleton et al., Dictionary of Microbiology and Molecular Biology (2nd ed. 1994); The Cambridge Dictionary of Science and Technology (Walker ed., 1988); The Glossary of Genetics, 5th Ed., R. Rieger et al. (eds.), Springer Verlag (1991); and Hale & Marham, The Harper Collins Dictionary of Biology (1991). As used herein, the following terms have the meanings ascribed to them below, unless specified otherwise.

As used herein, the phrases “nucleic acid,” “polynucleotide,” “oligonucleotide,” and “nucleic acid molecule” are used interchangeably to refer to a polymer of DNA and/or RNA, which can be single-stranded, double-stranded, or multi -stranded, synthesized or obtained (e.g., isolated and/or purified) from natural sources, which can contain natural, non-natural, and/or altered nucleotides, and which can contain natural, non-natural, and/or altered internucleotide linkages including, but not limited to phosphoroamidate linkages and/or phosphorothioate linkages instead of the phosphodiester found between the nucleotides of an unmodified oligonucleotide.

The term “gene” is well-known in the art and refers to a locatable region of a genomic sequence corresponding to a unit of inheritance, which is associated with regulatory regions, transcribed regions and/or other functional sequence regions. Genes typically comprise a coding sequence encoding a gene product, such as an RNA molecule or a polypeptide. As used herein, the term “polypeptide” refers to any polymer preferably consisting essentially of any of the 20 natural amino acids regardless of its size. Although the term “protein” is often used in reference to relatively large proteins, and “peptide” is often used in reference to small polypeptides, use of these terms in the field often overlaps. The term “polypeptide” refers generally to proteins, polypeptides, and peptides unless otherwise noted. Peptides useful in accordance with the present disclosure are generally between about 0.1 to 100 KD or greater up to about 1000 KD, preferably between about 0.1, 0.2, 0.5, 1, 2, 5, 10, 20, 30, and 50 KD as judged by standard molecule sizing techniques such as centrifugation or SDS-polyacrylamide gel electrophoresis.

As used herein, the term “immunotherapy” refers to any treatment that modulates a subject’s immune system. In particular, the term comprises any treatment that modulates an immune response, such as a humoral immune response, a cell-mediated immune response, or both. An immune response may typically involve a response by a cell of the immune system, such as a B cell, cytotoxic T cell (CTL), T helper (Th) cell, regulatory T (Treg) cell, antigen-presenting cell (APC), dendritic cell, monocyte, macrophage, natural killer T (NKT) cell, natural killer (NK) cell, basophil, eosinophil, or neutrophil, to a stimulus. In the context of anti-cancer treatments, immunotherapy may elicit, induce or enhance an immune response, such as in particular an immune response specifically against tumor tissues or cells, such as to achieve tumor cell death. Immunotherapy may modulate, such as increase or enhance, the abundance, function, and/or activity of any component of the immune system, such as any immune cell, such as without limitation T cells (e.g., CTLs or Th cells), dendritic cells, and/or NK cells. Immunotherapies can be categorized as active, passive or a combination thereof. Anti-cancer immunotherapy is based on the fact that cancer cells typically have molecules on their surface, known as tumor antigens that can be detected by the immune system. Active immunotherapy directs the immune system to attack tumor cells by targeting tumor antigens. Passive immunotherapies enhance existing antitumor responses, typically through binding and modifying the intracellular signaling of surface receptors. Immunotherapy comprises cell-based immunotherapy in which immune cells, such as T cells and/or dendritic cells, are transferred into the patient. The term also comprises an administration of substances or compositions, such as chemical compounds and/or biomolecules (e.g., antibodies, antigens, interleukins, cytokines, or combinations thereof), that modulate a subject’s immune system. Examples of cancer immunotherapy include, without limitation, treatments employing monoclonal antibodies, for example, immune checkpoint inhibitors, Fc- engineered monoclonal antibodies against proteins expressed by tumor cells, prophylactic or therapeutic cancer vaccines, adoptive cell therapy, and combinations thereof.

The terms “prevent,” “preventing,” “prevention,” “prophylactic treatment,” and the like refer to reducing the probability of developing a disorder or condition in a subject, who does not have, but is at risk of or susceptible to developing a disorder or condition. The term includes prevention of spread of infection in a subject exposed to the virus or at risk of having cancer.

The term “immune response,” as used herein, refers to any type of immune response, including, but not limited to, innate immune responses (e.g., activation of Toll receptor signaling cascade), cell-mediated immune responses (e.g., responses mediated by T cells (e.g., antigenspecific T cells) and non-specific cells of the immune system) and humoral immune responses (e.g. , responses mediated by B cells (e.g. , via generation and secretion of antibodies into the plasma, lymph, and/or tissue fluids). The term “immune response” is meant to encompass all aspects of the capability of a subject’s immune system to respond to antigens and/or immunogens (e.g., both the initial response to an immunogen (e.g., a pathogen) as well as acquired (e.g., memory) responses that are a result of an adaptive immune response).

The term “disease” as used herein is intended to be generally synonymous and is used interchangeably with the terms “disorder” and “condition” (as in medical condition), in that all reflect an abnormal condition of the human or animal body or of one of its parts that impairs normal functioning, is typically manifested by distinguishing signs and symptoms, and causes the human or animal to have a reduced duration or quality of life.

The term “effective amount,” “effective dose,” or “effective dosage” is defined as an amount sufficient to achieve or at least partially achieve a desired effect. A “therapeutically effective amount” or “therapeutically effective dosage” of a drug or therapeutic agent is any amount of the drug that, when used alone or in combination with another therapeutic agent, promotes disease regression evidenced by a decrease in severity of disease symptoms, an increase in frequency and duration of disease symptom-free periods, or a prevention of impairment or disability due to the disease affliction. A “prophylactically effective amount” or a “prophylactically effective dosage” of a drug is an amount of the drug that, when administered alone or in combination with another therapeutic agent to a subject at risk of developing a disease or of suffering a recurrence of disease, inhibits the development or recurrence of the disease. The ability of a therapeutic or prophylactic agent to promote disease regression or inhibit the development or recurrence of the disease can be evaluated using a variety of methods known to the skilled practitioner, such as in human subjects during clinical trials, in animal model systems predictive of efficacy in humans, or by assaying the activity of the agent in in vitro assays.

The term “agent” is used herein to denote a chemical compound, a mixture of chemical compounds, a biological macromolecule (such as a nucleic acid, an antibody, a protein or portion thereof, e.g., a peptide), or an extract made from biological materials such as bacteria, plants, fungi, or animal (particularly mammalian) cells or tissues. The activity of such agents may render it suitable as a “therapeutic agent,” which is a biologically, physiologically, or pharmacologically active substance (or substances) that acts locally or systemically in a subject.

The terms “therapeutic agent,” “therapeutic capable agent,” or “treatment agent” are used interchangeably and refer to a molecule or compound that confers some beneficial effect upon administration to a subject. The beneficial effect includes enablement of diagnostic determinations; amelioration of a disease, symptom, disorder or pathological condition; reducing or preventing the onset of a disease, symptom, disorder or condition; and generally counteracting a disease, symptom, disorder or pathological condition.

“Combination” therapy, as used herein, unless otherwise clear from the context, is meant to encompass administration of two or more therapeutic agents in a coordinated fashion and includes, but is not limited to, concurrent dosing. Specifically, combination therapy encompasses both co-administration (e.g., administration of a co-formulation or simultaneous administration of separate therapeutic compositions) and serial or sequential administration, provided that administration of one therapeutic agent is conditioned in some way on administration of another therapeutic agent. For example, one therapeutic agent may be administered only after a different therapeutic agent has been administered and allowed to act for a prescribed period of time. See, e.g., Kohrt e/ aZ. (2011) /c 117:2423.

Doses are often expressed in relation to bodyweight. Thus, a dose which is expressed as [g, mg, or other unit]/kg (or g, mg etc.) usually refers to [g, mg, or other unit] “per kg (or g, mg etc.) body weight,” even if the term “body weight” is not explicitly mentioned. As used herein, “in vitro’’’ refers to events that occur in an artificial environment, e.g., in a test tube or reaction vessel, in cell culture, etc., rather than within a multi-cellular organism.

As used herein, “in vivo” refers to events that occur within a multi-cellular organism, such as a non-human animal.

It is noted here that, as used in this specification and the appended claims, the singular forms “a,” “an,” and “the” include plural reference unless the context clearly dictates otherwise.

As used herein, “including,” “comprising,” “containing,” or “having” and variations thereof are meant to encompass the items listed thereafter and equivalents thereof as well as additional subject matter unless otherwise noted.

As used herein, the phrases “in one embodiment,” “in various embodiments,” “in some embodiments,” and the like do not necessarily refer to the same embodiment, but may unless the context dictates otherwise.

As used herein, the terms “and/or” or “/” means any one of the items, any combination of the items, or all of the items with which this term is associated.

As used herein, the term “substantially” does not exclude “completely,” e.g., a composition which is “substantially free” from Y may be completely free from Y. Where necessary, the word “substantially” may be omitted from the definition of the present disclosure.

As used herein, the term “approximately” or “about,” as applied to one or more values of interest, refers to a value that is similar to a stated reference value. In some embodiments, the term “approximately” or “about” refers to a range of values that fall within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less in either direction (greater than or less than) of the stated reference value unless otherwise stated or otherwise evident from the context (except where such number would exceed 100% of a possible value). Unless indicated otherwise herein, the term “about” is intended to include values, e.g., weight percents, proximate to the recited range that are equivalent in terms of the functionality of the individual ingredient, the composition, or the embodiment.

As used herein, the term “each,” when used in reference to a collection of items, is intended to identify an individual item in the collection but does not necessarily refer to every item in the collection. Exceptions can occur if explicit disclosure or context clearly dictates otherwise. As disclosed herein, a number of ranges of values are provided. It is understood that each intervening value, to the tenth of the unit of the lower limit, unless the context clearly dictates otherwise, between the upper and lower limits of that range is also specifically disclosed. Each smaller range between any stated value or intervening value in a stated range and any other stated or intervening value in that stated range is encompassed within the present disclosure. The upper and lower limits of these smaller ranges may independently be included or excluded in the range, and each range where either, neither, or both limits are included in the smaller ranges is also encompassed within the present disclosure, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the present disclosure.

The use of any and all examples, or exemplary language (e.g., “such as”) provided herein, is intended merely to better illuminate the invention and does not pose a limitation on the scope of the present disclosure unless otherwise claimed. No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the present disclosure.

All methods described herein are performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. In regard to any of the methods provided, the steps of the method may occur simultaneously or sequentially. When the steps of the method occur sequentially, the steps may occur in any order, unless noted otherwise. In cases in which a method comprises a combination of steps, each and every combination or sub-combination of the steps is encompassed within the scope of the disclosure, unless otherwise noted herein.

Each publication, patent application, patent, and other reference cited herein is incorporated by reference in its entirety to the extent that it is not inconsistent with the present disclosure. Publications disclosed herein are provided solely for their disclosure prior to the filing date of the present disclosure. Nothing herein is to be construed as an admission that the present disclosure is not entitled to antedate such publication by virtue of prior disclosure. Further, the dates of publication provided may be different from the actual publication dates, which may need to be independently confirmed.

It is understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application and scope of the appended claims.

Examples

EXAMPLE 1

Single cell atlas of the tumor microenvironment predicts response to adoptive T-cell therapy in melanoma patients

This example describes the clinical results of a single-center phase I study to assess feasibility, safety, and efficacy of TIL-ACT in melanoma patients (NCT03475134). Comprehensive translational studies were performed on patients’ tumor longitudinal samples, including multispectral immuno-fluorescence (mIF) imaging, bulk RNA-sequencing, and singlecell RNA-sequencing before and after TILs infusion (30 days). RECIST 1.1 criteria were applied to define radiological responses.

Thirteen patients successfully completed TIL-ACT therapy and were included in the analysis. The best overall response was 46.1% (6/13) at 3 months, including two patients with ongoing near-complete response 3 years after infusion. Multiplex immunofluorescence staining revealed that responders had increased densities of polyfunctional intra-epithelial (i.e.,) CD8 T cells marked by higher PD-1 expression at baseline, and these cells persisted 30 days after TILs infusion only in responders. Single-cell transcriptomic analyses confirmed higher cytotoxic and exhaustion CD8 T cell states at baseline. PDCD1 (PD-1), CXCL13, TNFRSF9 (CD137), GZMB, HAVCR2 (TIM3), and PRF1 were some of the most overexpressed genes in CD8 TILs and associated with TIL- ACT clinical responses. Although the overall infiltration rate of macrophages and dendritic cells did not differ according to clinical response, specific activated subtypes of macrophages overexpressing complement genes, type I IFN signatures, and CXCL9/10 chemokines were found in higher proportion in responders, and these cells were interacting with activated CD8 T cell subtypes. Finally, responders exhibited immunogenic tumor cell states with higher inferred CNVs, overexpressing DNA sensing/IFN, and class I antigen presentation-related genes.

Comprehensive TME profiling revealed divergent phenotypical and functional tumor, TME, and TIL states between responders and non-responders to TIL-ACT. Emerging TIL and TME derived biomarkers that predict therapeutic TIL-ACT efficacy serve to improve patient selection.

Identification and isolation of tumor reactive TCRs from physically interacting TIL- based doublets using lOXGenomics scRNAseq/TCRseq platform

In single-cell RNA sequencing (scRNA-seq) experiments, doublets are generally viewed as artifactual gene expression matrices generated from two cells. Doublets are traditionally considered undesirable since the main aim of most single cell transcriptomic studies is to characterize populations at the single-cell level. Some of the reasons include the biased interpretation of doublets as intermediate populations or transitory states that may not exist. For these reasons, several experimental strategies are available for doublet detection and filtering or removal from scRNAseq data analysis (Kang et al. 2018), (Stoeckius et al. 2018), (Dahlin et al. 2018).

In the ATATIL scRNAseq data (lOxGenomics platform) analysis, transcriptomic profiles and states of single cells (singlets) originating from malignant, stromal, lymphoid, and myeloid sources were extensively analyzed. In order to discover biomarkers and traits of tumor infiltrating lymphocytes (TIL) and tumor microenvironment (TME) associated with response or resistance to TIL ACT therapy, potential interactions from singlets of populations of interest were predicted using available bioinformatics tools (Armingol et al., 2021, review). At the same time, naturally occurring physical doublets of cell states predicted to interact from singlets gene expression analysis were also detected. The cell origins forming these doublets were deconvoluted using the gene expression profiles of the singlets, and it was found that these heterotypic doublets were composed by TILs and antigen-presenting cells (APC) states of different leukocyte lineages reflecting increased tumor reactivity, cytotoxicity, exhaustion and costimulation; all gene signatures discriminated responders from non-responding patients at steady state but also upon TIL immunotherapy.

These data indicate that the scRNAseq/TCRseq lOXGenomics-based workflow, despite focused primarily on transcriptomic analysis of singlets, permits detecting and deciphering doublets, their origins, and underlying molecular mechanisms.

The intentional isolation and detection of these doublets for adoptive immunotherapy applications using the already established workflow of lOxGenomics were performed. The T cell receptor repertoire information of TILs determined by scTCR sequencing found in these doublets, testing their tumor reactivity, and benchmarking this, are an effective approach for the isolation of antigen-specific TCRs for adoptive cell immunotherapy.

Table 1. Selected genes for best overall response (upregulated)

Table 2. Selected genes for best overall response (downregulated)

Table 3. Genes identified as tumor reactive (upregulated)

Table 4. Genes identified as tumor reactive (downregulated)

EXAMPLE 2

This example describes materials and methods for the subsequent example (Example 3).

Patients

From March 2018 to January 2021, fifteen patients were enrolled in a single-center phase 1 investigator-initiated trial designed to test the feasibility of ACT with TILs (ClinicalTrials.gov NCT03475134). Among them, two patients did not receive any trial treatment: one patient failed during the rapid expansion phase (REP)-TIL, and another one withdrew informed consent from the study before the start of the treatment. Thirteen eligible patients constitute the ‘per protocol’ cohort of the study for efficacy analysis. Eligible patients were adults with histologically proven unresectable locally advanced (stage IIIc) or metastatic (stage IV) melanoma who have progressed on at least one standard first-line therapy, including but not limited to chemotherapy, BRAF and MEK inhibitors, anti-CTLA4, anti-PD-1, anti-PD-Ll or anti-LAG3 antibodies and/or the combination. When feasible, a biopsy of one metastatic deposit was performed at screening to assess and quantify the intratumoral CD3⁺ and CD8⁺ infiltration by a dedicated Pathologist (JD). In some cases, archived FFPE material from diagnosis was retrieved for analysis.

Patients were required to have an accessible metastasis to procure for TILs with acceptable anticipated perioperative risk and also at least one separate additional measurable tumor lesion on CT. Patients were required to have a good general health status (ECOG PS < 2), sufficient cardiopulmonary function, including a cardiac stress test showing no reversible ischemia; adequate respiratory function with forced expiratory volume in 1 second (FEV1) > 65% predicted, forced vital capacity (FVC) > than 65% predicted and DLCO > than 50% predicted corrected; and left ventricular ejection fraction (LVEF) > 45%.

Patients with active brain metastases, autoimmune conditions or acquired immunodeficiency were excluded. Patients were required to have adequate normal organ and marrow function, defined as hemoglobin > 8 g/100 ml; absolute neutrophil count > 1.0 x 10⁹ (>1,000 mm^-3); platelet count >100 x 10⁹ (>100,000 mm^-3); serum creatinine < 1.5x of the institutional upper limit of normal; and AST and ALT < 3 of the institutional upper limit of normal. Patients with symptomatic and/or untreated brain metastases were excluded. Patients with definitively-treated brain metastases were considered for enrollment, as long as lesions were stable for > 14 days prior to beginning the chemotherapy, there were no new brain lesions, and the patient did not require ongoing corticosteroid treatment. Bridging therapy was allowed as per the investigator’s choice and upon discussion with the Principal Investigator.

Clinically eligible patients underwent surgery for TILs harvest and ex vivo expansion. Only patients having sufficient numbers of pre-REP TILs (TIL numbers > 50 million) were offered to receive TIL-ACT treatment. TILs were successfully expanded for all 13 patients from tumor deposits resected by surgery, and the median number of TILs infused was 55.0 billion cells (range: 12.8-84.7).

Study design

The primary endpoints were feasibility and safety of ACT using autologous TILs. Key secondary endpoints were feasibility and safety of nivolumab rescue following TIL- ACT, and the clinical efficacy of the treatment with respect to ORR, PFS, according to RECIST vl.l, and OS. Overall Survival (OS) was defined as the time from the start of NMA chemotherapy until death from any cause, for a maximum of 5 years. If there is no death date, the patient is censored on the last day known to be alive. Exploratory objectives included collection of exploratory translational data regarding the biological effects of the TIL-ACT and its interaction with the tumor microenvironment, using paired tumor biopsies before and after treatment, as well as blood samples. Tumor samples were collected at screening (if feasible), at surgery (tumor material for pre-REP), at day 30 after TILs- ACT, after 4 weeks of nivolumab treatment if applicable (optional), and at progression (optional). Adverse events were recorded according to NCI Common Terminology Criteria for Adverse Events (CTCAE v5.0).

Study conduct

Patients received a lymphodepletion regimen consisting of fludarabine (25 mg/m²/day) for 5 days and cyclophosphamide (60 mg/kg/day) for 2 (overlapping) days, followed by the infusion of T lymphocytes, which was followed by the administration of intravenous boluses of high dose IL-2 (720,000 lU/kg) starting 3 hours post- TIL infusion, then every 8h at minimum counting from the start of each administration, for a total of 8 doses maximum, with a maximum interval of 24h. In particular, ten patients (76.9%) received a full course of lymphodepleting chemotherapy (cyclophosphamide and fludarabine) without dose modification. All patients initiated high-dose IL-2 treatment and received a median of 5 doses of IL-2 (range 1-8). Adverse effects were primarily attributable to lymphodepletion and IL-2 administration. Common non-hematologic adverse events included nausea, hypophosphatemia and capillary leak syndrome (hypoalbuminemia, weight gain, and pulmonary edema).

Following day 30, at the completion of the TLT period, eligible patients started on nivolumab, as long as no TIL-ACT related toxicity > to grade 1 was observed. Nivolumab, at a dose of 240 mg IV every two weeks, was administered for the first 12 months, followed by nivolumab at a dose of 480 mg IV every four weeks for the next 12 months until unacceptable toxicities or confirmed disease progression.

Only 3 out of 11 patients (27%) were eligible for nivolumab rescue. Patient #7 progressed at 6 months after TILs infusion (Nov 2019) with new inguinal lymph nodal lesions and received six cycles of nivolumab treatment with initial stability of the disease and then new PD for which TKI therapy was started (Sept 2020) and still ongoing. Patient #9 progressed five months after TILs infusion and received five cycles of nivolumab with a further progression after three months. Patient #12 progressed at first-month assessment and received six cycles of nivolumab with further PD.

Synthetic Control Arm (SCA)

The SCA for the analysis is based on historical, real-world data from CHUV (Centre Hospitalier Universitaire Vaudois) patients, taking into account the current treatment strategies and the consistency of characteristics with the ATATIL patients. In order to build a synthetic comparator arm for patients who failed the standard of care, a search was conducted in the institutional clinical research data warehouse (June 2021) complemented by a number of fields curated specifically for the melanoma cohort. Uveal melanoma and patients who took part in an ACT-TIL trial were excluded. Patients who had progressed on first-line PD-1 or PD-1 + CTLA-4 blockades and, for BRAF V600 mutated melanoma, who subsequently failed BRAF + MEK inhibitors, were identified via a structured search by a physician. The BRAF status of the tumors has been used as a stratification factor. Of note, in the present analysis BRAF-positive subgroup refers precisely to BRAF-V600 mutated melanoma (potentially treated by BRAF inhibitors), while the BRAF-negative subgroup includes “non-mutated BRAF-V600” patients.

Immunohistochemistry, TIL assessment, and scoring For each case, representative tumor slides with the most viable component and higher inflammatory infiltrate were selected for immunohistochemical (IHC) staining on formalin-fixed paraffin-embedded (FFPE) whole tissue sections. CD3 (2GV6, CONFIRM, Rabbit Monoclonal, Roche, Basel, Switzerland), CD8 (C8/144B, Mouse Monoclonal, Dako, Glostrup, Denmark), PD- 1 (NAT105, Mouse Monoclonal, CellMarque, Rocklin, United States) and CXCL13 (BLC/BCA- 1, Goat Polyclonal, IgGR&D Systems, Minneapolis, United States) expressions were assessed. Briefly, immunohistochemical staining (IHC) was performed using the Ventana Benchmark Ultra (Roche Ventana Tucson, Arizona, USA) for the CD3, CD8, PD-1, and CXCL13 stains, by following the manufacturer’s instructions. Four pm -thick FFPE sections were subjected to routine deparaffmization, rehydration, and antigen retrieval procedures. Each section was incubated with the primary antibody, followed by the secondary antibody. The ultra View Universal DAB Detection Kit for CD3, CD8, and CXCL13 IHC (Ref 05269806001, Roche Ventana); and the OptiView DAB Detection Kit for the PD-1 IHC (Ref 06396500001, Roche Ventana) were used as a detection system. Tissue counterstaining was performed with Hematoxylin from Gil II solution (Ref 105175, MERCK). Sections of human tonsil were used as positive control. Evaluation was performed independently by one pathologist (JD) without knowledge of clinical information.

For each sample, at least 10 high-power fields (HPF) with a diameter of 500 pm were selected. These fields were distributed over the whole tumor sample. Intra-tumoral T lymphocytes were qualitatively assessed by CD3 (low to high), along with spatial distribution (stroma vs. tumor) and heterogeneity. CD8/CD3 ratio and CD8 ranges per HPF were then evaluated. The final TILs score was the mean intratumoral CD8⁺ cells in at least 10 HPFs. The tumor-infiltrating CD8⁺ T- cells were evaluated and classified as “intratumoral” if they were in direct contact with tumor cells. Cells stained positive in the stromal compartment and within the borders of the invasive tumor or in areas of necrosis were not evaluated. For cases with significant region variations in lymphocyte distribution, each region was evaluated separately, and an average value was assessed in > 10 HPF whenever possible. The number of PD-1 (membranous immune cell labeling) and CXCL13 (cytoplasmic and perinuclear cell labeling) positive cells in the same representative high-power fields (HPF) were also assessed and averaged.

Multispectral immunofluorescence tissue staining and image analyses For the multiplexed staining, FFPE sections were stained by an automated immunostainer (DISCOVERY ULTRA, Ventana Roche). First, the heat-induced antigen retrieval in EDTA buffer (pH 8.0) was performed for 92 min at 95°C. Multiplex staining was performed in consecutive rounds, each round consisting of protein blocking, primary antibody incubation, secondary HRP- labeled antibody incubation, OPAL detection reagents, and then antibodies heat denaturation. The Multiplex IF images were acquired on a Polaris imaging system (Perkin Elmer). Tissue- and panelspecific spectral libraries of the specific panel individual fluorophore and tumor tissue autofluorescence were acquired for an optimal IF signal unmixing (individual spectral peaks) and multiplex analysis. The IF-stained slides were pre-scanned at lOx magnification using the Phenochart whole-slide viewer. Using a Phenochart™ whole-slide viewer, regions of interest (ROI) representative of all samples were acquired. InForm 2.5.1 software was used for training and phenotyping analysis. The images were first segmented into specific tissue categories of tumor, stroma, and no tissue, based on the cytokeratin and DAPI staining using the inForm Tissue Finder™ algorithms. Individual cells were then segmented using the counterstained-based adaptive cell segmentation algorithm. Quantification of the immune cells was then performed using inForm active learning phenotyping algorithm by assigning the different cell phenotypes across several images chosen for the project. IF-stained cohorts were then batch processed on the whole image, and data were exported via an in-house developed R-script algorithm to retrieve every cell’s x,y coordinates and staining positivity and intensity. To calculate cell densities of different populations, the number of specific cell phenotypes was counted in both tumor and stroma compartments on a whole-slide basis. Counts of each cell type (tissue-specific) were divided by the area of tissue (mm²) to obtain a density (number of cells/mm²).

Neighborhood cell-to-cell distance analysis and density map

Given two sets of points in a two-dimensional space, a measure that estimates their vicinity can be implemented. Given the set A={ Ai, . . . ., AN}C, and the set B={Bi, . . . , BK}, the coordinate of each point A can be expressed as x_an, yan, and for each point B can be expressed as xbn, ybn.

For each A, the number of neighbors of type B was measured in a distance D < s, where D is:

This can be graphically viewed as counting the number of neighbors in a circle centered on the point Ai, with radius a. Therefore, the function “S” can be defined by applying to an Ai, to give the number of neighbors of type B around this element.

S(A_i, B) = v

The extremes of the function S goes from v = 0 to “K,” where 0 means no neighbors for an element of A, and “K” means that all the elements of the set B can be found below the distance a from that specific Ai.

From the function “S,” another function can be created:

The previous equation can be applied in a summation to get the number of element A that have at least one neighbor.

Once defined the function for the set “A,” this function can be reversed to count the number of neighbors of type A around a given element of the set B.

Indeed, to take into account the interaction between A to B, and the interaction between B to A, another function, “M” (referred to as mutual or mutual frequency), can be defined:

C(A, B) + C(B, A)

M(A, B)

N + K

This function can then be extended to multiple sets and apply these metrics to any kind of pair, changing starting point/ending point and the radius-diameter of interest (20 pm, 45 pm, 100 pm).

Using precise cell coordinates and phenotypes, it is possible to digitally represent them across the tissue. Particularly, to visualize densities of niches in the digital reconstructed tissue, the coordinates of such niches and a 2D Kernel Density Estimation (KDE) were used. The KDE allows us to approximate the density distribution of specific types of niches (or any other kind of species in a 2D surface) and shows the equiprobability lines for such densities. It is important to notice that the highest is the number of niches, and the highest will be the overlap between the equiprobability lines and the actual tissue.

Tissue processing

Resected tumors were chopped into 1-2 mm² pieces and, along with post-infusion biopsies, cryopreserved in 90% human serum + 10% dimethyl sulfoxide (DMSO), and additional pieces were snap-frozen for bulk RNA extraction. For single-cell experiments, both frozen and fresh materials were used as starting materials. PBMCs were isolated from blood collected in EDTA tubes and cryopreserved in 90% human serum + 10% DMSO.

On the day of the assay, path frozen pieces were thawed in RPMI + 10% FBS and chopped in small pieces using a scalpel. Tissue was dissociated in RPMI + 2% Gelatin (#G7041, Sigma- Aldrich) + 200 lU/mL Collagenase I (#17100-017, ThermoFisher Scientific) + 400 lU/mL Collagenase IV (#17104-019, ThermoFisher Scientific) + 5 lU/mL Deoxyribonuclease I (#D4527, Sigma-Aldrich) + 0.1% RNasin Plus RNase Inhibitor (#N2618, Promega) for 15-30min (depending of sample size and consistency) at 37°C and shaken at 160rpm. Digested cells were filtered using a 70 pm strainer and resuspended in PBS + 1% Gelatin + 0.1% RNasin. Cells were manually counted with a hematocymeter and then stained for viability with 50uM/mL of Calcein AM (#C3099, Thermo Fisher Scientific) and FcR blocked (#130-059-901, Miltenyi Biotec) for 15min at RT. After incubation and washing, cells were stained with CD45-APC (#304012, BioLegend) for 20min at 4°C. After washing, cells were resuspended in PBS + 0.04% BSA (Sigma- Aldrich) + 0.1% RNasin and DAPI staining (Invitrogen) was performed.

Bulk RNA sequencing library preparation and processing

Total RNA was extracted from snap-frozen tissues. Samples were lyzed in TRIzol reagent (Invitrogen) using a tissue lyzer (Qiagen), and the RNA was purified with the RNeasy kit (Qiagen). RNA quality was assessed with a Fragment Analyzer (Agilent) and Nanodrop spectrophotometer (Thermofisher). Quantification was performed with the Qubit HS RNA assay kit (InVitrogen). RNA sequencing libraries were prepared using the Illumina TruSeq Stranded Total RNA reagents according to the protocol supplied by the manufacturer and sequenced using HiSeq 4000. Illumina paired-end sequencing reads were aligned to the human reference GRCh37.75 genome using STAR aligner (version 2.6.0c) and the 2-pass method as briefly follows: the reads were aligned in a first round using the —runMode alignReads parameter, then a sample-specific splice-junction index was created using the —runMode genomeGenerate parameter. Finally, the reads were aligned using this newly created index as a reference. The number of counts was summarized at the gene level using htseq-count (version 0.9.1). The Ensembl ID was converted into gene symbols using the biomaRt package, and only protein-coding, immunoglobulin, and TCR genes were conserved for the analysis. Read counts were normalized into reads per kilobase per million (RPKM), and log2 transformed after addition of a pseudo-count value of 1 using the edgeR R package. As the data came in three different batches, a batch correction algorithm using the ComBat function of the sva R package was applied by using the patient origin as a covariate in the model.

FACS sorting, encapsulation, and library construction

40,000 CD45⁺ cells and 40,000 total live cells were sorted on a MoFlo Astrios (Beckman Coulter) and collected in separated 0.2mL PCR tubes containing lOpl in PBS + 0.04% BSA + 0.1% RNasin. After sorting, cells were manually counted with a hemocytometer, and viability was assessed using Trypan blue exclusion. Ex vivo CD45 cells from tumor were resuspended at a density of 600-1200 cells/p when possible with a viability of >90% and subjected to a lOx Chromium instrument for single-cell analysis. Single-cell RNA libraries were generated using the Chromium Next GEM Single Cell 5’ Library and Gel beads kit vl. l for the CD45-sorted population or using the Chromium Next GEM Single Cell 3’ Library and Gel beads kit v3.1 for the live-sorted cells, according to the manufacturer’s instructions. For each sample, 15’000 cells were loaded into the Chromium machine when possible (z.e., the total of sorted cells if <15’000), encapsulated, and barcoded following the manual (CG000388 for 3’ or CG000207 for 5’ technology). After encapsulation and reverse transcription, 14 PCR cycles were used to amplify cDNA. All library construction steps were performed according to the manufacturer’s protocol. For each sample, 3’GEX and 5’GEX enriched libraries were generated. Complementary DNA and library quality were examined on a Fragment Analyzer (Agilent), and quantification was performed with the Qubit HS dsDNA assay kit (Invitrogen). For the FACS sorting of post-ACT biopsies, depending on the percentage of CD45⁺ fraction in the sample, two sorting strategies were adopted. If the CD45⁺ fraction was inferior to 75% of total viable cells, 40,000 CD45⁺, and 40,000 CD45" viable cells were sorted in two separate tubes. If the CD45⁺ fraction was superior to 75% of total viable cells, 40,000 total viable cells were collected in one tube. All cells were sorted on a MoFlo Astrios (Beckman Coulter) and collected in separated 0.2mL PCR tubes containing lOpl in PBS + 0.04% BSA + 0.1% RNasin. After sorting, cells were manually counted with a hemacytometer, and viability was assessed using Trypan blue exclusion. If CD45⁺ and CD45" were sorted separately, cells were mixed with a ratio of 70% CD45⁺ and 30% CD45". All cells were resuspended at a density of 600-1200 cells/mL, if possible, with a viability >90% and subjected to a lOx Chromium instrument for single-cell analysis. Only the CD45⁺ compartment was analyzed in this study.

Alignment, filtering, and processing of scRNA-seq libraries

The following scRNAseq GEX datasets were analyzed: 1) 3’GEX from baseline sorted viable cells (total TME dataset) of 10/13 patients, which retained cell stoichiometry of the total TME, 2) 5’GEX from baseline CD45⁺-sorted cells of 13/13 patients which permits relative and deep phenotyping of only immune cells, including rare populations and 3) 5’GEX from CD45⁺- sorted cells of 7/13 patients at day 30 post TIL-ACT which enabled tracking of the dynamics of immune cell infiltration post TIL-ACT.

The scRNA-Seq reads were aligned to the GRCh38 reference genome and quantified using cellranger count (10X Genomics, version 4.0.0). Filtered gene-barcode matrices that contained only barcodes with a unique molecular identifier (UMI) counts that passed the threshold for cell detection were used and processed using the Seurat R package version 4.0.1. Two different Seurat objects were created: one containing CD45⁺-sorted cells from tumors (15 baseline tumor samples from 13 patients and 7 samples for post-ACT tumors from 7 patients using 5’ sequencing technology) and one containing all viable cells from baseline tumors (3’ technology of 12 samples from 10 patients using 3’ sequencing technology. The baseline tumor samples involving two different sites for the same patient (patients 10 and 13) were pooled for subsequent analyses unless otherwise mentioned.

For CD45⁺-sorted cells from tumor data (5’GEX), low-quality cells containing more than 10% of mitochondrial reads were defined and removed. A table summarizing the number of cells per sample before and after filtering appears below. For CD45⁺-sorted cells from tumor data, the number of genes expressed per cell averaged 1,240 (median: 1,458), and the number of unique transcripts per cell averaged 4,206 (median: 2,926). For sorted viable cells from tumor data (3’GEX), the number of genes expressed per cell averaged 2,136 (median: 1,811), and the number of unique transcripts per cell averaged 7,297 (median: 4,961). Cells from the baseline sorted viable cell dataset (3’) were not filtered out as this dataset was mainly built to derive the stoichiometry of cell types.

QC metrics and summary of cell number in single-cell RNA sequencing data The full baseline sorted viable cell dataset (3’) contained real stoichiometry for all cell types except for patient 1 where the scRNAseq library was enriched for malignant cells to reach an approximate rate of CD45⁺ and non-CD45⁺ cells of fifty percent each for a better characterization of the malignant compartment. Thus, for analyses involving stoichiometrical comparisons, the real stoichiometry of cell types was deduced and used by forcing the CD45⁺ cells to represent 93.9% (value obtained by FACS analysis) of the total number of cells. Single-cell clustering analysis

The data was processed using the Seurat R package (version 4.0.1) as follows: data counts were log-normalized using the NormalizeData function (scale. factor=10, 000) and then scaled using the ScaleData function by regressing the mitochondrial and ribosomal rate of read contents, the number of read count per cell (nCount RNA), and cell cycle parameters represented by S phase and G2/M phase scores (computed using the CellCycle Scoring function with the list of genes provided internally in the Seurat package). Dimensionality reduction was performed using the standard Seurat workflow by principal component analysis (RunPCA function) followed by tSNE (RunTSNE function with tsne.method=“Rtsne”) and UMAP projection (RunUMAP function with min.dist = 0.75) using the first 75 principal components. The k-nearest neighbors of each cell were found using the FindNeighbors function run on the first 75 principal components (nn.eps=0.5), and followed by clustering at several resolutions ranging from 0.1 to 10 using the FindClusters function. Unless otherwise mentioned, default parameters were used.

After the annotation of cells (see next section), the data was integrated by main cell type (CD8⁺ T cells with NK cells, myeloid cells, B cells) following an integration procedure using anchors and as follows: expression data Seurat objects were split per sample and then normalized individually using NormalizeData. The most variable features were individually identified using the FindVariableFeatures function. The 800 best genes selected for running the integration were found using SelectlntegrationFeatures on all split Seurat objects together. Every Seurat object was then scaled a PCA was run individually using only the selected features. Anchors were then found using the FindlntegrationAnchors function (reduction.method=“rpca” and 30 principal components), and these anchors were used for integration by using the IntegrateData function. This merge integrated object was then subjected to dimension reduction (20 principal components) and clustering, as explained above. For B cells integration, samples with too few B cells (patient #3, #5, #6, and #11) could not be integrated and were thus removed from the data.

Annotation of single cells

Main cell-type annotation

The annotation of main cell types was performed on both datasets using the combination of several different methods: (i) differential gene expression of clusters at different resolutions using the Seurat FindAllMarkers function followed by literature curation; (ii) investigation of the expression of known canonical gene markers for melanoma cells (MLANA, PRAME, SOXIO, S100B), immune cells (PTPRC), T cells (CD3E, CD8A, CD8B, CD4), B cells (CD79A, MS4AT), CAFs (DCN, FAP), endothelial cells PECAM1, VWF), plasma cells (immunoglobulins), Myeloid cells (CD68, HLA-DRA, LYZ, CD86), plasmacytoid DC (LILRA4),' (iii) automated annotation using the singleR package and using gene expression centroids (average gene expression profiles per cell type) derived from several studies: Yost et al. (K. E. Yost et al., Nat Med 25, 1251-1259 (2019)), Guo et al. (X. Guo et al., Nat Med 24, 978-985 (2018)), Zhang et al. (L. Zhang et al., Nature 564, 268-272 (2018)), Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)), Zilionis et al. (R. Zilionis et al., Immunity 50, 1317-1334 el310 (2019)) and the human primary cell atlas (HPCA) reference as inferred in the singleR package.

Malignant cell annotation

The malignant cell annotation from the sorted viable cell dataset was also confirmed by using two other methods. In the first method, a signature of genes specific to melanoma was derived using pan-cancer normalized bulk RNA sequencing data from the Cancer Genome Atlas (TCGA) (https://gdc.cancer.gov/about-data/publications/pancanatlas). Differential gene expression was performed using the regularized linear model as implemented in the Umma R package to extract the 50 most discriminant genes between melanoma (SKCM) and all other cancer types together. A melanoma-specific score was computed using the AUCell R package to ensure that private clusters found in the sorted viable cell data were melanoma using they were significantly higher than the score of normal CD45⁺-positive cells.

In the second method, the copy-number variation (CNV) was inferred using Copy KA T R package. For computational optimization, the sorted viable cell dataset was subsampled by randomly selecting 200 cells from each cluster at resolution 0.1. Then CopyKAT was run using the normal cells (immune, stromal, and endothelial) as reference. Inferred CNV profiles of 200 randomly selected cells per cluster exhibited clear CNVs, indeed confirming that these cells were of malignant origin. The copy-number alterations (CNA) were extracted from the CopyKAT output and represented as heatmaps by sorting genes per chromosomal location and cells per cell type using pheatmap R package. The number of genes falling in amplified (log ratio of the CNV > 0.2) or deleted (log ratio of the CNV < -0.2) per patient was then computed. Genomic locations of CNA regions as given by CopyKAT output were extracted, and the “full. anno” database consisting of genomic locations annotated with gene symbols was loaded. The overlap between amplified and deleted regions and the “full. anno” database was computed using the findOverlaps function from the IRanges R packages. The number of genes that were deleted, amplified or the sum of both (all CNV) per patient was then plotted using the violin plot function from the plotrix R package. Whether deleted and amplified genes were cancer driver genes was also checked by intersecting the gene list with a list of 204 oncogenic and tumor suppressor cancer driver genes specific for melanoma or pan-cancer that were extracted from Bailey et al. (M. H. Bailey et al., Cell 173, 371- 385 e318 (2018)).

High-resolution cell type annotation

Once major clusters were annotated for both datasets (malignant, T cells, B cells, myeloid cells, CAFs and endothelial cells), detailed and curated cell annotation was performed on the CD45⁺-sorted single-cell dataset, which has a higher resolution of the tumor microenvironment and then projected transcriptomics profiles on the sorted viable cell dataset for annotation as explained below.

High-resolution cell type annotation of T cells and NK cells

For finer annotation of the T/NK cells, the cells were first classified as CD8-positive, CD4- positive, double-negative (DN), double-positive (DP), NK cells, and Ty8 as follows: cells with non-null expression of CD8A and null expression of CD4 were defined as CD8-positive (and vice- versa for CD4-positive). Cells showing non-null expression of both genes were first classified as DP, then as doublets of CD4⁺ and CD8⁺ T cells, as the average number of genes expressed per cell equaled close to the double of CD4⁺ or CD8⁺ T cell singlets. Due to notorious dropout events in single-cell data, cells lacking the expression of both markers were classified as follows: if a cell belongs to a cluster (taking a fine resolution of 10) in which the 75^th percentile expression of CD8A was higher than its 75^th percentile expression of CD4, it was classified as CD8-positive (and vice- versa for CD4-positive cells). If the 75^th percentile expressions of both markers equal 0, the cells were classified as DN. Finally, cells with average expression scores of all TRG and ZKD-related genes higher than 0.5 (cutoff established after histogram visual inspection) were assigned to be Ty6 cells. The T-cell repertoire of CD45⁺-sorted cells was also characterized by scTCR-sequencing (VDJ) and compiled additional Ty6 cells for which TCR gamma or delta chains were found. In the general clustering of all CD45⁺-sorted cells, NK cells were clustering close to the CD8⁺ T cells and were annotated as NK1 and NK2 using the Zilionis et al. (R. Zilionis et al., Immunity 50, 1317-1334 el310 (2019)) centroid annotation by the singleR function from the singleR package.

The clustering of CD4⁺ T cells was obviously formed by three distinct clusters whose gene markers indicated CD4 CXCL13 (T follicular-helper) cells (CXCL13⁺, CD40LG⁺, BCL6⁺, € )200 ), Tregs (FOXP3 , CTLA4⁺, IL2RA⁺) and CD4 T helper 1 (Thl) cells (IL7R , SELL⁺, LEFT).

For the CD8⁺ T cell subtyping, CD8⁺ T cells and NK cells were integrated by sample as explained above and by removing the TCR genes in order to prevent clustering based on clonotypes. Several methods were employed to annotate the clusters: (1) differential gene expression and differential regulon/TF activity (see below) analysis was performed. The differential regulon/TF activity (see below) analysis was computed using the FindAllMarkers function with a 0.7 resolution. (2) signature score (using the AUCell R package) was computed for cytotoxicity (CCL3, CCL4, CST7, GZMA, GZMB, IFNG, NKG7, PRF1), exhaustion (CTLA4, HAVCR2, LAG3, PDCD1, TIGIT) and naiveness/memory (CCR7, LEF1, SELL, TCF7) taken from the Table S3 of Jerby-Amon et al. list of genes (L. Jerby-Arnon et al., Cell 175, 984-997 e924 (2018)). The gene list was extracted from Figure 7 of Andreatta et al. (M. Andreatta et al., Nat Commun 12, 2965 (2021)) as specific markers. The prediction from the annotation coming from public studies, namely Yost et al. (K. E. Yost et al., Nat Med 25, 1251-1259 (2019)), Guo et al. (X. Guo et al., Nat Med 24, 978-985 (2018)), Zhang et al. (L. Zhang et al., Nature 564, 268-272 (2018)), and Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)), were used.

A cluster displaying elevated levels of CCR7, LTB, SELL, and IL7R gene expression, high TCF7 and LEF1 regulon activity, high concordance with Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)) predictions of naive cells was then named CD8 naive-like. Two clusters were displaying signs of stress or apoptosis: one with elevated levels of mitochondrial-related genes (gene names starting by MT-), expression of the MALAT1 gene, and a lower number of genes expressed per cell was then named CD8 low-quality, and another one with overexpression of heatshock protein (HSP) genes and /WAZ-related genes which was named CD8 HSP. A cluster displaying concomitant expression of CD8A and FOXP3 genes, without any signs of a higher number of genes per cell (which could highlight doublets of CD8⁺ T cells and CD4 Tregs) and high concordance with Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)) predictions of Treg-like cells was then named CD8 FOXP3. A cluster was obviously driven by the high expression of type-I interferon genes (ISG15, MX1, IFI16, IFIT3, IFIT1, ISG20, OASF) and was then named CD8 type-I IFN. A small cluster was found very close to the one of NK cells and displayed CD8A expression with co-expression of NK cell markers such as KLRC2 and KLRD1, and high concordance with Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)) predictions of NK-like cells was then named CD8 NK-like. Another small cluster that was captured at a finer resolution of 2, displayed high expression of CX3CR1 and FGFBP2, high levels of cytotoxic and low levels of exhaustion signatures, high concordance with Guo etal. (X. Guo et a!., Nat Med 24, 978-985 (2018)) and Zhang et al. (L. Zhang et al., Nature 564, 268-272 (2018)) predictions of CD8 CX3CR1 cells was then named CD8 CX3CR1. Four clusters with similar gene expression profiles, displaying high expression of GZMK (a marker for effector memory cells according to Andreatta etal. (M. Andreatta etal., Nat Commun 12, 2965 (2021)), modest cytotoxic and low exhaustion signature levels, high concordance with Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)) predictions of CD8 Effector-Memory cells were then combined and named CD8 Effector-Memory.

Two clusters showing obvious signs of exhaustion with high levels of HAVCR2 and PDCD1 expression, high exhaustion signature levels, and elevated EOMES regulon activity were first classified as exhausted T cells, and then subclassified as follows: one cluster has higher HAVCR2, a typical sign of late exhaustion, and high concordance with Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)) predictions of Terminal -Exhausted cells and was then named CD8 Exhaustion; and the other cluster displayed higher levels of TOX, XCL1, XCL2 and CRTAM, which are markers of precursor exhaustion according to Andreatta et al. (M. Andreatta et al., Nat Commun 12, 2965 (2021)), lower activity of TBX21 regulon and high concordance with Oliveira et al. (G. Oliveira et al., Nature 596, 119-125 (2021)) predictions of Progenitor-Exhausted cells and was thus named CD8 Precursor Exhausted.

Finally, three clusters were driven by proliferation markers. Since proliferation is not a T- cell state in itself, an average expression profile (centroid) was generated for CD8⁺ T cell subsets, and the T-cell state of proliferating CD8⁺ T cells was predicted by automated annotation using this centroid and the singleR R package. The full list of differentially expressed genes and TFs per cell type was computed using the FindAllMarkers function from the Seurat package. Pseudotime analysis was performed using Monocle3 as inferred in the SeuratWrappers R library. High-resolution cell type annotation of myeloid cells

Myeloid cell subtyping was achieved by integrating per sample, as explained above. Several methods were used to annotate the clusters. First, differential gene expression was computed using the FindAllMarkers function with a resolution of 1. The prediction from the annotation coming from Zilionis et al. (R. Zilionis et al., Immunity 50, 1317-1334 el310 (2019)) and microphage subtyping with specific genes (I. Vazquez-Garcia et al., bioRxiv 2021.08.24.4545192021) was used. The dendritic cell (DC) was first annotated according to the predictions made using the Zilionis et al. (R. Zilionis et al., Immunity 50, 1317-1334 el310 (2019)) centroids. DC clusters were attributed to DC1, DC2, DC3, MonoDC, and pDC calls. It was also ensured that specific DC markers were expressed in the corresponding DC subsets (CLEC9A in DC1, CD207 in DC2, CCL22 in DC3, CLEC10A in MonoDC, and LILRA4 in pDC). The monocyte/macrophage component was annotated using the overlap between differential gene expression per cluster and specific genes extracted in Vazquez-Garcia et al. ((I. Vazquez-Garcia et al., bioRxiv 2021.08.24.4545192021)).

Clusters overexpressing the S100A8, TREM2, CXCL9, and Complement genes were found, for which the names Macro S100A8, Macro TREM2, Macro CXCL9, and Macro Complement were then attributed, respectively. A small population clustering close to the macrophages with high FCGR3A expression (the gene encoding CD 16) was annotated as CD 16 Monocytes. As for the CD8⁺ T cells, clusters with mitochondrial gene expression and a low number of genes expressed per cell were also isolated and attributed to be Macro low-quality, as well as a cluster overexpressing type-I interferon genes that were further named Macro Type-I IFN. A cluster with co-expression of both myeloid markers and T cell markers was isolated, for which the number of genes was significantly higher than for the macrophage, and attributed to Myeloid-T cell doublets. As for the CD8⁺ T cell, myeloid cell-specific average expression profiles and reannotated proliferating cells (cluster with high expression of cell cycle genes such as MKI67) using the singleR package were generated. No mast cell, neutrophil, or basophil was found in the data. The full list of differentially expressed genes and TFs per cell type was computed using the FindAllMarkers function from the Seurat package.

High-resolution cell type annotation of B cells B cell subtyping was achieved by integrating per sample, as explained above. Several methods were used to annotate the clusters. Differential gene expression was computed using the FindAllMarkers function with a resolution of 0.6. An immunoglobulin (Ig) signature score was computed (using the A UCell R package) by capturing all genes whose names started with IGH or IGL. Four major subtypes of B cells were found in the dataset and could be annotated using specific markers: Plasma cell overexpressing MZ 1, ./CHAIN, SDC1 and displaying high levels of the Ig signature. Naive B cells are characterized by high levels of FCER2, TCL1, and IGHD. Memory B cells were characterized by higher expression levels of CD27 and less naive B-cell markers. A germinal center (GC) population characterized by expression of CD38 and MEF2B was also identified. Isolated small populations clustered separately from the Memory B cells but nevertheless expressing memory B-cell markers (CD27). These clusters overexpressed specific immunoglobulin chains. It is likely that they correspond to clonally expanded B cells but kept their annotation as Memory B cells. As for the myeloid subtyping, it was found doublets of B cells, with both myeloid and T cells exhibiting more expressed genes per cell. Proliferating B cells were found and reannotated as described for myeloid and CD8⁺ T cells. In addition, low-quality B cells characterized by high expression of mitochondrial genes and a lower number of expressed genes per cell were also found. The full list of differentially expressed genes and TFs per cell type was computed using the FindAllMarkers function from the Seurat package.

Annotation of the full sorted viable cell dataset using CD45⁺-sorted annotation

To ensure the high consistency between the baseline scRNAseq datasets, differential gene expression was performed for major immune cell populations between Rs and NRs in both the CD45⁺-sorted (5’GEX) and total TME scRNAseq (3’GEX) datasets. Log fold-change correlations between the two scRNAseq libraries yielded highly similar DGE, thus allowing the CD45⁺- enriched data to be confidently utilized for the relative profiling and phenotyping of the immune compartment and the total TME data for the malignant phenotyping and quantification of cell stoichiometries. After the CD45⁺-sorted dataset (5’GEX) was annotated at higher resolution, averaged gene expression profiles (centroids) were then generated per cell subtype and used to annotate the sorted viable cell dataset (3’GEX) by the singleR package. Briefly, major cell populations (CD4⁺ T cell, CD8⁺ T cell, myeloid and B cells) were isolated, and automated annotation was performed using singleR by only using centroids from the specific populations (/. e. , CD4 Thl, CD4 CXCL13 and T Regs for the annotation of CD4 T cells). The final annotation was not kept as predicted by the package. Instead, fine resolution clusters were used, and each cluster was attributed to the cell type whose prediction was the most abundant.

Doublet annotation

In order to deconvolute the cell subtypes composing the doublets, averaged expression profiles derived from singlets were created and, using the singleR tool that annotates cells based on correlation with a pre-existing gene expression profile, the heterotypic doublet annotation was guided into single cells of finer subtypes (z.e., a doublet involving a T cell and a myeloid cell was annotated into both a T cell and a myeloid state).

Gene signature analysis

For single-cell data analysis and unless otherwise mentioned, gene signature scores were computed using the A UCell package. For bulk RNA sequencing data, gene signature scores were computed using single-sample geneset enrichment anylsis (ssGSEA) method as inferred in the gsva function from the GSVA R package (with mx.diff=FALSE and ssgsea.norm=FALSE parameters). Individual gene signatures were taken from cited publications, and collections were selected from the MSigDB portal (http://www.broadinstitute.org/gsea/msigdb); the Hallmarks and the Reactome collections were used. Differential analysis of gene signature scores was achieved using the regularized linear model as implemented in the Umma package. Enrichment Barcode plots were generated using the barcodeplot function from the Umma R package.

Transcription factor activity in single-cell data

The transcription factor activity was estimated using the regulon signature of each transcription factor. Regulons were inferred using the SCENIC pipeline (https://scenic.aertslab.org), which integrates three algorithms grnBoost2, RcisTarget, and AUCelT) corresponding to three consecutive steps:

Step 1 : First, a gene regulatory network (GRN) was inferred from all tumors and ACT products transcriptomic together using grnBoost2, a faster implementation of the original Genie3 algorithm. grnBoost2 takes as the input scRNAseq transcriptomics data to infer causality from the expression levels of the transcription factors to the targets based on co-expression patterns. Basically, the prediction of the regulatory network between n given genes is split into n different regression problems, and expression of a given target gene was predicted from the expression patterns of all the transcription factors using tree-based ensemble methods, Random Forests or Extra-Trees. The ranking of the relative importance of each transcription factor in the prediction of the target gene expression pattern is taken as an indication of a putative regulatory event. The aggregation targets into raw putative regulons were performed using the runSCENIC_l_coexNetwork2modules function from the SCENIC R package with default parameters.

Step 2: Co-expression modules (raw putative regulons, /.<?., sets of genes regulated by the same transcription factor) derived from the GRN generated in Step 1 were refined by removing indirect targets by motif discovery analysis using cisTarget algorithm and a cis-regulatory motif database. In particular, hgl9-500bp-upstream-7species.mc9nr.feather and hgl9-tss-centered- 10kb-7species.mc9nr.feather databases were used. The motif database includes a score for each pair motif-gene, which allows the generation of a motif-gene ranking. A motif enrichment score was then calculated for the list of transcription factor selected targets by calculating the Area Under the recovery Curve (AUC) on the motif-gene ranking using the RcisTarget R package (https://github.com/aertslab/RcisTarget). If a motif was enriched among the list of transcription factor targets, a regulon was derived, including the target genes with a high motif-gene score.

Step 3: Finally, AUCell was used to quantify the regulon activity in each individual cell (https://github.com/aertslab/AUCell). AUCell provides an AUC score for each regulon and cell; regulons with less than five constituent elements were discarded, as the estimation of the activity of small regulons is less reliable. For the calculation of the AUC, the parameter aucMaxRank of the AUCell calcAUC function was set with a fixed number of 1500 features.

Differential gene expression, regulon, and pathway analyses

During the process of annotating the cells, to find markers of cell population , the FindAllMarkers function from the Seurat package was used with the following parameters: only.pos = TRUE, min. pct = 0.25 and logfc.threshold = 0.25. To derive gene expression, regulon activity and pathway score differential analyses according to clinical status, linear regressions as inferred in the ImFit function of the Umma R package were performed. The mitochondrial and ribosomal contents were used as covariates in the regressions. Only the genes with average expression higher than a fixed threshold of 0.3 were kept in the final DEG table. No filtering was used for regulon and pathway analyses. In a differential analysis involving clinical status, the comparison of log fold-change analyses derived from single-cell based with those of patient-based data (by averaging the data) was reported. This approach helps to minimize the weight from variation of cell numbers among patients. This confirms that the single cell-based observations were representative of all patients of each clinical response category and were not driven by outstanding cases. Statistical /?-values for both single-cell and patient-based analyses are shown in Table 5.

Reactome pathway enrichment analysis

The 30 most upregulated or downregulated genes (computed by averaging logFC values from single-cell and patient-averaged analyses) according to clinical status were subjected to Reactome pathway enrichment analysis as follows: these genes were first converted into Entrez ID using maplds from the AnnotationDbi package then were subjected to Reactome enrichment analysis using the enrichPathway function ReactomePA package. Reactome pathways with q- values lower than 0.01 were kept.

Ligand-Receptor Interaction analysis

A methodology was developed to infer cell-cell interaction considering cell type stoichiometry and expression of ligands and receptors in both cell types, as described below. A database of ligands-receptors (LR) that was initially taken from the SingleCellSignalR package (LRdb.rda file) was used. Five different pathways were isolated from this database. In particular, the complement pathway was isolated by capturing LR pairs containing the word “Complement” in the “ligand. name” or “receptor.name” columns. Similarly, the Interferon and Interleukin pathways were isolated by using the words “Interferon” and “Interleukin” respectively. The costimulation and co-inhibitory gene list was extracted from Figure 1 of Chen et al. (Chen L., et al. Nat Rev Immunol. 2013 Apr; 13 (4): 227-42), and LR pairs containing these genes were isolated from the LR database. Similarly, Chemokine gene list was taken from Hornburg et al. (Hornburg, D., et al. Sci Rep 9, 17401 (2019)), and LR pairs containing these genes were isolated from the LR database. To ensure capturing all existing interactions, the LR pairs for the five selected pathways were then enriched using another database (human_lr_pair.rds) taken from the CellTalkDB package. LR pairs with the same receptor or ligand that were not present in the initial LR database were appended to it. Cell-cell interaction scores were computed based on the average expression of a ligand and its cognate receptors in two specific cell type (finer resolution) weighted by their corresponding proportions (real stoichiometry extracted from the sorted viable cell dataset (3’GEX)) and computed in every possible combination of ligands and receptors in all cell types as follows:

Interaction score = (Prop CellTypel')(AvgExpGenel in CelTypel^Prop CellType2~) AvgExpGene2 in CelType2~) where Prop is the proportion of this cell type in the full sorted viable cell dataset, and AvgExp is the average expression of the gene in this particular cell type.

Student’ s t-test was then run according to clinical status for all these possible combinations and extracted significantly different interaction scores between responders and non-responders (uncorrected p-value <= 0.05). Significant interactions were then selected for plotting using heatmap and circos plotting. These analyses were run in the sorted viable cell dataset for baseline tumor and in the CD45⁺-sorted dataset for post- ACT T30 data as sorted viable data was not available for this time point.

Plotting description and statistical analyses

All heatmaps were performed using the pheatmap function from the pheatmap R package. Violin plots were achieved by first plotting the violins using the violin plot function from the plotrix R package, followed by addition of symbols by using the stripchart function from the graphics R package. Plotting of UMAP and Dot plot showing expression of genes per cell subtypes were achieved using the DimPlot, respectively DotPlot function from the Seurat R package. UMAP highlighting gene expression or signature score in density was performed using the plot density function of the Nebulosa R package. UMAP showing the performance of two signature scores at the same time was plotted using the plot function from the base R package, and the script to convert the two gene signatures scores for exhaustion and cytotoxicity into a doublecolor code was taken from Wu et al. (T. D. Wu et al., Nature 579, 274-278 (2020)). Scatter plots, Ridge plots, and Fraction plots were generated using the ggplot2 R package. Alluvial plots were performed using the alluvial R package. Bar plots and line plots tracking points at different time points were generated using the default graphics R package. Circos plots were generated in R using the circlize package. Figures were reprocessed using Adobe Illustrator 2020 for esthetical purposes. The method used for statistical analysis appears directly in the method section related to the specific technique that was used or directly in the figure legend.

ROC analyses were performed using the prediction function from the ROCR package. The true-positive rate, false-positive rate, and area under the curve (AUC) were then extracted using the performance function of the ROCR package. Statistical tests used for specific analyses directly appear in the corresponding figure legends.

EXAMPLE 3

Steady state melanoma microenvironment of patients subjected to ACT

Thirteen patients with metastatic melanoma who received TIL-ACT after failing immune checkpoint blockade (ICB) therapy were analyzed. Patients received ex vivo expanded TILs and bolus intravenous IL-2 support following non-myeloablative chemotherapy. Patient characteristics are detailed in Supplementary Materials. Objective responses were observed in 6/13 patients (best overall response by RECIST v.1.1 46% at 3 months), resulting in median progression-free survival (PFS) of 5.6 months (95% CI 1.2 - 8.5), and median overall survival (OS) of 8.8 months (95% CI 7.5 - not reached, with a median follow-up of 30.2 months, IQR: 27.0 - 36.2), which compared favorably to a synthetic control arm from the same institution of clinically matched melanoma patients who received no TIL-ACT following ICB (median OS: 6.7 months). In particular, two patients obtained an ongoing complete response (CR) and four partial responses (PR) at three months (further grouped as Responders (Rs)), three stable disease (SD), and four with progressive disease (PD) at three months (further grouped as non-responders (NRs)).

To understand the baseline cellular landscape of tumors that responded or did not to TIL- ACT, 12 surgical tumor samples from 10 patients (6 Rs and 4 NRs) harvested at enrollment were interrogated. First, a cell-type map from scRNA-seq data was constructed, and 89,090 cells distributed into 21 major clusters were identified, including 17,965 T cells, 3,777 B cells; 4,276 myeloid cells, and other stroma such as vascular endothelial cells. Major-lineage non-malignant populations clustered well in UMAP space, and cells from different patients were found to be intermixed. The pan-cancer Cancer Genome Atlas (TCGA) was used to build a melanoma-cell specific gene signature score and identified 59,958 tumor cells, all of which clustered separately for each patient, indicating high patient specificity. Comparison of malignant or major immune cell type proportions according to clinical response did not yield any statistically significant differences. Even though NRs exhibited a lower trend in total TILs this was masked by high interpatient variability.

Baseline tissue of ACT responders exhibited tumor-intrinsic immunogenicity and genomic instability

The malignant compartment in baseline tumors was further interrogated. Copy-number variation (CNV), including amplification and deletions together, inferred from gene expression data, was significantly more prevalent in melanoma cells from Rs relative to NRs, indicating higher genomic instability. Notably, CNVs were mostly private, and no recurrent alteration of tumor suppressor or oncogenic drivers was found across patients. Consistent with genomic instability, activation of the DNA sensing/interferon pathway was observed, evidenced by STAT1, STAT2, STAT5A, IRF1, IRF2, and IRF9 emerging among the top activated regulons in melanoma cells of Rs by SCENIC analysis (S. Aibar et al., Nat Methods 14, 1083-1086 (2017)). Notably, SOX10 was among the top differentially activated regulons in responders, in agreement with histopathology evaluation, which demonstrated well-differentiated epithelioid-predominant melanoma enrichment in Rs. Corroborating the above inference, differential analysis of singlesample gene set enrichment analysis (ssGSEA) showed activation of immunogenic programs such as double-stranded (ds)DNA/IFN, interferon-a and -y response, immune checkpoints, antigenpresentation class I-II MHC and the complement system in melanoma cells of Rs. In agreement, by differential gene expression analysis, higher signatures of MHC class I-II antigen-presentation but also immunoproteasome activation in tumor cells from Rs were observed, as evidenced by the upregulation of B2M, TAPI, HLA-E,A,C, PSMB8, and PSMB9 (Table 5).

To search for melanoma-intrinsic programs associated with lack of response to TIL- ACT, a broader collection of Reactome signatures were interrogated. Among all, the MET-driven PI3K- AKT pathway and PI3K cascade driven by FGFR2, and resolution of D-loop structures, and DNA replication initiation gene expression programs, indicating DNA repair competence, were significantly overexpressed in melanoma cells of NRs, while FASL-CD95 and TRAIL apoptosis signaling pathways were overexpressed in tumor cells of responders. Notably, SIRT6 (NAD- dependent protein deacetylase sirtuin-6), required for genomic stability and known to control the expression of multiple glycolytic genes, was another hyperactive regulon in baseline tumors of NRs. Melanomas responding to TIL-ACT are enriched in pre-existing CD8⁺ TILs exhibiting features of cytotoxicity, exhaustion, costimulation, and type-I IFN signaling

Next, the TIL compartment at baseline was interrogated. The proportions of CD4⁺ and CD8⁺ TILs or CD45⁺ leukocytes annotated from scRNAseq data, were not significantly different between Rs and NRs. However, as shown by differential reactome pathway analysis, CD8⁺ TILs from Rs, overexpressed transcriptomic programs of IL-2, IL-27, and PD-1 signaling, type-I IFN and IFN-y activation, costimulation by the CD28 family, PEC AMI interactions denoting increased trans-endothelial migration, DNA amplification and repair indicating overall for a qualitatively superior TIL compartment in the tumors of Rs. In agreement, CD8⁺ TILs from Rs overexpressed genes of tissue residence and tumor reactivity (CXCL13, TNFRSF9 exhaustion (HAVCR2, CTLA4, PDCD1, ENTPDF), activation (HLA class-II genes), DNA repair (APOBEC3G), recruitment chemokines such as CCL4 and CCL5, adhesion to endothelium VCAM1 (Vascular cell adhesion molecule 1) and effector molecules such as PRF1 and NKG7. These pathways were regulated by TFs such as ZNF831 (Zinc finger protein 831) and HIVEP1 (Zinc finger protein 40), TBX21 (T-box transcription factor 21), EOMES, PRDM1 which encodes BLIMP-1; ETV1 and RUNX3, all TFs involved in the generation, activity and retention of antigen-experienced CD8⁺ TILs. Conversely, failure of TIL-ACT was associated with prevalent with naive-related and potentially bystander TILs at baseline, as highlighted by the upregulation of IL7R and LTB as well as higher activity of different TFs implicated in the WNT/p-catenin or TGF-P signaling pathways including LEF1, TCF7 and SMAD3, respectively (Figs. 1A-B, Table 5).

To gain more depth into the heterogeneity of the TIL compartment at baseline, scRNAseq data from Rs and NRs were interrogated together. A diverse group of transcriptional states were identified, including: nine different CD8⁺ TIL states: naive-like; effector-memory (EM); precursor exhausted (Pex); exhausted (Tex); heat-shock (HSP) genes⁺; FOXP3⁺; CX3CR1⁺; type-I interferon (IFN) activated; NK-like CD8 TILs; one NK-cell cluster and three CD4⁺ T-cell subsets: T-helper 1 (Thl); CXCL13⁺ T-follicular helper (Tfh)-like; and T-regulatory (Treg) cells. Clusters were annotated based on known marker genes and cross-referenced with previously described CD4⁺ and CD8⁺ T-cell clusters.

Then, CD8⁺ TIL state phenotypic divergence was validated by computing pseudotime differentiation trajectories. The data revealed that CX3CR1⁺ and Tex CD8⁺ cells were the most differentiated states by pseudotime progression, considering T naive-like as the starting state. In addition, CD8⁺ TILs branched across 3 main trajectories: interferon-stimulated genes (type-I IFN), cytotoxic/effector (CX3CR1⁺), and dysfunctional T-cell states (Pex and Tex).

Gene expression and prediction of regulons revealed the most distinctive genes and inferred TFs for each CD8⁺ or CD4⁺ TIL state. Reactome pathway analysis further revealed transcriptional programs characteristic of each CD8⁺ TIL cluster supporting the gene and regulon predictions. Naive-like, NK-like CD8⁺ TILs, and NK cells overexpressed highly IL7R, SELL, CCR7 genes and pathways of stemness/memory, regulated by TCF1, LEF1, and FOXP1, among others. NK cells also overexpressed TLR3/4/7/8/9 cascades and eicosanoid ligand-binding receptors signatures. Pex and Tex TILs overexpressed genes and pathways of activation and TCR reactivity (i.e., granzymes, PRF1, TNFRSF9, NKG7, CCL5, CXCL13) together with genes of increased carbohydrate metabolism and glycolysis (GAPDH, PKM), response to y_c family cytokines (IL2Ry and IL21K), transendothelial migration (VCAMF), tissue residence and retention (i.e., CXCL13, ITGAE, CRTAM, CXCR6, CXCR3) but also inhibitory molecules and exhaustion (PDCD1, HAVCR2, LAG3, TIGIT, CTLA4, TOX, and PROMT). Those were regulated by BATF, HIVEP1, EOMES, RUNX3, CREM, STAT3, PRDM1, and IKZF2 TFs. Compared to Tex, Pex TILs displayed higher regulation mediated by MYC, E2F2, MYODI, HDAC2 chromatin modulators, metabolic enzyme ENO1 regulating glycolysis and gluconeogenesis, but displayed down-regulated activity of TBX21, CREM, RUNX3, and ETS1. In agreement, Pex TILs overexpressed the most DNA amplification and DNA repair pathways as well as proliferation signatures. Meanwhile, type-I IFN CD8⁺ TILs overexpressed a dominant signature of IFN response genes and TFs such as IRFs and STATs. CD8 CX3CR1 TILs overexpressed cytotoxicity genes and signatures but downregulated exhaustion programs, in agreement with their pseudotime trajectory.

When focusing on the CD4⁺ compartment, CD4 Thl TILs were associated mostly with stemness/memory signatures, IL7R, FOS, JUN genes, and SMAD3, TCF7, and MYC TFs. CXCL13⁺ CD4 TILs overexpressed CXCL13, glucocorticoid receptor NR3C1, TOX and displayed signatures of exhaustion, CD28 costimulation, and TCR signaling regulated by NR3C1, MYODI, PPARG, STAT3, and PRDM1. Tregs overexpressed F0XP3, IL2RA, TNFRSF family molecules (IB, 4, 9, 18), TIGIT, and CTLA4 as well as IL32, BATF, STAT1 and displayed signatures of activation and exhaustion regulated by IKZF2, BATF, ETV7, and HIVEPL In the scRNAseq analyses, highly proliferating Ty6 cells were identified, which overexpressed gamma (TRGC2) and delta (TRDV1/TRDC) TCR chains, KLR molecules (KLR D1IK1IC2IC4IG1 CCL5, NKG7 and granzymes regulated by RUNX3, TBX21, EOMES, XBP1.

When analyzing the abundance of each TIL state, a significant enrichment in CD8 Tex and Pex TILs among CD45⁺ cells of Rs was found, consistent with their antigen-experienced, TCR- activated and CD28-costimulated phenotype. Albeit not statistically significant, CD8 type-I IFN, CD4 CXCL13, and CD4 TRegs were also higher in the baseline tumors of Rs, indicating an active involvement in T cell responses. On the contrary, the TIL compartment of NRs was slightly enriched in NK cells and CD8 NK-like T cells and significantly enriched in Ty6 cells. Validating their potential contribution to clinical response, CD8 Tex and Pex were the CD8⁺ TIL states with the highest overexpression of the CD8⁺ T-cell clinical response signature. These findings were corroborated by multispectral immunofluorescence (mIF) microscopy and immunohistochemistry (H4C) interrogation of baseline tumors, showing more intratumoral CD3⁺/CD8⁺ TILs expressing PD-l or CXCL13 in Rs.

Thus, in situ T-cell activation, cytotoxicity, costimulation, and exhaustion, along with immunogenic tumor cells are hallmarks of response to TIL-ACT.

TIL-ACT responding melanoma is infiltrated by activated macrophages and dendritic cells

The important role of tumor-resident B cells in supporting antitumor T-cell responses as well as response to ICB in melanoma and other cancers is being increasingly recognized. In baseline tumors, three tumor-infiltrating follicular ( S'- A/CD2CT) B-cell states were identified, including: naive (TCL1A⁺, FCER2⁺, and IGHI y.. several clusters of memory B cells (CD27⁺\ some of them driven by BCR chain expression (JGHV3-66, IGHV3-48, IGHV3-20, IGHV4-3iy, one germinal center (GC) B cells (CD38⁺/A7E 2 ⁺); plasma cells (MZBl⁺y Gene expression and regulon prediction revealed the most distinctive genes and regulons for each B-cell state. GC B cells overexpressed signatures of BCR signaling, class-II antigen presentation and proliferation. Memory B cells displayed high expression of IFN signaling and antigen-presentation cells (APC) maturation genes like CD80, CD86, and CD40. A comparative B cell differential gene expression between Rs and NRs revealed that those from Rs overexpressed PD-1 and interferon signaling, costimulation by the CD28 family as well as MHC class-II antigen presentation, particularly in memory B cells (data not shown). Despite the presence of these functional B-cell phenotypes, there is no difference in their proportions between Rs and NRs.

Recently, T-cell networks involving intratumoral antigen-presenting myeloid cells overexpressing CXCL9 were uncovered (D. Dangaj et al., Cancer Cell 35, 885-900 e810 (2019)), supporting effector functions of Tex in ovarian cancer by providing CD28-mediated costimulation. Thus, it was sought to understand the myeloid compartment in the context of ACT. Differential gene expression analyses from pseudobulked scRNAseq revealed that M(|)s of tumors from Rs significantly overexpressed genes and pathways for activation of complement (C1QA-C, C3), interferon signaling (JFI27, IFITM1, IFITM3, IFI6, STA T I), IFN-inducible T-cell recruiting chemokines CXCL9 and CXCL10, and MHC class-I (HLA-B, HLA-C, B2M) and class-II (HLA- DQA1, HLA-DQA2, HIA-DRB5) antigen presentation and processing (Fig. 2, Table 5). Similarly, dendritic cells (DCs) from Rs overexpressed IRF4-7, JUNB, CXCR4, ICAM1, and STA T representing mostly interferon response signaling (Fig. 3, Table 5).

To comprehensively capture the diversity of myeloid-cell compartments, they were subclustered to 11 different transcriptomic states, including four DC subtypes (DC1, DC2, DC3, and pDC), a transitory Monocyte-to-DC subset (MonoDC) and six monocyte/macrophage (M(|>) states: CD16⁺ monocytes; CXCL9⁺; type-I IFN; S100A8⁺; TREM2⁺; Complement-associated macrophages). Gene expression and prediction of regulons revealed the most distinctive genes and inferred TFs for each myeloid state. To further understand their functionality, defined Ml- or M2- associated signatures or other pathways associated with antigen presentation and costimulation were analyzed. Notably, transitory Mono-DC and differentiated DC clusters 1, 2, and 3 overexpressed Ml-like but not M2-like gene signatures. DC2 expressed the highest levels of class- II antigen presentation and CD28 costimulation but low levels of IFNs, immune inhibitory and APC maturation signatures. DC3 highly overexpressed highly immune inhibitory and APC maturation signatures. CD16⁺ monocytes appeared undifferentiated with a lack of expression of the above signatures. Macrophage clusters CXCL9 and type-I IFN overexpressed both Ml and M2 signatures together with signatures of class-II antigen presentation and CD28 costimulation, IFN response, and immune checkpoint inhibitory molecules consistent with macrophage polarization by type-I IFNs and IFN-y. CXCL9 and type-I IFN also overexpressed CXCL9, 10, and 11, known chemokine ligands of CXCR3 receptor, key for T-cell recruitment and response to ICB. In contrast, macrophage clusters S100A8, TREM2, and Complement overexpressed only M2- but not Ml- associated signatures, pointing to rather discrete immunosuppressive phenotypes.

When comparing M(|) abundance, it was found that NRs exhibited higher proportions of major lineage Mono/M(|)s populations within their CD45⁺ compartment, while no differences were observed in the DCs compartment, which represented a smaller fraction of the immune TME. Similarly, no difference was observed in the proportions of finer M(|) or DC clusters between Rs and NRs, nor in the ratio of classically-defined Ml or M2 M(|), indicating that specific transcriptional programs rather than overall myeloid-cell states drive responses to ACT. In line with this, the M(|)-specific genes and signatures associated with clinical response selectively localized in CXCL9⁺ and type-I IFN M(|) subtypes, while they were underrepresented in CD16⁺ monocytes, TREM2⁺ and S100A8⁺ M(|). Similarly, DC-specific clinical response genes and signatures are selectively localized to the pDC subtype (Fig. 3).

These results indicate that myeloid- and potentially B cell-derived transcriptional programs of complement activation, interferon signaling, and IFN-inducible T-cell recruiting chemokines mediate T-cell response networks. These transcriptional networks acting on the malignant compartment, but also on B cells and myeloid cells, could recruit, elicit and stimulate a tumor- reactive pool of pre-existing TILs available for ex vivo IL-2-based expansion and in vivo ACT.

Melanoma responding to TIL-ACT are characterized by rich cellular crosstalk, while non-responders lack cell-to-cell communication

The above findings demonstrate important cellular interactions within the TME of melanoma tumors. To explore this, cell-to-cell crosstalks were computed by ligandome analysis which relied on putative receptor-ligand gene expression of two cell types and their proportions in the TME and revealed far more cell-to-cell interactions in tumors from Rs relative to NRs. The origin of these cellular cross-talks was identified. Malignant cells from NRs mostly interacted with endothelial cells and myeloid cells, specifically Macro S100A8 and DC2 but displayed no other interactions. However, in Rs, malignant cells interacted strongly with Tregs and CXCL13⁺ CD4, CD8 EM, and exhausted CD8⁺ TILs (Tex and Pex). In addition, Rs displayed interaction clusters involving CD8⁺ TILs and macrophages, CD4⁺ TILs and macrophages, CD4⁺ and CD8⁺ TILs.

To increase resolution in the interactome, the ligandome analysis was carried out in five selected pathways known to be involved in T cell networks: interferon, complement, chemokines, interleukins, and costimulation/co-inhibition. Again, cell-to-cell interaction analyses summarized in major cell types or in finer cell states revealed statistically significant and denser putative interactions in Rs and between TILs and myeloid cells, TILs and tumor cells but also CD4⁺ and CD8⁺ TILs. Those interactions occurred mainly in the chemokine, complement and interferon pathways. Strikingly, barely any crosstalk through the five signaling pathways was detected in NRs, with the only predicted interactions occurring between malignant cells and DC2, Macro SI 008 A and Macro TREM2 clusters and mostly involving complement signaling.

Having described that TCR-engaged CD28-costimulated/exhausted TILs exhibit increased effector fitness when in close proximity with tumor-resident mAPC, it was sought to unravel the states of interacting CD8⁺ TILs and myeloid cells. Exhausted CD8⁺ TILs of responders exhibited striking interaction levels with all macrophage clusters, especially with Macro Complement and Macro CXCL9 populations and at a lower extent with DC1, DC3, and pDCs. CD8 EM and Naive- like CD8⁺ TILs had higher interactions with Macro TREM2 but also Macro Complement and Macro CXCL9, while NRs totally lacked such interactions (Fig. 4A).

Specific examples of ligand-receptor interactions occurring in Rs included ICAX1L ICAM2, CCL5, CXCL9, and CXCL13 expressed in Macro CXCL9 with ITGB2, ITGAI., CCR5, CXCR3, and CXCR5 respectively expressed in CD8 Pex and Tex TILs indicating higher recruitment and tighter cellular adhesion between interacting populations (Fig. 4B). It was predicted higher interactions between IL15RA expressed in Macro Complement and IL12RG and IL15RA overexpressed in CD8 type-I IFN under exhausting conditions (Fig. 4B). In addition to these canonical LR interactions, less documented interactions between C3 expressed in T cells and C3AR1 expressed in Macro Complement and CXCL9 were also discovered (Fig. 4B). Overall, the results indicate that melanoma of Rs has constructed highly recruiting and activating TIL-myeloid networks.

In the process of sc annotation, naturally occurring physical doublets between T cells and myeloid cells were detected, as well as between B cells and T cells. CD4:CD8 doublets were also identified. In line with higher inferred cellular crosstalk from singlets, a higher frequency of T myeloid cell doublets within the myeloid compartment of Rs and a similar trend in the frequency of T:B and CD4:CD8 cell doublets were found. Importantly, myeloid:T-cell doublets from Rs, overexpressed exhaustion and costimulation signatures, B: T-cell doublets exhibited higher expression of costimulatory and tumor-reactivity scores while CD4:CD8 T-cell doublets overexpressed scores of cytotoxicity and tumor reactivity implying that TIL cell states residing in these doublets could be highly relevant for ACT. By deconvoluting cell states involved in T myeloid doublets, it was found that in Rs, those were enriched in Macro CXCL9 and exhausted Pex and Tex CD8⁺ TILs, while T:myeloid doublets of NRs were enriched in DC2 and CD8 NK- like states.

Considering that cell doublets represent true cellular interactions that remained intact or reformed after tissue dissociation, these cell doublets were searched in baseline tumors in situ using mIF, examining total cell densities and mutual cell-to-cell distances among CD8⁺/PDl⁺ or CD87PDl⁺ -thus exhausted- TILs, CDl lc⁺DCs, CD68⁺ macrophages, and CD19⁺B cells. Higher levels of cell neighborhood interactions in Rs were detected compared to NRs, in particular pairs between overall CD8⁺ or CD8⁺/PD1⁺ T cells and CDl lc⁺ cells, widely distributed both in tumor islets and in the stroma compartment. Furthermore, CD8⁺/PD1⁺:CD68⁺, CD8⁺/PD1⁺:CD19⁺ or CD8⁺/PD1⁺:PD1⁺/CD8‘ (z.e., CD4) pairs were also higher in the stroma of Rs.

These analyses revealed that the baseline TME of Rs is highly interconnected, as opposed to NRs, and involves intratumoral T celkmyeloid niches as well as other T-cell networks which sustain the ability of tumor-reactive TILs to exert antitumor activities and persist in the TME.

Effective ACT-TIL therapy reprograms myeloid populations and reconstitutes antitumor CD8 TIL-myeloid cell networks

Next, it was sought to determine how TIL-ACT affects TME dynamics. Thus, comparatively baseline and post TIL-ACT were interrogated. Bulk RNAseq as well as scRNAseq from tumors at baseline (TO) and biopsies acquired 30 days post ACT (T30) were performed. Reactome pathways analyses of bulk RNAseq data in pairwise patient comparisons revealed divergent changes in the TME between Rs and NRs. Whereas ERBB2, ERBB4, and PI3K signaling pathways were downregulated in Rs, they were upregulated in NRs, possibly reflecting tumor cell expansion or adaptation to TIL-ACT. Importantly, responding tumors exhibited an increase in the nicotinamide salvage pathway; inflammatory signatures, including downstream signaling in the alternative complement activation, TLR3, NF-KB, IL-18, and IL-10 pathways; and T-cell activation, including PD-1 and CTLA-4, TCR, CD28 costimulation, and IL-2 signaling. The latter pathways were already lower in baseline melanoma of NRs relative to Rs, and were lost in NRs” T30 biopsies.

To infer evolution dynamics of TME cell subets, bulk RNAseq data at baseline and T30 were interrogated using gene signatures discriminating immune cell subtypes derived from the scRNAseq data. Importantly, in Rs, post-ACT, a reconstitution of some key TME hallmarks observed at baseline was detected, including CD8 Pex and Tex and CD4 CXCL13 TILs, in addition to an increase in CXCL9⁺ and type-I IFN macrophages, reflecting reprogramming of the myeloid compartment by TIL- ACT. In contrast, NRs lost signatures of CD8 Pex and Tex, CD4 CXCL13 TILs, Tregs, as well as DC2 (Fig. 5). These predictions were validated by applying curated independent gene signatures, indicating again a major loss of global T cells and exhaustion-related features post-ACT in NRs (Fig. 5).

These observations were corroborated by mIF of baseline and T30 tumor samples. Responders maintained their CD8⁺/PD1⁺/GZMB^+/" tumor reactive and polyfunctional TIL pool upon ACT. However, NRs had significantly lower tumor-reactive CD8⁺ TILs at day 30 post ACT. Notably, an overall decrease in intratumoral CD8⁺/PD1‘ -potentially bystander- TILs in both Rs and NRs was observed, which was attributed to lymphodepletion.

Since many myeloid cell types were increased at T30, whether cellular cross-talks were also reconstituted following TIL-ACT was investigated. This was addressed in paired baseline and T30 post-ACT biopsies from 7 patients. It was found that compared to baseline, Rs increased their CD8:myeloid-cell interactions at day 30, especially with CXCL9⁺ macrophages (Fig. 5). At day 30 post, denser networks were observed between several TIL states: naive, EM, Tex, Pex CD8⁺ TILs, CXCL13⁺ CD4 TILs with myeloid cells; EM, Tex, IFN, and CXCR3⁺ CD8 TILs with memory B cells as well as Pex with Thl CD4⁺ TILs. Similarly, as in baseline tumors, NRs displayed minimal interactions and mainly between CD16⁺ monocytes and monoDC with CD8 EM TILs. These results indicate that a successful ACT broadened the repertoire of TIL:TME interactions with polarization of CXCL9⁺ myeloid APCs as a hallmark of this interactome.

Some selected ligandome interactions between CD8 EM, type-I IFN, and CXCL9⁺ macrophages include IFNG and IFNGR1/2, indicating a direct macrophage polarization in producing CXCL9 and CXCL10 chemokines. IFNG and IFNGR1/2 interactions were also observed between CD8 Tex and memory B cells. CCL3I4I5 expressed by CXCL9⁺ macrophages were significantly interacting with CCRJ-expressing naive-like T cells, indicating a mechanism of their recruitment in the TME.

Finally, the above data indicate that these cellular cross-talks are a key architectural hallmark of the immunoreactive melanoma TME organization, culminating with the key observation that association of Tex or Pex with APCs at the steady state was associated with superior functional TIL states, which gave rise to successful TIL-ACT. Since this was an important feature at baseline, and it reconstituted after TIL-ACT specifically in responders, whether CD8- APC neighborhoods could, in fact, predict the outcome of TIL-ACT, providing a bona fide biomarker for suitable patient selection was investigated. By interrogating both single-cell and mIF data, interactions between CD8⁺ TILs and myeloid cells had better performance at predicting clinical response than their mere individual presence in this discovery cohort.

Thus, successful TIL- ACT leads to the elimination of melanoma through the establishment of a favorable TME, with re-engraftment of tumor-reactive TILs following transfer, and improvement of the myeloid compartment after immune reconstitution. The close association of exhausted CD8⁺ TILs with CDl lc⁺ DCs represents a potential powerful biomarker for patient selection.

Discussion

Adoptive cell therapy (ACT) using ex vivo expanded tumor-infiltrating T lymphocytes (TILs) can mediate responses in metastatic melanoma, but long-term efficacy remains limited to a fraction of patients. In this example, tumor-microenvironment (TME) cellular states and interactions of longitudinal samples from 13 metastatic melanoma patients treated with TIL- ACT in the clinical study (NCT03475134) were investigated. Single-cell (sc) RNA-seq and spatial proteomic analyses were performed in pre- and post-ACT tumor tissues and showed that responders exhibited higher tumor cell-intrinsic immunogenicity. Also, endogenous CD8⁺ TILs and myeloid cells of responders were characterized by increased cytotoxicity, exhaustion and costimulation and type-I IFN signaling, respectively. Cell-cell interaction prediction analyses corroborated by spatial neighborhood analyses revealed that responders have rich baseline intratumoral and stromal tumor-reactive T-cell networks with activated myeloid populations. Successful TIL- ACT therapy further reprogrammed the myeloid compartment and increased TIL- myeloid networks. This systematic target discovery study reveals CD8⁺ T-cell network-based biomarkers that can improve patient selection and guide the design of adoptive cell therapy clinical trials.

This example describes the most comprehensive single-cell profiling of longitudinal melanoma samples during TIL- ACT, providing insights into the cellular composition and T-cell interaction network that are associated with clinical response to TIL-ACT. The comparative analysis as described herein reveals that divergent phenotypic and functional malignant, but also immune TME states are already present at baseline. Baseline tissues of responders to adoptive cell therapy exhibit immunogenic tumor-intrinsic malignant cell states with higher predicted CNVs, DNA-sensing/IFN, and class-I antigen presentation-related transcriptomic profiles. The data also indicates that activating myeloid cell subsets conjointly participate in CD8⁺ T-cell networks in situ, elicit tumor-reactive TIL populations, and are able to expand under IL-2 ex vivo. Myeloid cells of Rs overexpressed antigen processing and presentation, costimulatory and complement genes, as well as type-I IFN signatures and CXCL9/10 chemokines. These myeloid compartments have been recently linked to effective ICB response and indicates the presence of CXCL13⁺ T cell in the TME. Indeed, melanoma tissues from Rs are associated with increased densities of poly functional intratumoral CD8⁺ T cells marked by higher expression of PD-1, CXCL13, and TNFRSF9 (CD 137) at steady state, and (neo-)antigen-specific, which persisted after TIL infusion.

In addition, in baseline melanoma responsive to adoptive cell therapy, a strong in situ tumor reactive CD8⁺ TIL crosstalk and adhesion with activating myeloid cells was observed. In contrast, the TME of NRs displayed minimal numbers of those interactions and almost no cellular crosstalk when focusing on pathways such as costimulation, complement, type-I IFN, and chemokines. These findings indicate that cellular crosstalk among key immune cell subsets is not only relevant to drive efficient responses to immune checkpoint blockade but also can be the underlying mechanism for successful TIL-ACT. These are hinged on the cooperation of CCL5, CXCL9, and CXC10 chemokines, which ensure recruitment but also retention of TILs in the TME.

Furthermore, by comparing the evolution dynamics of TIL and TME states from baseline to 30 day post ACT, it was found that in responders, (i) tumor-reactive but not bystander TILs were able to home to the tumor, and (ii) their interactome with myeloid cells increased and diversified including tumor-reactive TILs and other TIL states, indicating that effective ACT-TIL therapy reprograms myeloid populations and increases antitumor CD8⁺ TILs-myeloid cell networks. In turn, an immunologically interactive TME can sustain the persistence and reactivity of ex vivo re-educated and adoptively transferred T cells via chemokine-mediated retention and costimulation.

Notably, homotypic and heterotypic cell doublets with key qualitative features important for antitumor immune responses were identified by scRNA-seq analyses. In single-cell analyses, cell doublets are traditionally considered undesirable and are thus often filtered out from any subsequent analyses (N. J. Bernstein et al., Cell Syst 11, 95-101 el05 (2020)). While doublets are generally viewed as artifactual gene expression matrices generated from two cells (E. A. K. DePasquale et al., Cell Rep 29, 1718-1727 el718 (2019)), new evidence as presented herein demonstrates the relevance of immune cell interaction. It provides new approaches for the intentional investigation of physically interacting cells. The transcriptomic profiling of doublets revealed that CD4:CD8 TIL doublets from responders were characterized by high cytotoxicity and tumor reactivity programs, B:TIL doublets by high tumor-reactivity and costimulation, while TIL:myeloid doublets displayed increased costimulation and exhaustion features.

The disclosed detailed TME states and dynamics, as well as cell-cell interactome description, facilitate the development of biomarkers of response and guide the next generation of adoptive cell-based immunotherapies to achieve maximal clinical benefit. In addition, tissue signatures demonstrate the presence of intratumoral tumor-reactive TIL: myeloid niches with traits of polyfunctionality, fitness, and costimulation can be utilized to select patients that can maximally benefit from TIL-ACT approaches with traditional IL-2 -based TIL expansion methodologies.

Table 5. Differential Expressed Genes (DEG) per cell population (malignant) according to clinical response.

Positive logFC indicates expression values that are higher in responders and vice-versa. The “ sc” and “_patient” labels indicate values for which the analyses were performed at single-cell and patient-averaged levels, respectively.

Table 5 (continued). Differential Expressed Genes (DEG) per cell population (CD8 T cell) according to clinical response.

Table 5 (continued). Differential Expressed Genes (DEG) per cell population (macrophages) according to clinical response.

Table 5 (continued). Differential Expressed Genes (DEG) per cell population (Dendritic cells) according to clinical response.

The present disclosure is not to be limited in scope by the specific embodiments described herein. Indeed, various modifications of the invention, in addition to those described herein, will become apparent to those skilled in the art from the foregoing description and the accompanying figures. Such modifications are intended to fall within the scope of the appended claims.

Claims

CLAIMS What is claimed is:

1. A method of predicting responsiveness to a cancer therapy in a subject, comprising: determining an expression level of each of a set of biomarkers in a sample from the subject; determining a change in the expression level of each of the set of biomarkers as compared to a respective reference expression level for each of the set of biomarkers; determining a distribution of changes of expression levels of the set of biomarkers; and assessing a likelihood of a therapeutic response to the cancer therapy by comparing the distribution of changes of expression levels to a reference distribution of changes of expression levels, wherein the reference distribution of changes of expression levels is correlated positively or negatively with the therapeutic response to the cancer therapy, and wherein the reference distribution of changes of expression levels is associated with one or more characteristics in a tumor or tumor microenvironment thereof in the subject.

2. The method of claim 1, wherein the one or more characteristics comprising: increased tumor-intrinsic immunogenicity; increased genomic instability; increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ tumor-infiltrating lymphocytes (TILs); increased activation of macrophages or dendritic cells; increased cell-cell interaction; or reprogramed myeloid populations and reconstituted antitumor CD8 TIL-myeloid cell networks.

3. The method of claim 1 or 2, wherein the step of assessing the likelihood of the therapeutic response comprises identifying the subject as having an increased likelihood of the therapeutic response to the cancer therapy if the distribution of changes of expression levels of the set of biomarkers is identical to the reference distribution of changes of expression levels.

4. The method of any one of the preceding claims, wherein the reference distribution of changes of expression levels is determined from samples of one or more subjects who have responded positively to the cancer therapy.

92

5. The method of any one of the preceding claims, comprising determining the reference distribution of changes of expression levels by: determining an expression level of each of a plurality of biomarkers in each of the samples from a plurality of subjects who have responded positively to the cancer therapy; determining whether the determined expression level is increased, decreased, or unchanged as compared to a reference value for each of the plurality of biomarkers to provide a biomarker expression profile of each of the samples; performing an aggregated analysis on the biomarker expression profiles of the samples; identifying a group of biomarkers having increased or decreased expression levels as compared to the reference value; and determining a reference distribution of changes of expression levels of the group of biomarkers.

6. The method of claim 5, wherein the group of biomarkers are associated with the one or more characteristics in tumor microenvironments in the plurality of subjects.

7. The method of any one of claims 5 to 6, wherein the plurality of subjects have been administered the cancer therapy.

8. The method of any one of the preceding claims, wherein the respective reference expression level is determined from samples of one or more subjects who have not been administered the cancer therapy.

9. The method of any one of the preceding claims, wherein the change in the expression level of each of the set of biomarkers is an increase or decrease in the expression level.

10. The method of any one of the preceding claims, wherein the distribution of changes comprises an increase or decrease in expression levels of the set of biomarkers.

11. The method of any one of the preceding claims, wherein the increased tumor-intrinsic immunogenicity or genomic instability is characterized by increased copy-number variation.

93

12. The method of any one of the preceding claims, wherein the increased cytotoxicity, exhaustion, costimulation, or type-I IFN signaling in CD8+ TILs is characterized by overexpressed genes of tissue residence and tumor reactivity, exhaustion, activation (HLA class- II genes), DNA repair, recruitment chemokines, adhesion to endothelium, or effector molecules.

13. The method of any one of the preceding claims, wherein the increased activation of macrophages or dendritic cells is characterized by overexpressed genes and pathways for activation of complement, interferon signaling, IFN-inducible T-cell recruiting chemokines, class-II antigen presentation and processing, or CD28 costimulation.

14. The method of any one of the preceding claims, wherein the set of biomarkers comprise a first group of biomarkers expressed in a first population of cells and a second group of biomarkers expressed in a second population of cells, and wherein the first population of cells interact with the second population of cells.

15. The method of claim 14, wherein the first population of cells comprise myeloid cells, B cells, CD4 cells, or dendritic cells.

16. The method of any one of claims 14 to 15, wherein the second population of cells comprise T cells or CD8 cells.

17. The method of any one of claims 14 to 16, wherein the first population of cells comprise myeloid cells, and the second population of cells comprise T cells.

18. The method of any one of claims 14 to 17, wherein the first population of cells comprise B cells, and the second population of cells comprise T cells.

19. The method of any one of claims 14 to 18, wherein the first population of cells comprise dendritic cells, and the second population of cells comprise T cells.

94

20. The method of any one of claims 14 to 19, wherein the first population of cells comprise CD4 cells, and the second population of cells comprise CD8 cells.

21. The method of claim 19, wherein the T cells comprise progenitor exhausted T cells or CD8+ TILs.

22. The method of any one of the preceding claims, wherein the increased cell-cell interaction comprises increased myeloid: T cell interaction, increased B cell: T cell interaction, increased dendritic cell: T cell interaction, or increase CD4 cell: CD8 cell interaction.

23. The method of any one of the preceding claims, wherein the set of biomarkers comprise a first group of biomarkers associated with a first signaling pathway and a second group of biomarkers associated with a second signaling pathway.

24. The method of any one of the preceding claims, wherein the reprogramed myeloid populations and reconstituted antitumor CD8 TIL-myeloid cell networks are characterized by increased number of progenitor exhausted T cells, increased number of CD8+ TILs, increased number of CD4 CXCL13 TILs, increased number of CXCL9⁺ macrophages, increased number of type-I IFN macrophages, maintained number of CD8⁺/PD1⁺/GZMB^+/" tumor reactive and polyfunctional TILs.

25. The method of any one of the preceding claims, wherein the likelihood of the therapeutic response in the subject comprises complete or partial response as defined by response evaluation criteria in solid tumors (RECIST), stable disease as defined by RECIST, or long-term survival in spite of disease progression or response as defined by immune-related response criteria (irRC).

26. The method of any one of the preceding claims, wherein the cancer cell therapy comprises a cancer immunotherapy.

27. The method of any one of the preceding claims, wherein the cancer therapy comprises an immune cell therapy.

95

28. The method of claim 27, wherein the immune cell therapy comprises a T cell.

29. The method of claim 27, wherein the immune cell therapy comprises a tumor infiltrating lymphocyte (TIL).

30. The method of any one of the preceding claims, wherein the cancer therapy comprises an adoptive cell therapy (ACT).

31. The method of claim 30, wherein the adoptive cell therapy comprises a T-cell receptor (TCR) T cell therapy or a chimeric antigen receptor (CAR) T cell therapy.

32. The method of claim 30, wherein the adoptive cell therapy comprises an adoptive cell therapy with TILs.

33. The method of any one of the preceding claims, wherein the set of biomarkers comprise one or more biomarkers set forth in Tables 1-5.

34. The method of any one of the preceding claims, wherein the set of biomarkers comprise:

(c) one or more differentially expressed genes in macrophages selected from: IFI27, C1QB, C1QA, CCL4L2, C1QC, IFITM3, FCGR3A, STAT1, CCL3L1, HLA-C, SERPING1, LY6E, IFI6, GBP1, HLA-DQA2, PSAP, B2M, HLA-DQA1, CXCL10, VAMP5, IFITM1,

96 PLAAT4, CTSC, LGALS3BP, CXCL9, APOCI, PSME2, APOE, HLA-DRB5, HSPA8, HLA- B, WARS, GBP4, C3, NCF1, RPS4Y1, IER2, FN1, RPS21, RPS29, YBX1, and RPS2; or

35. The method of any one of the preceding claims, wherein the set of biomarkers comprise one or more biomarkers selected from:

(i) TOX, PKM, PRF 1 , LYST, TNFRSF9, ITM2A, GAPDH, PARK7, HAVCR2, CTLA4, PDCD1, SLA, CBLB, RGS1, KLRC2, STAT3, PHLDA1, GNLY, PTPN6, SH2D2A, GZMB, CD7, IFNG, CYTOR, SUB1, VCAM1, RBPJ, NPM1, APOBEC3C, EIF4A1, TPI1, MIF, LAG3, SAMSN1, DUSP4, CXCL13, ARPC1B, DYNLL1, ATP5MC2, CSNK2B, RPL12, SRGN, S100A4, CTSD, and FXYD5;

(iii) GZMK, AHNAK, IL32, CCL4, FOS, TSC22D3, CD52, GZMM, TXNIP, SEPTIN9, DNAJB1, ANXA1, LTB, SPOCK2, CD48, WIPF1, EMP3, ITM2C, CCNH, KLRG1, THEMIS, AOAH, PTPRC, TC2N, VIM, KLF6, ZFP36L2, CNN2, CYBA, CD69, SELPLG, LIME1, BIN2, SLC2A3, TRAT1, MBP, LCP1, KLRK1, TAPBP, ITGAL, LINC02446, TUBA4A, GZMH, KRT86, DDIT4, and SKAP1; or

(iv) MTRNR2L8, CD52, ANXA1, ZFP36L2, S100A10, VIM, BTG1, DUSP2, RPS29, GPR183, RPS2, LTB, EMP3, PLP2, RPL38, S100A4, IL7R, MTRNR2L12, SLC2A3, AHNAK, TAGLN2, CD44, RPL17, RPS21, TXNIP, FXYD5, TC2N, RPL27A, RPL39, AL138963.4, S100A11, EML4, ANXA2, HSPA1A, HSPA1B, DUSP1, DNAJB1, HSPE1, and DDIT4

97

36. The method of any one of the preceding claims, wherein the sample is obtained from neoplasia tissue, tumor microenvironment, or tumor-infiltrating immune cells.

37. The method of any one of the preceding claims, wherein the sample comprises a biological sample that comprises a plasma sample, a blood sample, or a tissue sample.

38. The method of any one of the preceding claims, wherein the sample is obtained from a primary tumor or a metastasis.

39. The method of any one of the preceding claims, wherein the sample comprises immune cells.

40. The method of claim 39, wherein the immune cells are selected from T cells, macrophages, dendritic cells, fibroblasts, NK cells, NKT cells, and NK-DC cells.

41. The method of any one of the preceding claims, wherein the sample comprises protein, DNA, or RNA.

42. The method of any one of the preceding claims, wherein the expression level comprises a mRNA or protein level.

43. The method of claim 42, wherein the mRNA level is determined by at least one technique selected from reverse transcription polymerase chain reaction (RT-PCR), competitive RT-PCR, real-time RT-PCR, ribonucleic acid sequencing (RNA-seq), immunohistochemistry (IHC), immunofluorescence, RNaseprotection assay (RPA), northern blotting, and DNA chip.

44. The method of claim 42, wherein the protein level is determined by at least one technique selected from western blot, enzyme linked immunosorbent assay (ELISA), radioimmunoassay (RIA), radioimmunodiffusion, Ouchterlony immunodiffusion, rocket immunoelectrophoresis, immunohistochemical staining, immunoprecipitation assay, complement fixation assay, fluorescence activated cell sorter (FACS), and protein chip.

98

45. The method of any one of the preceding claims, wherein the subject has a cancer.

46. The method of claim 45, wherein the cancer is a carcinoma, a sarcoma, a lymphoma, a melanoma, a pediatric tumor, or a leukemia.

47. The method of claim 45, wherein the cancer is selected from adrenal gland tumors, biliary cancer, bladder cancer, brain cancer, breast cancer, carcinoma, central or peripheral nervous system tissue cancer, cervical cancer, colon cancer, endocrine or neuroendocrine cancer or hematopoietic cancer, esophageal cancer, fibroma, gastrointestinal cancer, glioma, head and neck cancer, Li-Fraumeni tumors, liver cancer, lung cancer, lymphoma, melanoma, meningioma, multiple neuroendocrine type I and type II tumors, nasopharyngeal cancer, oral cancer, oropharyngeal cancer, osteogenic sarcoma tumors, ovarian cancer, pancreatic cancer, pancreatic islet cell cancer, parathyroid cancer, pheochromocytoma, pituitary tumors, prostate cancer, rectal cancer, renal cancer, respiratory cancer, sarcoma, skin cancer, stomach cancer, testicular cancer, thyroid cancer, tracheal cancer, urogenital cancer, and uterine cancer.

48. A method of treating cancer in a patient in need thereof with a cancer therapy, comprising: selecting a patient who is likely responsive to treatment of the cancer therapy according to the method of any one of the preceding claims; and administering to the patient the cancer therapy.

49. The method of claim 48, wherein the cancer therapy comprises an adoptive cell therapy (ACT).

50. The method of claim 49, wherein the adoptive cell therapy comprises a T-cell receptor (TCR) T cell therapy or a chimeric antigen receptor (CAR) T cell therapy.

51. The method of claim 49, wherein the adoptive cell therapy comprises an adoptive cell therapy with TILs.

99

52. The method of any one of claims 48 to 51, wherein the cancer is a carcinoma, a sarcoma, a lymphoma, a melanoma, a pediatric tumor, or a leukemia.

100