Neoantigen selection for personalized cancer vaccines
Patent Information
- Application Number
- JP2023527473
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2020-11-06
- Filing Date
- 2021-11-05
- Publication Date
- 2024-11-12
AI Technical Summary
Current cancer therapies, including ablative techniques and chemical agents, are associated with serious risks, toxic side effects, and high costs, while traditional cancer vaccines targeting tumor-associated antigens face challenges due to immune tolerance and uncertain efficacy.
A method for selecting tumor-specific neoantigens using machine learning platforms to predict immunogenicity, RNA expression, and genetic diversity, followed by formulation of personalized immunogenic compositions to elicit a targeted immune response.
The approach enhances the accuracy of neoantigen selection, reducing immune tolerance and improving the safety and efficacy of cancer vaccines by focusing on tumor-specific neoantigens that are immunogenic and sufficiently expressed.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
[Technical Field]
[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 110,711, filed November 6, 2020, the entire contents of which are incorporated herein by reference.
[0002] Sequence Listing Reference This application contains a Sequence Listing in computer readable form, which is incorporated herein by reference. The above ASCII copy, created on October 28, 2021, is named 146401_091524_SL.txt and is 75,598 bytes in size. [Background technology]
[0003] Cancer is the leading cause of death worldwide, accounting for one-quarter of all deaths (Siegel et al., CA: A Cancer Journal for Clinicians, 68:7-30(2018)). In 2018, there were 18.1 million new cancer cases and 9.6 million cancer-related deaths (Bray et al., CA: A Cancer Journal for Clinicians, 68(6):394-424). Several existing standard-of-care cancer therapies exist, including ablative techniques (e.g., surgery and radiation) and chemical techniques (e.g., chemotherapy). Unfortunately, such therapies are often associated with serious risks, toxic side effects, and significant costs, as well as uncertain efficacy.
[0004] Cancer immunotherapy (e.g., cancer vaccines) has emerged as a promising cancer treatment. The goal of cancer immunotherapy is to harness the immune system for selective destruction of cancer while leaving normal tissues unharmed. Traditional cancer vaccines typically target tumor-associated antigens, which are typically present in normal tissues but are overexpressed in cancer. However, because these antigens are often present in normal tissues, immune tolerance may prevent immune activation. Several clinical trials targeting tumor-associated antigens have not shown sustained beneficial effects compared to standard treatments. Li et al., Ann Oncol., 28(Suppl 12):xii11-xii17(2017).
[0005] Neoantigens represent attractive targets for cancer immunotherapy. Neoantigens are non-self proteins with individual specificity. Neoantigens originate from random somatic mutations in tumor cell genomes and are not expressed on the surface of normal cells. Id. Because neoantigens are expressed only on tumor cells and therefore do not induce central immune tolerance, cancer vaccines targeting cancer neoantigens have potential advantages, including reduced central immune tolerance and an improved safety profile. Id.
[0006] The mutational landscape of cancer is complex, and tumor mutations are generally unique to each individual subject. Most somatic mutations detected by sequencing do not result in effective neoantigens. Only a small percentage of mutations in tumor DNA or tumor cells are transcribed, translated, and processed into tumor-specific neoantigens with sufficient precision to design a vaccine that is likely to be effective. Furthermore, not all neoantigens are immunogenic. In fact, the percentage of T cells that spontaneously recognize endogenous neoantigens is approximately 1%–2%. See Karpanen et al., Front Immunol., 8:1718 (2017). Furthermore, the cost and time associated with producing neoantigen vaccines are significant.
[0007] Therefore, efficiently and accurately predicting, prioritizing, and selecting neoantigen candidates for immunogenic compositions remains a challenge. Thus, there is a significant unmet need for integrated methods to characterize tumor genomic material in order to identify neoantigens, identify which neoantigens are targeted by the immune system, and select which neoantigens are likely suitable for effective immunogenic compositions.
[0008] Summary of the Invention The present disclosure relates to a method for selecting one or more tumor-specific neoantigens from a subject's tumor for a subject-specific immunogenic composition. The disclosure also relates to a method for treating cancer in a subject in need thereof by administering an immunogenic composition comprising tumor-specific neoantigens selected using a novel approach for selecting tumor-specific neoantigens and formulating the immunogenic composition comprising the selected tumor-specific neoantigens. The approach involves obtaining sequence data from the tumor. The sequence data is used to obtain data representing the polypeptide sequences of one or more tumor-specific neoantigens. The sequence data can be nucleotide sequence data, polypeptide sequence data, exome sequence data, transcriptome sequence data, or whole genome nucleotide sequence data. The sequence data can be whole exome sequence data, RNA sequence data, whole genome sequence data, or a combination thereof. The sequence data can be a combination of whole exome sequence data, RNA sequence data, and whole genome sequence data.
[0009] The subject's polypeptide sequence(s) and MHC molecule(s) are then input into a machine learning platform. The machine learning platform is used to identify whether the tumor-specific neoantigens are immunogenic (e.g., whether the one or more tumor-specific neoantigens induce an immune response in the subject). Based on these predictions, the machine learning platform generates a probability numerical score that the one or more tumor-specific neoantigens will induce an immune response in the subject.
[0010] The MCH molecule(s) of interest can be an MHC class I molecule and / or an MHC class II molecule. The polypeptide sequence encoding one or more tumor-specific neoantigens can be derived from a short polypeptide. Short polypeptides are typically presented on MHC class I molecules. Alternatively, the polypeptide sequence encoding one or more tumor-specific neoantigens can be derived from a long polypeptide.
[0011] The immune response in the subject can include presentation of one or more tumor-specific neoantigens on the surface of tumor cells, presentation of one or more tumor-specific neoantigens by one or more MHC molecules on tumor cells, or presentation of one or more tumor-specific neoantigens to T cells by antigen-presenting cells.
[0012] The immune response in the subject can be a CD4+ or a CD8+ mediated response. Typically, the immune response is either a CD4+ or a CD8+ mediated response.
[0013] A tumor-specific neoantigen with a higher probability numerical score compared to a lower probability numerical score indicates that the tumor-specific neoantigen elicits a greater immune response in the subject.
[0014] RNA expression, preferably mRNA expression, of one or more tumor-specific neoantigens in the tumor is also quantified to further identify one or more tumor-specific neoantigens that are sufficiently expressed to elicit an immune response in the subject. The tumor clones can then be optionally characterized to ensure that the tumor-specific neoantigens are sufficiently representative across the tumor (e.g., have genetic diversity). In embodiments, suitable tumor-specific neoantigens may represent approximately 1% of the tumor. In other cases, suitable tumor-specific neoantigens may represent approximately 5% of the tumor.
[0015] These parameters are used to calculate a tumor-specific neoantigen score for one or more tumor-specific neoantigen scores. The tumor-specific neoantigen score is used to select tumor-specific neoantigens suitable for formulating a subject-specific immunogenic composition. A higher tumor-specific neoantigen score compared to a lower tumor-specific neoantigen score indicates that the neoantigen has stronger immunogenicity and is therefore more likely to induce a strong immune response and induce a stable therapeutic effect (i.e., more likely to be suitable for an immunogenic composition). In embodiments, at least about 10 tumor-specific neoantigens are selected for formulating a subject-specific immunogenic composition. In embodiments, at least about 20 tumor-specific neoantigens are selected for formulating a subject-specific immunogenic composition.
[0016] The methods disclosed herein can further include measuring the ability of one or more tumor-specific neoantigens to induce an autoimmune response against normal tissue. Tumor-specific neoantigens that induce an autoimmune response in normal tissue have a lower tumor-specific neoantigen score compared to tumor-specific neoantigens that do not induce an autoimmune response. Tumor-specific neoantigens that induce an autoimmune response are not selected for the immunogenic composition.
[0017] The formulated immunogenic composition may contain at least about 10 tumor-specific neoantigens or at least about 20 tumor-specific neoantigens. The tumor-specific neoantigens may be encoded by short or long polypeptides. The immunogenic composition may comprise a nucleotide sequence, a polypeptide sequence, RNA, DNA, a cell, a plasmid, a vector, a dendritic cell, or a synthetic long peptide. The immunogenic composition may further comprise an adjuvant.
[0018] The present disclosure also relates to a method of treating cancer in a subject in need thereof, comprising administering a personalized immunogenic composition comprising one or more tumor-specific neoantigens selected using the methods described herein. The methods disclosed herein may be suitable for treating any number of cancers. The tumor may be derived from melanoma, breast cancer, ovarian cancer, prostate cancer, kidney cancer, stomach cancer, colon cancer, testicular cancer, head and neck cancer, pancreatic cancer, brain cancer, B-cell lymphoma, acute myeloid leukemia, chronic myeloid leukemia, chronic lymphocytic leukemia, T-cell lymphocytic leukemia, bladder cancer, or lung cancer. Preferably, the cancer is melanoma, breast cancer, lung cancer, or bladder cancer. [Brief explanation of the drawings]
[0019] [Figure 1] FIG. 1 is a schematic diagram showing an approach for selecting one or more tumor-specific neoantigens. [Figure 2] FIG. 1 is a schematic flow diagram showing bioinformatics analysis of next-generation sequencing data (input / output). [Figure 3] FIG. 1 is a modular flow diagram for clonal deconvolution. DETAILED DESCRIPTION OF THE INVENTION
[0020] The present disclosure relates to a novel approach for selecting tumor-specific neoantigens with high accuracy for potent personalized cancer immunogenic compositions (e.g., subject-specific immunogenic compositions). The present disclosure also relates to methods for treating cancer in a subject in need thereof by administering an immunogenic composition containing tumor-specific neoantigens selected using the novel approach for selecting tumor-specific neoantigens and formulating an immunogenic composition containing the selected tumor-specific neoantigens. The inventors have developed the following approach: 1) sequencing DNA and / or RNA encoding the polypeptide sequence of one or more neoantigens; 2) determining whether the tumor-specific neoantigens are immunogenic (e.g., whether the neoantigens are capable of eliciting an immune response in the subject); 3) determining whether the tumor expresses the neoantigens in sufficient amounts to elicit an immune response; and 4) optionally, determining whether the neoantigens represent a sufficient proportion of the tumor. Currently available methods rely on MHC binding affinity prediction to rank and select neoantigens or neoantigens for the probability that they will be presented by MHC molecules. These methods do not predict immunogenicity. Furthermore, current methods are not capable of assessing all of these factors with high accuracy.
[0021] This approach begins with sequencing the polypeptide sequences of tumor-specific neoantigens obtained from tumor biopsies. A predictive machine learning platform is then used to identify which neoantigens are recognized by the subject's MHC molecules. The platform can determine whether the tumor-specific neoantigens are immunogenic (e.g., that they will elicit an immune response in the subject). Based on these predictions, the machine learning platform generates a probability numerical score that the tumor-specific neoantigen will elicit an immune response. The RNA expression, preferably mRNA expression, of the tumor-specific neoantigens is also quantified to focus on tumor-specific neoantigens that are abundantly expressed and likely to elicit an immune response. Tumor clones are then optionally characterized to ensure that the tumor-specific neoantigens represent sufficient genetic diversity across the tumor. These parameters are used to generate a tumor-specific neoantigen score for the tumor-specific neoantigen. The tumor-specific neoantigen score is used to select tumor-specific neoantigens suitable for formulating personalized vaccines. Compared to a lower tumor-specific neoantigen score, a higher tumor-specific neoantigen score indicates that the neoantigen has stronger immunogenicity and is therefore more likely to induce a stronger immune response and induce a more stable therapeutic effect.
[0022] All publications and patent documents cited in this disclosure are incorporated herein by reference in their entirety. To the extent that material incorporated by reference contradicts or is inconsistent with this specification, this specification supersedes such material. The citation of any reference herein is not an admission that such reference is prior art to this disclosure. When a range of values is expressed, it includes embodiments using any specific value within that range. Furthermore, reference to values stated in a range includes each and every value within that range. All ranges are inclusive of endpoints and are combinable. By use of the antecedent "about," when values are expressed as approximations, it will be understood that the particular value forms another embodiment. Reference to a particular numerical value includes at least that particular value unless the context clearly dictates otherwise. The use of "or" means "and / or" unless the specific context of its use dictates otherwise.
[0023] Throughout the specification and claims, various terms relating to aspects of the present disclosure are used. Unless otherwise indicated, such terms shall be given their ordinary meaning in the art. Other terms that are specifically defined shall be interpreted consistent with the definitions set forth herein. The techniques and procedures described or referenced herein are generally well understood by those of skill in the art and commonly used conventional methods, such as the widely applied molecular cloning methodology described in Sambrook et al., Molecular Cloning: A Laboratory Manual 4th ed. (2012) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Where appropriate, procedures involving the use of commercially available kits and reagents are generally performed according to manufacturer-defined protocols and conditions unless otherwise specified.
[0024] As used herein, the singular forms "a," "an," and "the" include the plural forms unless the context clearly dictates otherwise. Terms such as "include," "such as," and the like are intended to encompass without limitation, unless expressly stated otherwise.
[0025] Unless otherwise indicated, the terms "at least," "less than," and "about," or similar terms preceding a series of elements or ranges, should be understood to refer to every element in the series or range. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims.
[0026] The term "cancer" refers to a physiological condition in a subject in which a cell population is characterized by uncontrolled proliferation, immortality, metastatic potential, rapid growth and proliferation rate, and / or specific morphological features. Often, cancer may be in the form of a tumor or mass, but may exist alone within a subject or circulate in the bloodstream as independent cells, such as leukemia or lymphoma cells. The term cancer includes all types of cancer and metastases, including hematological malignancies, solid tumors, sarcomas, carcinomas, and other solid and non-solid tumors. Examples of cancer include, but are not limited to, carcinoma, lymphoma, blastoma, sarcoma, and leukemia. More specific examples of such cancers include squamous cell carcinoma, small cell lung cancer, non-small cell lung cancer, lung adenocarcinoma, lung squamous cell carcinoma, peritoneal cancer, hepatocellular carcinoma, gastrointestinal cancer, pancreatic cancer, glioblastoma, cervical cancer, ovarian cancer, liver cancer, bladder cancer, breast cancer (e.g., triple-negative breast cancer, hormone receptor-positive breast cancer), osteosarcoma, melanoma, colon cancer, colorectal cancer, endometrial (e.g., serous) or uterine cancer, salivary gland cancer, kidney cancer, liver cancer, prostate cancer, vulvar cancer, thyroid cancer, liver cancer, and various forms of head and neck cancer. Triple-negative breast cancer refers to breast cancer in which the expression of the estrogen receptor (ER), progesterone receptor (PR), and Her2 / neu genes is negative. Hormone receptor-positive breast cancer refers to breast cancer in which at least one of ER or PR is positive and Her2 / neu (HER2) is negative.
[0027] The term "neoantigen" as used herein refers to an antigen that has at least one alteration different from its corresponding parent antigen, for example, through a mutation in tumor cells or a tumor cell-specific post-translational modification. Mutations can include frameshifts, indels, missense or nonsense substitutions, splice site alterations, genomic rearrangements or gene fusions, or any genomic expression alteration that gives rise to a neoantigen. Mutations can include splice mutations. Tumor cell-specific post-translational modifications can include aberrant phosphorylation. Tumor cell-specific post-translational modifications can also include proteasome-generated splicing antigens. See Lipe et al., Science, 354(6310):354:358 (2016). Generally, point mutations account for approximately 95% of mutations in tumors and indels, with frameshift mutations accounting for the remaining mutations. See Snyder et al., N Engl J Med., 371:2189-2199 (2014).
[0028] As used herein, the term "tumor-specific neoantigen" is a neoantigen that is present in tumor cells or tissues of a subject but is not present in normal cells or tissues of the subject.
[0029] The term "next generation sequencing" or "NGS," as used herein, refers to a sequencing technology with increased throughput compared to traditional approaches (e.g., Sanger sequencing), with the ability to generate hundreds of thousands of sequence reads at a time.
[0030] As used herein, the term "neural network" refers to a machine learning model for classification or regression that consists of multiple layers of linear transformations followed by element-wise nonlinearities that are typically trained via stochastic gradient descent and back-propagation.
[0031] The term "subject," as used herein, refers to any animal, such as any mammal, including, but not limited to, humans, non-human primates, rodents, etc. In some embodiments, the mammal is a mouse. In some embodiments, the mammal is a human.
[0032] As used herein, the term "tumor cell" refers to a cell that is or is derived from a cancer cell. The term "tumor cell" can also refer to a cell that exhibits cancer-like properties, such as uncontrolled reproduction, resistance to anti-proliferative signals, the ability to metastasize, and the loss of the ability to undergo programmed cell death.
[0033] A description of the method and guidance for carrying out the method is provided below.
[0034] I. Methods for selecting tumor-specific neoantigens Disclosed herein are methods for selecting suitable tumor-specific neoantigens from a subject's tumor for a subject-specific immunogenic composition. Suitable tumor-specific neoantigens are those that are likely to be presented on the cell surface of the tumor, likely to be immunogenic, predicted to be expressed in sufficient amounts to elicit an immune response in the subject, and optionally, display sufficient diversity throughout the tumor.
[0035] The first step in selecting one or more tumor-specific neoantigens from a tumor of interest involves obtaining sequence data from the tumor. The sequence data is used to obtain data representing the polypeptide sequences of the one or more tumor-specific neoantigens. Generally, the sequence data representing the polypeptide sequences of the one or more tumor-specific neoantigens is determined by subjecting a tumor sample to sequence analysis.
[0036] The sequence data can be exome sequence data, transcriptome sequence data, whole genome nucleotide sequence data, nucleotide sequence data, or polypeptide sequence data. The sequence data can be whole exome sequence data, RNA sequence data, whole genome sequence data, or a combination thereof. The sequence data can be a combination of whole exome sequence data, RNA sequence data, and whole genome sequence data.
[0037] Various methods for obtaining sequence data can be used in the methods described herein. Sequencing methods are well known in the art and include, but are not limited to, PCR-based methods, including real-time PCR, whole-exome sequencing, deep sequencing, high-throughput sequencing, or a combination thereof. In some embodiments, the techniques and procedures described above are carried out according to the methods described in, for example, Sambrook et al., Molecular Cloning: A Laboratory Manual 4th ed. (2012) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. See also Austell et al., Current Protocols in Molecular Biology, ed., Greene Publishing and Wiley-Interscience New York (1992) (regularly updated).
[0038] Sequencing methods can also include, but are not limited to, high-throughput sequencing, single-cell RNA sequencing, RNA sequencing, pyrosequencing, sequencing by synthesis, single molecule sequencing, nanopore sequencing, semiconductor sequencing, sequencing by synthesis, sequencing by ligation, sequencing by hybridization, RNA-Sew (Illumina), digital gene expression (Helicos), next-generation sequencing, single molecule sequencing by synthesis (SMSS) (Helicos), massively parallel sequencing, clonal single molecule array (Solexa), shotgun sequencing, Max Highbury or Sanger sequencing, whole genome sequencing, whole exome sequencing, primer walking, sequencing using PacBio, SOLid, Ion Torrent or Nanopore, or any known sequencing method.The sequencing method used herein to obtain sequence data is preferably high-throughput sequencing.High-throughput sequencing technology can sequence multiple nucleic acid molecules in parallel, allowing millions of nucleic acid molecules to be sequenced at once. See Churko et al., Circ. Res. 112(12):1613-1623 (2013).
[0039] In some cases, whole exome sequencing, RNA sequencing, whole genome sequencing, or a combination thereof can be performed. In some cases, a combination of whole exome sequencing, RNA sequencing, and whole genome sequencing can be performed.
[0040] In some cases, high-throughput sequencing can be next-generation sequencing. Several different next-generation platforms exist that use different sequencing technologies (e.g., using the HiSeq or MiSeq instruments available from Illumina (San Diego, CA)). Any of these platforms can be utilized to sequence the genetic material disclosed herein. Next-generation sequencing is based on the sequencing of multiple independent reads, with each read representing anywhere between 10 and 1,000 bases of nucleic acid. Sequencing by synthesis is a common technique used in next-generation sequencing. Generally, sequencing involves hybridizing a primer to a template to form a template / primer duplex and contacting the duplex with a polymerase in the presence of detectably labeled nucleotides under conditions that allow the polymerase to add nucleotides to the primer in a template-dependent manner. The steps are then repeated sequentially to identify the incorporated bases using the signal from the detectable label and determine the linear order of nucleotides within the template. Exemplary detectable labels include radioactive labels, fluorescent labels, enzymatic labels, and the like. A number of techniques are known for detecting sequences, such as the Illumina NextSeq platform by cycle end sequencing.
[0041] Once sequence data representing the polypeptide sequences of one or more tumor-specific neoantigens are obtained, the sequence data, along with the subject's MHC molecules, are input into a machine learning platform, which generates a probability numerical score that predicts whether the one or more tumor-specific neoantigens will be immunogenic (e.g., elicit an immune response in the subject).
[0042] MHC molecules transport and present peptides on the cell surface. MHC molecules are classified as class I and class II MHC molecules. MHC class I is present on the surface of almost all cells in the body, including most tumor cells. MHC class I proteins typically contain antigens derived from endogenous proteins or intracellular pathogens and are presented to cytotoxic T lymphocytes (i.e., CD8+). MHC class I molecules can include HLA-A, HLA-B, or HLA-C. Class II MHC molecules are present only on dendritic cells, B lymphocytes, macrophages, and other antigen-presenting cells. They primarily present peptides from external antigen sources, i.e., from the outside of cells, to T helper (Th) cells (i.e., CD4+). MHC class II molecules can include HLA-DPA1, HLA-DPB1, HLA-DQA1, HLA-DQB1, HLA-DRA, and HLA-DRB1. In some cases, MHC class II molecules can also be expressed on cancer cells.
[0043] MHC class I molecules and / or MHC class II molecules can be input into the machine learning platform. Typically, either MHC class I molecules or MHC class II molecules are input into the machine learning platform. In some embodiments, MHC class I molecules are input into the machine learning platform. In other embodiments, MHC class II molecules are input into the machine learning platform.
[0044] MHC class I molecules bind to short peptides. MHC class I molecules can generally accommodate peptides about 8 to about 10 amino acids in length. In embodiments, the sequence data encoding one or more tumor-specific neoantigens are short peptides about 8 to about 10 amino acids in length. MHC class II molecules bind to long peptides. MHC class II molecules can generally accommodate peptides about 13 to about 25 amino acids in length. In embodiments, the sequence data encoding one or more tumor-specific neoantigens are long peptides about 13 to 25 amino acids in length.
[0045] The sequence data encoding one or more tumor-specific neoantigens can be about 5 amino acids in length, about 6 amino acids in length, about 7 amino acids in length, about 8 amino acids in length, about 9 amino acids in length, about 10 amino acids in length, about 11 amino acids in length, about 12 amino acids in length, about 13 amino acids in length, about 14 amino acids in length, about 15 amino acids in length, about 16 amino acids in length, about 17 amino acids in length, about 18 amino acids in length, about 19 amino acids in length, about 20 amino acids in length, about 21 amino acids in length, about 22 amino acids in length, about 23 amino acids in length, about 24 amino acids in length, about 25 amino acids in length, about 26 amino acids in length, about 27 amino acids in length, about 28 amino acids in length, about 29 amino acids in length, or about 30 amino acids in length.
[0046] The machine learning platform predicts the likelihood that one or more tumor-specific neoantigens are immunogenic (e.g., induce an immune response).
[0047] Immunogenic tumor-specific neoantigens are not expressed in normal tissues. They can be presented to CD4+ and CD8+ T cells by antigen-presenting cells to generate an immune response. In embodiments, the immune response elicited in a subject by one or more tumor-specific neoantigens comprises presentation of one or more tumor-specific neoantigens on the surface of tumor cells. More specifically, the immune response elicited in a subject by one or more tumor-specific neoantigens comprises presentation of one or more tumor-specific neoantigens by one or more MHC molecules on tumor cells. The immune response elicited by one or more tumor-specific neoantigens is expected to be a T cell-mediated response. The immune response elicited in a subject by one or more tumor-specific neoantigens may comprise one or more tumor-specific neoantigens that can be presented to T cells by antigen-presenting cells, such as dendritic cells. Preferably, the one or more tumor-specific neoantigens can activate CD8+ T cells and / or CD4+ T cells.
[0048] In embodiments, the machine learning platform can predict the likelihood that one or more tumor-specific neoantigens will activate CD8+ T cells. In embodiments, the machine learning platform can predict the likelihood that one or more tumor-specific neoantigens will activate CD4+ T cells. In some cases, the machine learning platform can predict the antibody titers that one or more tumor-specific neoantigens may induce. In other cases, the machine learning platform can predict the frequency of CD8+ activation by one or more tumor-specific neoantigens.
[0049] The machine learning platform can include a model trained with training data. The training data can be obtained from a set of different subjects. The training data can include data obtained from healthy subjects and subjects with cancer. The training data can include a variety of data that can be used to generate a probability score indicating whether one or more tumor-specific neoantigens will induce an immune response in a subject. Exemplary training data can include data representing nucleotide or polypeptide sequences from normal tissues and / or cells, data representing nucleotide or polypeptide sequences from tumor tissue, data representing MHC peptidome sequences from normal and tumor tissues, peptide-MHC binding affinity measurements, or combinations thereof. The reference data can further include mass spectrometry data, DNA sequencing data, RNA sequencing data, clinical data from healthy subjects and subjects with cancer, cytokine profiling data, T cell cytotoxicity assay data, peptide-MHC mono- or multimer data, and proteomic data from synthetic proteins, normal and tumor human cell lines, fresh and frozen primary samples, and monoallelic cell lines engineered to express a given MHC allele that are subsequently exposed to a T cell assay.
[0050] The machine learning platform can be a supervised learning platform, an unsupervised learning platform, or a semi-supervised learning platform. The machine learning platform can use a sequence-based approach to generate a probability value that one or more tumor-specific neoantigens can induce an immune response (e.g., induce a high or low antibody response or a CD8+ response). Sequence-based predictions can include supervised machine learning modules, including artificial neural networks (e.g., deep or otherwise), support vector machines, K-nearest neighbors, logistic multiple network constrained regression (LogMiNeR), regression trees, random forests, Adaboost, XGBoost, or hidden Markov models. These platforms require a training dataset containing known MHC-binding peptides.
[0051] A number of predictive programs are used to predict whether tumor-specific neoantigens are presented on MHC molecules and are capable of eliciting an immune response. Exemplary prediction programs include, for example, HLAminer (Warren et al., Genome Med., 4:95 (2012), which predicts HLA types by directing the assembly of shotgun sequence data and comparing it to a reference allele sequence database), VariantEffect Predictor Tool (McLaren et al., Genome Biol., 17:122 (2016)), NetMHCpan (Andreatta et al., Bioinformatics., 32:511-517 (2016): an artificial neural network-based sequence comparison method and prediction of peptide-MHC-I type molecule affinity), UCSC Browser (Kent et al., Genome Res., 12:996-1006 (2002)), CloudNeo pipeline (Bais et al., Bioinformatics, 33:3110-2 (2017)), OptiType (Szolek et al. al.,Bioinformatics,30:3310-316(2014)), ATHLATES(Liu C et al.,Nucleic Acids Res.41:e142(2013)), pVAC-Seq(Hundal et al.,Genome Med.8:11(2016)), MuPeXI(Bjerregaard et al.,Cancer Immunol Immunother.,66:1123-30(2017)), Strelka(Saunders et al.,Bioinformatics.28:1811-7(2012)), Strelka2(Kim et al.,Nat Methods.2018;15:591-4.), VarScan2(Koboldt et al.,Genome Res.,22:568-76(2012)), Somaticseq(Fang L et al., Genome Biol., 16:197 (2015)), SMMPMBEC (Kim et al., BMC Bioinformatics.,10:394(2009)), NeoPredPipe(Schenck RO,BMC Bioinformatics.,20:264(2019)), Weka(Witten et al.,Data mining:practical machine-learning tools and techniques.4. th ed. Elsevier, ISBN: 97801280435578 (eBook) (2017), or Orange (Demsar et al., Orange: Data Mining Toolbox in Python., J. Mach Learn Res., 14:2349-2353 (2013)). Any known predictive program can be used as a machine learning platform to generate a probability numerical score indicating whether a neoantigen will induce an immune response.
[0052] Depending on the machine learning platform used, additional filters can be applied to prioritize tumor-specific neoantigen candidates, including removing hypothetical (RIKEN) proteins, using antigen processing algorithms to remove epitopes that are unlikely to be produced constitutively or proteolytically by the immunoproteasome, and prioritizing neoantigens with higher predicted binding affinity than the corresponding wild-type sequence.
[0053] The probability numeric score may be a number between 0 and 1. In embodiments, the probability numeric score may be a number of 0, 0.0001, 0.0002, 0.0003, 0.0004, 0.0005, 0.0006, 0.0007, 0.0008, 0.0009, 0.001, 0.002, 0.003, 0.004, 0.005, 0.006, 0.007, 0.008, 0.009, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90, or 1. A tumor-specific neoantigen with a higher probability numerical score compared to a lower probability numerical score indicates that the tumor-specific neoantigen will induce a greater immune response in a subject and is therefore more likely to be a suitable candidate for an immunogenic composition. For example, a tumor-specific neoantigen with a probability numerical score of 1 is more likely to induce a greater immune response in a subject than a tumor-specific neoantigen with a probability numerical score of 0.05. Similarly, a tumor-specific neoantigen with a probability numerical score of 0.5 is more likely to induce a greater immune response in a subject than a tumor-specific neoantigen with a probability numerical score of 0.1.
[0054] Higher probability numerical scores are preferred over lower probability numerical scores. Preferably, tumor-specific neoantigens have a probability numerical score of at least 0.8, 0.81, 0.82. A probability numerical score of 0.83, 0.84, 0.85, 0.86, 0.87, 0.88, 0.89, 0.9, 0.95, 0.96, 0.97, 0.98, 0.99, or 1 indicates a high likelihood of eliciting an immune response in a subject.
[0055] Although a higher probability numerical score is preferred, a lower probability numerical score may still indicate that the tumor-specific neoantigen is capable of eliciting a sufficient immune response, such that the tumor-specific neoantigen is likely to be a suitable candidate.
[0056] In some cases, the machine learning platform described herein can also predict the likelihood that one or more tumor-specific neoantigens are presented by MHC molecules on tumor cells. The machine learning platform can predict the likelihood that one or more tumor-specific neoantigens are presented by MHC class I molecules or MHC class II molecules.
[0057] The method for selecting one or more tumor-specific neoantigens may further include measuring in silico the affinity of one or more tumor-specific neoantigens for binding to MHC molecules of a subject. Tumor-specific neoantigens with a binding affinity to an MHC molecule of less than about 1000 nM indicate that the one or more tumor-specific neoantigens may be suitable for an immunogenic composition. Tumor-specific neoantigens with a binding affinity to an MHC molecule of less than about 500 nM, less than about 400 nM, less than about 300 nM, less than about 200 nM, less than about 100 nM, or less than about 50 nM may indicate that the one or more tumor-specific neoantigens may be suitable for an immunogenic composition. The affinity of one or more tumor-specific neoantigens for binding to an MHC molecule in a subject can predict tumor-specific neoantigen immunogenicity. Alternatively, median affinity can be an effective method for predicting tumor-specific neoantigen immunogenicity. Median affinity can be calculated using epitope prediction algorithms such as NetMHCpan, ANN, SMM, and SMMPMBEC.
[0058] The RNA expression of one or more tumor-specific neoantigens is also quantified. The RNA expression of one or more tumor-specific neoantigens is quantified to identify one or more neoantigens that induce an immune response in the subject. Various methods exist for measuring RNA expression. Known techniques that can measure RNA expression include RNA-seq, in situ hybridization (e.g., FISH), Northern blot, DNA microarray, tiling array, and quantitative polymerase chain reaction (qPCR). Other techniques known in the art can be used to quantify RNA expression. The RNA can be messenger RNA (mRNA), short interfering RNA (siRNA), microRNA (miRNA), circular RNA (circRNA), transfer RNA (tRNA), ribosomal RNA (rRNA), small nuclear RNA (snRNA), Piwi-interacting RNA (piRNA), long non-coding RNA (long ncRNA), subgenomic RNA (sgRNA), RNA from integrating or non-integrating viruses, or any other RNA. Preferably, mRNA expression is measured.
[0059] The methods disclosed herein can optionally include sequencing the tumor clone. The tumor clone is sequenced to identify one or more tumor-specific neoantigens that represent a sufficient portion of the tumor. The tumor clone can be sequenced, for example, using the sequencing techniques disclosed herein, as well as other known sequencing techniques known to those skilled in the art.
[0060] In embodiments, tumor-specific neoantigens having a tumor clonal fraction of at least about 1%, at least about 2%, at least about 3%, at least about 4%, at least about 5%, at least about 6%, at least about 7%, at least about 8%, at least about 9%, at least about 10%, at least about 15%, at least about 20%, at least about 25%, or at least about 30% across the tumor indicate that the tumor-specific neoantigens represent a sufficient fraction of the tumor, which indicates that the tumor-specific neoantigens provide sufficient genetic diversity across the tumor.
[0061] The method can further include measuring the ability of one or more tumor-specific neoantigens to induce an autoimmune response in normal tissue. Tumor-specific neoantigens with sequences similar to normal antigens are expected to induce an autoimmune response in normal tissue. For example, tumor-specific neoantigens that are at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% similar to normal antigens can induce an autoimmune response. Tumor-specific neoantigens predicted to induce an autoimmune response are not prioritized for immunogenic compositions. Tumor-specific neoantigens predicted to induce an autoimmune response are typically not selected for immunogenic compositions. The method can further include measuring the ability of one or more tumor-specific neoantigens to induce immune tolerance. Tumor-specific neoantigens predicted to induce immune tolerance are not prioritized for immunogenic compositions. Tumor-specific neoantigens predicted to induce immune tolerance are not prioritized for immunogenic compositions.
[0062] The tumor-specific score is calculated based on data generated by obtaining a probability numerical score that one or more tumor-specific neoantigens will induce an immune response in a subject and the RNA expression level of the one or more tumor-specific neoantigens. The tumor clonal proportion across the tumor can optionally be included in addition to the above calculations used to calculate the tumor-specific score. Priority is given to tumor-specific neoantigens that have a high probability numerical score (e.g., the tumor-specific neoantigen is immunogenic) and a high level of RNA expression. In comparison, tumor-specific antigens predicted to induce an autoimmune response have a low tumor-specific neoantigen score compared to tumor-specific neoantigens that do not induce an immune response and are not selected for inclusion in the immunogenic composition. Priority is given to tumor-specific neoantigens that have a high probability numerical score (e.g., the tumor-specific neoantigen is immunogenic), a high level of RNA expression, and provide a sufficient tumor clonal proportion across the tumor.
[0063] A tumor-specific neoantigen with a high probability numerical score (e.g., the tumor-specific neoantigen is immunogenic) and optionally providing a sufficient tumor clonal proportion throughout the tumor, but with a low RNA expression level, has a lower tumor-specific score compared to a tumor-specific neoantigen with a high probability numerical score, a high RNA expression level, and optionally providing a sufficient tumor clonal proportion throughout the tumor. In this example, a tumor-specific neoantigen with a lower tumor-specific score is not prioritized over a tumor-specific neoantigen with a higher tumor-specific score. A tumor-specific neoantigen with a high probability numerical score (e.g., the tumor-specific neoantigen is immunogenic) and a sufficient level of RNA expression to elicit an immune response, but not providing a sufficient tumor clonal proportion throughout the tumor, has a lower tumor-specific score compared to a tumor-specific neoantigen with a high probability numerical score, a high RNA expression level, and providing a sufficient tumor clonal proportion throughout the tumor. Tumor-specific neoantigens that have sufficient levels of RNA expression to elicit an immune response and provide sufficient tumor fraction throughout the tumor, but have a low probability numerical score, will have a low tumor-specific score compared to tumor-specific neoantigens that have a high probability numerical score, high RNA expression levels, and provide sufficient tumor fraction throughout the tumor.
[0064] Finally, one or more tumor-specific neoantigens based on the tumor-specific score are selected for formulation of a target-specific immunogenic composition. In embodiments, at least about 1, at least about 2, at least about 3, at least about 4, at least about 5, at least about 6, at least about 7, at least about 8, at least about 9, at least about 10, at least about 11, at least about 12, at least about 13, at least about 14, at least about 15, at least about 16, at least about 17, at least about 18, at least about 19, at least about 20, at least about 25, at least about 30, at least about 35, at least about 40, at least about 50, or more tumor-specific neoantigens are selected for the immunogenic composition. Typically, at least about 10 tumor-specific neoantigens are selected. In other cases, at least about 20 tumor-specific neoantigens are selected.
[0065] II. Treatment method The present disclosure also relates to methods of treating cancer in a subject in need thereof, comprising administering an individualized immunogenic composition comprising one or more tumor-specific neoantigens selected using the methods described herein.
[0066] The cancer may be any solid tumor or any hematological tumor. The methods disclosed herein are preferred for solid tumors. The tumor may be a primary tumor (e.g., a tumor at the original site where the tumor first developed). Solid tumors may include, but are not limited to, breast cancer, ovarian cancer, prostate cancer, lung cancer, kidney cancer, stomach cancer, testicular cancer, head and neck cancer, pancreatic cancer, brain cancer, and melanoma. Hematological tumors may include, but are not limited to, tumors derived from lymphoma (e.g., B-cell lymphoma) and leukemia (e.g., acute myeloid leukemia, chronic myeloid leukemia, chronic lymphocytic leukemia, and T-cell lymphocytic leukemia).
[0067] The methods disclosed herein can be used with any suitable cancerous tumor, including hematological malignancies, solid tumors, sarcomas, carcinomas, and other solid and non-solid tumors. Exemplary suitable cancers include, for example, acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), adrenocortical carcinoma, anal cancer, appendix cancer, astrocytoma, basal cell carcinoma, brain tumor, bile duct cancer, bladder cancer, bone cancer, breast cancer, bronchial tumor, cancer of unknown primary origin, cardiac tumor, cervical cancer, chordoma, colon cancer, colorectal cancer, craniopharyngioma, ductal carcinoma, embryonal tumor, endometrial cancer, ependymoma, esophageal cancer, neuroblastoma, and fibrous histiocytoma. , Ewing's sarcoma, eye cancer, germ cell tumor, gallbladder cancer, gastric cancer, gastrointestinal carcinoid tumor, gastrointestinal stromal tumor, gestational trophoblastic disease, glioma, head and neck cancer, hepatocellular carcinoma, histiocytosis, Hodgkin's lymphoma, hypopharyngeal cancer, intraocular melanoma, islet cell tumor, Kaposi's sarcoma, kidney cancer, Langerhans cell histiocytosis, laryngeal cancer, lip and oral cavity cancer, liver cancer, lobular carcinoma in situ, lung cancer, macroglobulinemia, malignant fibrous histiocytoma, melanoma, melanoma Lancet cell carcinoma, mesothelioma, metastatic squamous neck cancer of unknown primary, midline tract carcinoma with NUT gene, oral cancer, multiple endocrine gland adenoma, multiple myeloma, mycosis fungoides, myelodysplastic syndrome, myelodysplastic / myeloproliferative neoplasms, nasal cavity and pleuropulmonary mean cancer, nasopharyngeal carcinoma, neuroblastoma, non-small cell lung cancer, oropharyngeal carcinoma, osteosarcoma, ovarian cancer, pancreatic cancer, papillomatosis, paraganglioma, parathyroid carcinoma, penile cancer, pharyngeal cancer, Examples of cancers include pheochromocytoma, pituitary tumor, blastoma, primary central nervous system lymphoma, prostate cancer, rectal cancer, renal cell carcinoma, renal pelvis and ureter cancer, retinoblastoma, rhabdomyosarcoma, salivary gland cancer, Sézary syndrome, skin cancer, small cell lung cancer, small intestine cancer, soft tissue sarcoma, spinal cord tumor, gastric cancer, T-cell lymphoma, teratoid tumor, testicular cancer, laryngeal cancer, thymoma and thymic cancer, thyroid cancer, urethral cancer, uterine cancer, vaginal cancer, vulvar cancer, and Wilms' tumor. Preferably, the cancer is melanoma, breast cancer, ovarian cancer, prostate cancer, kidney cancer, stomach cancer, colon cancer, testicular cancer, head and neck cancer, pancreatic cancer, brain cancer, B-cell lymphoma, acute myeloid leukemia, chronic myeloid leukemia, chronic lymphocytic leukemia, T-cell lymphocytic leukemia, bladder cancer, or lung cancer. Melanoma is of particular interest. Breast, lung, and bladder cancers are also of particular interest.
[0068] The immunogenic composition stimulates the subject's immune system, particularly the response of specific CD8+ T cells or CD4+ T cells. Interferon gamma produced by CD8+ and T-helper CD4+ cells regulates the expression of PD-L1. PD-L1 expression on tumor cells is upregulated upon attack by T cells. Therefore, tumor vaccines may induce the production of specific T cells while simultaneously upregulating PD-L1 expression, which may limit the effectiveness of the immunogenic composition. Furthermore, during immune system activation, the expression of the T cell surface reporter CTLA-4 correspondingly increases, which binds to the ligands B7-1 / B7-2 on antigen-presenting cells and exerts an immunosuppressive effect. Therefore, in some cases, the subject may further be administered an anti-immunosuppressant or immunostimulatory agent, such as a checkpoint inhibitor. Checkpoint inhibitors may include, but are not limited to, anti-CTL4-A antibodies, anti-PD-1 antibodies, and anti-PD-L1 antibodies. These checkpoint inhibitors bind to immune checkpoint proteins on T cells, eliminating tumor cell-mediated inhibition of T cell function. Blockade of CTLA-4 or PD-L1 with antibodies can enhance a patient's immune response to cancer cells. CTLA-4 has been shown to be effective when used in vaccination protocols.
[0069] Immunogenic compositions comprising one or more tumor-specific neoantigens can be administered to subjects who have been diagnosed with cancer, who already have cancer, who have recurrent cancer (i.e., recurrence), or who are at risk for developing cancer. Immunogenic compositions comprising one or more tumor-specific neoantigens can be administered to subjects who are resistant to other forms of cancer treatment (e.g., chemotherapy, immunotherapy, or radiation). Immunogenic compositions comprising one or more tumor-specific neoantigens can be administered to subjects prior to other standard of care cancer therapies (e.g., chemotherapy, immunotherapy, or radiation). Immunogenic compositions comprising one or more tumor-specific neoantigens can be administered to subjects simultaneously with, after, or in combination with other standard of care cancer therapies (e.g., chemotherapy, immunotherapy, or radiation).
[0070] The subject can be a human, dog, cat, horse, or any animal in which a tumor-specific response is desired.
[0071] The immunogenic composition is administered to a subject in an amount sufficient to elicit an immune response to tumor-specific neoantigens and destroy or at least partially arrest symptoms and / or complications. In embodiments, the immunogenic composition can provide a long-lasting immune response. A long-lasting immune response can be established by administering booster doses of the immunogenic composition to the subject. The immune response to the immunogenic composition can be prolonged by administering booster doses to the subject. In embodiments, at least one, at least two, at least three or more booster doses can be administered to alleviate cancer. A first booster dose can increase the immune response by at least 50%, at least 100%, at least 200%, at least 300%, at least 400%, at least 500%, or at least 1000%. A second booster dose can increase the immune response by at least 50%, at least 100%, at least 200%, at least 300%, at least 400%, at least 500%, or at least 1000%. The third boosting dose may increase the immune response by at least 50%, at least 100%, at least 200%, at least 300%, at least 400%, at least 500%, or at least 1000%.
[0072] An amount sufficient to elicit an immune response is defined as a "therapeutically effective amount." Amounts effective for this use will depend, for example, on the particular composition of the vaccine regimen administered, the mode of administration, the stage and severity of the disease, the patient's general health, and the judgment of the prescribing physician. It should be noted that immunogenic compositions may generally be used in severe disease states, i.e., life-threatening or potentially life-threatening situations, particularly when cancer has metastasized. In such cases, given the minimization of foreign material and the relative non-toxicity of the neoantigens, administering substantial excesses of these immunogenic compositions is possible and may be felt to be desirable by the treating physician.
[0073] Immunogenic compositions containing one or more tumor-specific neoantigens can be administered to a subject alone or in combination with other therapeutic agents. The therapeutic agent can be, for example, a chemotherapy agent, radiation therapy, or immunotherapy. Any suitable therapeutic treatment for the particular cancer can be administered. Exemplary chemotherapeutic agents include, but are not limited to, aldesleukin, altretamine, amifostine, asparaginase, bleomycin, capecitabine, carboplatin, carmustine, cladribine, cisapride, cisplatin, cyclophosphamide, cytarabine, dacarbazine (DTIC), dactinomycin, docetaxel, doxorubicin, dronabinol, epoetin alfa, etoposide, filgrastim, fludarabine, fluorouracil, gemcitabine, granisetron, hydroxyurea, and idarucizumab. Examples of therapeutic agents include bicine, ifosfamide, interferon-alpha, irinotecan, lansoprazole, levamisole, leucovorin, megestrol, mesna, methotrexate, metoclopramide, mitomycin, mitotane, mitoxantrone, omeprazole, ondansetron, paclitaxel (Taxol®), pilocarpine, prochlorperazine, rituximab, tamoxifen, taxol, topotecan hydrochloride, trastuzumab, vinblastine, vincristine, and vinorelbine tartrate. The subject may also be administered a small molecule or targeted therapy (e.g., a kinase inhibitor). The subject may also be administered an anti-CTLA-4 antibody, an anti-PD-1 antibody, or an anti-PD-L1 antibody. Blocking CTLA-4 or PD-L1 with antibodies can enhance the patient's immune response to cancer cells.
[0074] III. Immunogenic composition The present invention further relates to personalized (i.e., subject-specific) immunogenic compositions (e.g., cancer vaccines) comprising one or more tumor-specific antigens selected using the methods described herein. Such immunogenic compositions can be formulated according to standard procedures in the art. The immunogenic compositions can raise specific immune responses.
[0075] The immunogenic composition can be formulated such that the selection and number of tumor-specific neoantigens are tailored to the subject's particular cancer. For example, the selection of tumor-specific neoantigens can depend on the particular type of cancer, the state of the cancer, the subject's immune status, and the subject's MHC type.
[0076] The immunogenic composition can comprise 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, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 37, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, or more tumor-specific neoantigens. The immunogenic composition can contain about 10-20 tumor-specific neoantigens, about 10-30 tumor-specific neoantigens, about 10-40 tumor-specific neoantigens, about 10-50 tumor-specific neoantigens, about 10-60 tumor-specific neoantigens, about 10-70 tumor-specific neoantigens, about 10-80 tumor-specific neoantigens, about 10-90 tumor-specific neoantigens, or about 10-100 tumor-specific neoantigens. Preferably, the immunogenic composition contains at least about 10 tumor-specific neoantigens. Also preferred is an immunogenic composition containing at least about 20 tumor-specific neoantigens.
[0077] The immunogenic composition can further comprise a natural or synthetic antigen, which can increase the immune response. Exemplary natural or synthetic antigens include, but are not limited to, the pan-DR epitope (PADRE) and the tetanus toxin antigen.
[0078] The immunogenic composition can be in any form, for example, a synthetic long peptide, RNA, DNA, a cell, a dendritic cell, a nucleotide sequence, a polypeptide sequence, a plasmid, or a vector.
[0079] Tumor-specific neoantigens can also be included in viral vector-based vaccine platforms, such as vaccinia, avian pox, self-replicating alphavirus, Maraba virus, and adenovirus (see, e.g., Tatsis et al., Molecular Therapy, 10:616-629 (2004)), or lentiviruses, including, but not limited to, any generation of second-, third-, or hybrid second / third-generation lentiviruses and recombinant lentiviruses designed to target specific cell types or receptors (see, e.g., Hu et al., Immunol Rev., 239(1):45-61 (2011); Sakma et al., Biochem J., 443(3):603-18 (2012)). Depending on the packaging capacity of the viral vector-based vaccine platform described above, this approach can deliver one or more nucleotide sequences encoding one or more tumor-specific neoantigen peptides. The sequence may be adjacent to non-mutated sequences, separated by a linker, or preceded by one or more sequences that target subcellular compartments (see, e.g., Gros et al., Nat Med., 22(4):433-8 (2016); Stronen et al., Science., 352(6291):1337-1341 (2016); Lu et al., Clin Cancer Res., 20(13):3401-3410 (2014)). Upon introduction into a host, infected cells express one or more tumor-specific neoantigens, thereby eliciting a host immune (e.g., CD8+ or CD4+) response against the one or more tumor-specific neoantigens. Vaccinia vectors and methods useful in immunization protocols are described, for example, in U.S. Patent No. 4,722,848. Another vector is BCG (Bacille Calmette Guerin). BCG vectors are described in Stover et al. (Nature 351:456-460 (1991)). A wide variety of other vaccine vectors useful for therapeutic administration or immunization of neoantigens, as will be apparent to those skilled in the art from the description herein, can also be used.
[0080] The immunogenic composition can contain components that are individualized according to the individual needs of a particular subject.
[0081] The immunogenic compositions described herein can further include an adjuvant. An adjuvant is any substance that, when added to an immunogenic composition, increases or otherwise enhances and / or boosts the immune response to tumor-specific neoantigens, but does not generate an immune response to tumor-specific neoantigens when administered alone. The adjuvant preferably generates an immune response to the neoantigen and does not generate an allergic or other adverse reaction. It is contemplated herein that the immunogenic composition can be administered prior to, together with, in combination with, or after the administration of the immunogenic composition.
[0082] Adjuvants can enhance immune responses by several mechanisms, including, for example, lymphocyte recruitment, stimulation of B and / or T cells, and stimulation of macrophages. When the immunogenic compositions of the invention comprise an adjuvant or are administered with one or more adjuvants, adjuvants that can be used include, but are not limited to, mineral salt or mineral salt gel adjuvants, particulate adjuvants, microparticulate adjuvants, mucosal adjuvants, and immunostimulatory adjuvants. Examples of adjuvants include aluminum salts (alum) (such as aluminum hydroxide, aluminum phosphate, and aluminum sulfate), 3De-O-acylated monophosphoryl lipid A (MPL) (see GB2220211), MF59 (Novartis), AS03 (GlaxoSmithKline), AS04 (GlaxoSmithKline), polysorbate 80 (Tween 80; ICL Americas, Inc.), imidazopyridine compounds (see International Application No. PCT / US2007 / 064857, published as WO2007 / 109812), imidazoquinoxaline compounds (see International Application No. PCT / US2007 / 064858, published as WO2007 / 109813), and saponins, such as QS21 (Kensil et al, in Vaccine Design: The Subunit and Adjuvant Approach (eds. Powell & Newman, Plenum Press, NY, 1995), U.S. Patent No. 5,057,540). In some embodiments, the adjuvant is Freund's adjuvant (complete or incomplete). Another adjuvant is an oil in water emulsion (such as squalene or peanut oil), optionally combined with an immunostimulant such as monophosphoryl lipid A (see Stoute et al., N. Engl. J. Med. 336, 86-91 (1997)).
[0083] CpG immunostimulatory oligonucleotides have also been reported to enhance the effect of adjuvants in vaccine settings. Other TLR-binding molecules, such as RNA-binding TLR7, TLR8, and / or TLR9, may also be used.
[0084] Other examples of useful adjuvants include, but are not limited to, chemically modified CpG (e.g., CpR, Idera), poly(I:C) (e.g., polypoly:CI2U), poly-ICLC, non-CpG bacterial DNA or RNA, and immunologically active small molecules and antibodies such as cyclophosphamide, sunitomib, bevacizumab, Celebrex (celecoxib), NCX-4016, sildenafil, tadalafil, vardenafil, sorafinib, XL-999, CP-547632, pazopamb, ZD2171, AZD2171, ipilimumab, tremelimab, and SC58175, which may act therapeutically and / or as an adjuvant. In embodiments, poly-ICLC is a preferred adjuvant.
[0085] Immunogenic compositions can contain one or more tumor-specific neoantigens described herein, either alone or together with a pharmaceutically acceptable carrier. One or more tumor-specific neoantigens can be used, particularly isotonic aqueous suspensions, dispersions, or suspensions or dispersions in amphiphilic solvents. The immunogenic compositions can be sterilized and / or contain excipients, such as preservatives, stabilizers, wetting agents and / or emulsifiers, solubilizers, salts for adjusting osmotic pressure and / or buffers, and are prepared in a manner known per se, for example, by conventional dispersion and suspension processes. In certain embodiments, such dispersions or suspensions can contain viscosity-adjusting agents. The suspensions or dispersions can be maintained at a temperature of about 2°C to 8°C, or can be preferentially frozen for longer storage and then thawed immediately before use. For injection, vaccines or immunogenic preparations can be formulated in aqueous solutions, preferably physiologically compatible buffers such as Hank's solution, Ringer's solution, or physiological saline buffer. The solutions can contain formulating agents such as suspending, stabilizing, and / or dispersing agents.
[0086] In certain embodiments, the compositions described herein further comprise a preservative, e.g., the mercury derivative thimerosal. In certain embodiments, the pharmaceutical compositions described herein comprise 0.001% to 0.01% thimerosal. In other embodiments, the pharmaceutical compositions described herein are preservative-free.
[0087] An excipient can exist independently of an adjuvant. The function of an excipient can be, for example, to increase the molecular weight of the immunogenic composition, increase activity or immunogenicity, confer stability, increase biological activity, or increase serum half-life. An excipient can also be used to aid in the presentation of one or more tumor-specific neoantigens to T cells (e.g., CD4+ or CD8+ T cells). The excipient can be a carrier protein, such as, but not limited to, a serum protein such as keyhole limpet hemocyanin or transferrin, bovine serum albumin, human serum albumin, thyroglobulin or ovalbumin, an immunoglobulin, or a hormone such as insulin or palmitic acid. For human immunization, the carrier is generally a physiologically acceptable carrier that is tolerated and safe by humans. Alternatively, the carrier can be a dextran, e.g., Sepharose.
[0088] Cytotoxic T cells recognize antigens in the form of peptides bound to MHC molecules, rather than intact foreign antigens themselves. MHC molecules themselves are located on the cell surface of antigen-presenting cells. Therefore, when a trimeric complex of peptide antigen, MHC molecule, and antigen-presenting cell (APC) is present, cytotoxic T cells can be activated. If not only one or more tumor-specific antigens are used to activate cytotoxic T cells, but additional APCs bearing the respective MHC molecules are added, it can enhance the immune response. Therefore, in some embodiments, the immunogenic composition further contains at least one APC.
[0089] The immunogenic composition can contain an acceptable carrier (e.g., an aqueous carrier). A variety of aqueous carriers can be used, such as water, buffered water, 0.9% saline, 0.3% glycine, hyaluronic acid, and the like. These compositions can be sterilized by conventional, well-known sterilization techniques or sterile filtered. The resulting aqueous solutions can be packaged for use as is or lyophilized, with the lyophilized preparation being combined with a sterile solution prior to administration. The composition can contain pharmaceutically acceptable auxiliary substances necessary to approximate physiological conditions, such as pH adjusting and buffering agents, tonicity adjusting agents, wetting agents, and the like, for example, sodium acetate, sodium lactate, sodium chloride, potassium chloride, calcium chloride, sorbitan monolaurate, triethanolamine oleate, and the like.
[0090] Neoantigens can also be administered via liposomes, which target specific cellular tissues, such as lymphoid tissues. Liposomes are also useful for increasing half-life. Liposomes include emulsions, foams, micelles, insoluble monolayers, liquid crystals, phospholipid dispersions, lamellar layers, and the like. In these preparations, the neoantigen to be delivered is incorporated as part of the liposome, either alone or in combination with a molecule that binds to a receptor prevalent among lymphoid cells, such as a monoclonal antibody that binds to the CD45 antigen, or with other therapeutic or immunogenic compositions. Thus, liposomes filled with the desired neoantigen can be targeted to the site of lymphoid cells, where they then deliver the selected immunogenic composition. Liposomes can be formed from standard vesicle-forming lipids, typically including neutral and negatively charged phospholipids and sterols, such as cholesterol. The choice of lipid is generally guided by considerations, for example, of liposome size, acidity, and stability of the liposomes in the bloodstream. Various methods for preparing liposomes are available, for example, Szoka et al., An. Rev. Biophys. Bioeng. 9;467 (1980), U.S. Pat. Nos. 4,235,871, 4,501,728, 4,501,728, 4,837,028, and 5,019,369.
[0091] To target immune cells, the ligand incorporated into the liposomes can include, for example, an antibody or fragment thereof specific for a cell surface determinant of the desired immune system cell. The liposome suspension can be administered intravenously, topically, locally, etc., at doses that vary depending, inter alia, on the method of administration, the peptide being delivered, and the stage of the disease being treated.
[0092] An alternative method for targeting immune cells, components of immunogenic compositions, such as antigens (i.e., tumor-specific neoantigens), ligands, or adjuvants (e.g., TLRs), can be incorporated into poly(lactic acid-co-glycol) microparticles, which can act as endosomal delivery devices to entrap components of immunogenic compositions.
[0093] Nucleic acids encoding the tumor-specific neoantigens described herein can also be administered to patients for therapeutic or immunization purposes. Several methods are suitable for delivering nucleic acids to patients. For example, nucleic acids can be delivered directly as "naked DNA." For this approach, see, for example, Wolff et al., Science 247:1465-1468 (1990), and U.S. Patent Nos. 5,580,859 and 5,589,466. Nucleic acids can also be administered using ballistic delivery, as described, for example, in U.S. Patent No. 5,204,253. Particles consisting solely of DNA can be administered. Alternatively, DNA can be attached to particles, such as gold particles. Approaches for delivering nucleic acid sequences can include viral vectors, mRNA vectors, and DNA vectors, with or without electroporation. Nucleic acids can also be complexed to cationic compounds, such as cationic lipids, for delivery.
[0094] The immunogenic compositions provided herein can be administered to a subject by routes including, but not limited to, oral, intradermal, intratumoral, intramuscular, intraperitoneal, intravenous, topical, subcutaneous, transdermal, intranasal, and inhalation routes, and can be administered via scarification (e.g., using a bifurcated needle to scrape the top layer of skin). The immunogenic compositions can be administered to a tumor site to induce a local immune response against the tumor.
[0095] The dosage of one or more tumor-specific neoantigens may depend on the type of composition, as well as the subject's age, weight, body surface area, individual condition, individual pharmacokinetic data, and mode of administration.
[0096] Also disclosed herein are methods for producing immunogenic compositions comprising one or more tumor-specific neoantigens selected by performing the steps of the methods disclosed herein. The immunogenic compositions described herein can be produced using methods known in the art. For example, a method for producing a tumor-specific neoantigen or vector (e.g., a vector comprising at least one sequence encoding one or more tumor-specific neoantigens) disclosed herein can include culturing host cells under conditions suitable for expression of the neoantigen or vector, where the host cells comprise at least one polynucleotide encoding the neoantigen or vector, and purifying the neoantigen or vector. Standard purification methods include chromatography, electrophoresis, immunological methods, precipitation, dialysis, filtration, concentration, and chromatofocusing.
[0097] The host cells may include Chinese hamster ovary (CHO) cells, NS0 cells, yeast cells, or HEK293 cells. The host cells may be transformed with one or more polynucleotides comprising at least one nucleic acid sequence encoding one or more tumor-specific neoantigens or vectors disclosed herein. In certain embodiments, the isolated polynucleotide may be a cDNA.
[0098] IV. Sample The methods disclosed herein include selecting one or more tumor-specific neoantigens derived from a tumor. The method for selecting one or more tumor-specific neoantigens includes obtaining sequence data derived from the tumor. Such sequence data can be derived from a tumor sample from a subject. The tumor sample can be obtained from a tumor biopsy.
[0099] Tumor samples can be obtained from human or non-human subjects. Preferably, tumor samples are obtained from humans. Tumor samples can be obtained from a variety of biological sources, including cancerous tumors. Tumors can be derived from tumor sites or from circulating tumor cells derived from blood. Exemplary samples include, but are not limited to, bodily fluids, tissue biopsies, blood samples, serum plasma, stool, skin samples, and the like. The source of the sample can be a solid tissue sample, such as a tumor tissue biopsy. Tissue biopsy samples can be, for example, biopsies from the lung, prostate, colon, skin, breast tissue, or lymph nodes. Samples can also be samples of bone marrow, including, for example, bone marrow aspirates and bone marrow biopsies. Samples can also be liquid biopsies, such as circulating tumor cells, cell-free circulating tumor DNA, or exosomes. Blood samples can be whole blood, partially purified blood, or portions of whole blood or partially purified blood, such as peripheral blood mononuclear cells (PBMCs).
[0100] The tumor sample described herein can be obtained directly from a subject, derived from a subject, or derived from a sample obtained from a subject, such as a biological fluid or tissue sample, derived from cultured cells. The tumor biopsy can be a fresh sample. The fresh sample can be fixed after removal from the subject using any known fixative (e.g., formalin, Zenker's fixative, or B-5 fixative). The tumor biopsy can also be an archived sample, such as a frozen sample of cells obtained directly from a subject, a cryopreserved sample, or the like, derived from cells obtained from a subject. Preferably, the tumor sample obtained from a subject is a fresh tumor biopsy.
[0101] Tumor samples can be obtained from subjects by any means, including but not limited to tumor biopsy, needle aspiration, scraping, surgical resection, surgical incision, venipuncture, or other means known in the art.Tumor biopsy is the preferred method for obtaining tumors.Tumor biopsy can be obtained from any cancerous site, for example, primary tumor or secondary tumor.Tumor biopsy from primary tumor is generally preferred.Those skilled in the art will recognize other suitable techniques for obtaining tumor samples.
[0102] Tumor samples can be obtained from a subject in a single procedure. Tumor samples can be obtained from a subject repeatedly over a period of time. For example, tumor samples can be obtained daily, weekly, monthly, semi-annually, or annually. Obtaining multiple samples over a period of time can be useful for identifying and selecting new tumor-specific neoantigens. Tumor samples can be obtained from the same tumor or different tumors.
[0103] Tumor samples can be obtained from the primary tumor, one or more metastases, and / or individual sites of tumor growth (e.g., bone marrow from different skeletal parts such as hips, bones, or vertebrae). Tumor samples can be obtained from the same site or different sites.
[0104] equivalent Other suitable modifications and adaptations of the methods of the present invention described herein will be apparent and will be readily apparent to those skilled in the art, which may be made using suitable equivalents without departing from the scope of the present disclosure or embodiments. Having described certain compositions and methods in detail, the same will be more clearly understood by reference to the following examples, which are not intended to be limiting and are presented for purposes of illustration only. [Example]
[0105] The following are examples of methods and compositions of the present invention. Upon reading the summary provided herein, it will be understood that various other embodiments may be practiced.
[0106] Example 1. Selection of neoantigen peptides This example describes the individual procedural steps for selecting neo-antigenic immunogenic peptides identified from next generation sequencing data generated from patient tumor and normal tissues.
[0107] 1.1. Sample preparation and generation of WES, WGS, and RNA-Seq data Tumor biopsies or surgical implants were collected from study participants with informed consent and transported in tissue culture medium on ice to the Clinical Trials CLIA laboratory, where the samples were received and assigned a specific, unique sample identifier. The tissue was then weighed, divided, and placed in five (5) volumes of RNAlater Stabilization Solution (ThermoFisher, Cat. No. AM7020). The samples were then left overnight at 4°C, removed from the RNAlater solution, placed in a cryotube containing 1 mL of STEMCELL CroyStor 10 (Cat. No. 07952), and transferred to a CoolCell (Corning, Cat. No. 432000) at -80°C.
[0108] Peripheral blood collected in ACD tubes from associated protocol participants was transported to the specimen processing and research cell bank for PBMC processing using Ficoll according to SOPs. PBMC processing may be performed prior to tumor biopsy, allowing for simultaneous shipment of PBMC and tumor biopsy tissue to the sequencing provider. [Table 1]
[0109] All specimens (tumor biopsies and PBMCs) were sent for sequencing provided overnight on dry ice by a CLIA lab.
[0110] DNA, RNA, and miRNA were simultaneously isolated from the same tissue or cell specimen using the AllPrep DNA / RNA / miRNA Universal Kit (QIAGEN).
[0111] The quality and quantity of the DNA / RNA samples were assessed using appropriate methods (e.g., Qubit, BioAnalyzer) and the following metrics were recorded:
[0112] DNA: concentration (ng / μL), total amount (ng), volume (μL)
[0113] RNA: Concentration (ng / μL), total amount (ng), volume (μL), purity (RIN) [Table 2]
[0114] DNA sample aliquots containing greater than 200 ng of genomic DNA were prepared, 200 ng was used for WES, and the remaining DNA was shipped to another sequencing provider for WGS.
[0115] NGS was performed and the library preparation and sequencing strategy is outlined in Table 3 below. [Table 3]
[0116] WES data from paired tumor / normal samples were used to identify germline and somatic variants in patients and their tumors.
[0117] WES was performed at a commercial sequencing vendor using the Agilent SureSelect All Exon v6 bait kit, and libraries were generated and sequenced on an Illumina NovaSeq6000 instrument. DNA from PBMC samples was subjected to WES using a 100bp PE strategy, 75x average coverage, and a target of 38 million reads. DNA from tissue samples was subjected to WES using a 100bp PE strategy, 125x average coverage, and a target of 63 million reads.
[0118] RNA sequencing data were used to identify neoantigens encoding RNA transcripts with sufficiently high expression as well as to independently confirm WES-derived somatic variants.
[0119] RNA was sequenced using Illumina Stranded mRNA sequencing using a 50 bp PE strategy and a target of 100 million reads.
[0120] Sequencing libraries were generated using the Illumina TruSeq Stranded mRNA method, which utilizes polyadenylated tails to preferentially select for messenger RNA.
[0121] WGS data was used to perform CNV calling and identify subclones in tumor samples.
[0122] WGS was performed in a commercial CLIA-validated laboratory from tumor and normal genomic DNA prepared as detailed above. Two pooled libraries from the same individual were sequenced on an Illumina S4 flow cell (FC) with a read length of 2 x 10. Data generated with FC were passed through for demultiplexing with a Q30 > 80% and an error rate < 3%.
[0123] NGS data in FASTQ format were transferred from the sequencing vendor for further bioinformatics analysis.
[0124] HLA typing was performed using molecular assays in an accredited clinical immunogenetics laboratory.
[0125] Bioinformatics analysis of NGS data Mapping and alignment of NGS reads to the hg19 reference genome was performed using the Illumina DRAGEN Bio-IT platform (v3.8.5). Details of the input and output files and the automated execution of the processing steps used for mapping and alignment are shown in Figure 2.
[0126] All NGS analyses were performed using the hg19 human reference genome assembly (initial UCSC reference assembly based on the initial GRCh37 release, md5 checksum a244d8a32473650b25c6e8e1654387d6, downloaded from the Sentieon reference bundle). Mappings between known ENSEMBL genes and hg19 chromosomal coordinates (GTF files) were downloaded from UCSC. These files are used to quantify RNA gene expression. The DRAGEN Bio-IT platform was used to generate a series of hash table files required for mapping, alignment, and variant calling. The generated hash table files and hg19 ensemble GTF files were uploaded to an S3 data storage bucket.
[0127] The DNA sequence mapping and alignment step takes NGS FASTQ files as input and aligns reads to the provided reference genome hash table, independently for normal and tumor (if provided) samples. Normal mapping / alignment included the generation of germline variant calls used in module A3 for somatic CNV calling (referred to as the B allele file). [Table 4-1] [Table 4-2]
[0128] In addition to the mapping / alignment algorithms, this module utilized DRAGEN's reporting features (including mapping statistics and trimming reports) which were used for quality control.
[0129] For mapping / alignment of tumor samples, the command lines are outlined in Table 5. [Table 5-1] [Table 5-2]
[0130] The RNA-Seq FASTQ files were aligned to the UCSC hg19 human reference genome using the DRAGEN Bio-IT platform (v3.8.5) RNA module, which allows for quantification and gene fusion detection. The GENCODE hg19 / GRCg37.p13 GTF file was used to map genes and gene transcripts (Ensembl gene and transcript IDs) to genomic regions. [Table 6-1] [Table 6-2]
[0131] Normal (PBMC) WES and WGS .BAM files aligned to the UCSC hg19 human reference genome were used separately to generate a list of germline variant calls (.VCF files) showing the differences between the study participant's genome and the reference genome. Detected variants were single or multiple base mutations, insertions, and deletions. Structural variants were not processed. Detailed command line arguments for deriving germline variant calls and generating unfiltered and filtered germline .VCF files using the DRAGEN Bio-IT platform (v3.8.5) are described below.
[0132] The resulting germline .VCF file was used to confirm that the candidate vaccine peptides did not represent the germline sequences (self-peptides) of the study participants, and CNV calls were also provided as the input B-allele frequency file.
[0133] The DRAGEN Bio-IT platform was used to identify somatic variants arising as differences between tumor and normal study participant samples from .BAM files aligned with WES and WGS. In the first step, tumor and normal .BAM files were compared to identify tumor-specific (somatic) DNA mutations, which were output as .VCF files: one unfiltered and one filtered for high-confidence variants. In the second step, for WGS only, CNV calls were generated and output as .VCF files. The inputs for this step were the tumor and normal .BAM files, as well as the hard-to-filter germline variant call file used as input for B-allele frequencies.
[0134] Additional details and command line interface options for performing somatic variant and CNV calling are provided in Table 8. [Table 7-1] [Table 7-2]
[0135] The somatic .VCF files (difficult to filter) were used in the downstream vaccine peptide selection module.
[0136] CNV calling in DRAGEN requires the B-Allele.VCF file. CNV calling was preceded by germline calling and normal / tumor.BAM files from module A1. [Table 8-1] [Table 8-2]
[0137] To analyze the validity of the mapping / alignment across contigs, module A-QC1 employed the mosdepth analysis program with the .BAM file as input. Additionally, for WES, a .BED file was provided to restrict the analysis to a defined genomic region. The hg19 BED file for the Agilent Sure Select All Exon v6 capture bait set was downloaded from the Agilent website and saved to S3 for automatic download by the processing pipeline. [Table 9-1] [Table 9-2]
[0138] The mosdepth summary file is a tab-delimited text file showing the average coverage (column average) of all contigs present in the .BAM file (column chrome). If a .BED file is provided, the suffix "_region" indicates the metric of the .BED-restricted contig. An abbreviated listing of the mosdepth summary file is shown (e.g., Table 10). [Table 10]
[0139] For analysis and quality control, tumor and normal mosdepth summary files were used to generate specification reports.
[0140] Examples of mapping alignment results are shown in Tables 11 to 15 below. [Table 11-1] [Table 11-2]
[0141] [Table 12-1] [Table 12-2]
[0142] [Table 13]
[0143] [Table 14]
[0144] [Table 15]
[0145] The results of the variant calling process are shown in Tables 16-19. [Table 16]
[0146] [Table 17]
[0147] [Table 18]
[0148] [Table 19]
[0149] Subsequent modules in the workflow process tumor somatic variants and copy number variants from tumor-normal WGS samples, as called by Module A. They output the membership probability of each somatic variant into a set of N tumor-specific subclones, where N is a parameter output by the module. They also output an estimate of the cellular prevalence of each variant. This module also performed bulk deconvolution of variants into subclones. WGS files generated from DRAGEN in Module A were processed by this module to estimate subclonality from the cellular prevalence of somatic variants. Figure 3 shows a workflow diagram of the module for clonality deconvolution.
[0150] 1.3.Selection of peptides from NGS data The vaccine peptides to be produced were selected by peptide prediction and machine learning algorithms within the personalized peptide prediction pipeline (p4vax). All components of this software solution are briefly described herein.
[0151] The mapping / alignment, RNA gene expression, germline variant, somatic variant, and CNV caller output files, and HLA haplotyping results are ingested by the peptide prediction workflow to: 1. Identify variants with coding effects on expressed RNA 2. Select putative MHC class I and II binding peptides spanning non-synonymous variants 3. Confirm the presence of the variant in the RNA sequencing reads 4. Estimation of MHC Class I and II Processing, Presentation, and Immunogenicity 5. Correcting for the hg19 reference in germline variant calling confirms that the mutation is present only in tumor DNA but not in germline (normal) DNA 6. Filtering potentially toxic peptides or products of endopeptidase or exopeptidase metabolism 7. Ranking peptides for MHC class I and II processing, presentation, and immunogenicity 8. Maximizing the expected percentage of tumor cells targeted by the peptide.
[0152] Specifically, WES based somatic variant calling and results from WES were used for peptide selection.
[0153] Using the Tumor-Subclone Deconvolution algorithm, the workflow processes tumor somatic variants and copy number variants called by Module A from tumor-normal WGS samples and further outputs the membership probability of each somatic variant into a set of N tumor-specific clusters of mutation / cell prevalence. The results of this step were used in Step 8.
[0154] The output of the peptide prediction pipeline is an exhaustive list of potential vaccine peptides ranked by a combination of MHC class I and II binding / presentation scores, immunogenicity, and tumor cell prevalence clusters.
[0155] From this ranked list of potential peptide synthesis candidates, we selected a set of up to 80 peptides that ranked highest within the list. After manually verifying the presence of somatic variants, the peptide sequences resulting from the verified somatic variants were communicated to a peptide manufacturer for synthesis.
[0156] Peptide selection was performed for the example samples DDDDD, FFFFF, and GGGGG (Tables 20 to 22), and 20 peptides were further selected as candidates for a pooled formulation. [Table 20-1] [Table 20-2] [Table 20-3] [Table 20-4] [Table 20-5] [Table 20-6]
[0157] [Table 21-1] [Table 21-2] [Table 21-3] [Table 21-4] [Table 21-5]
[0158] [Table 22-1] [Table 22-2] [Table 22-3] [Table 22-4] [Table 22-5]
[0159] 1.4. Peptide Production Vaccine peptide production and pool formulation was performed at a peptide manufacturer that performed peptide synthesis, quality control, dissolving and mixing of peptides and peptide pools.
[0160] Peptides were prepared by solid-phase peptide synthesis, purified using a RP-HPLC column, and analyzed for quality (identity, purity, peptide content, acetate / TFA content, residual organic solvents).
[0161] The results of peptide synthesis are shown in Tables 23 to 25. [Table 23-1] [Table 23-2] [Table 23-3] [Table 23-4] [Table 23-5]
[0162] [Table 24-1] [Table 24-2] [Table 24-3] [Table 24-4] [Table 24-5]
[0163] [Table 25-1] [Table 25-2] [Table 25-3] [Table 25-4] [Table 25-5] [Table 25-6] [Table 25-7] [Table 25-8]
[0164] The produced peptides, as determined by passing quality control criteria (appearance, identity, peptide content, peptide purity, acetic acid content, TFA content, residual organic solvents), were then subjected to a pool optimization algorithm.
[0165] Briefly, the algorithm distributed selected peptides into four pools, optimizing the distribution so that each pool contained peptides with high MHC class I and class II scores. After first excluding sequences containing more than two cysteines to avoid multimerization during pool formulation, the optimization algorithm identified the best combination of approximately 14 long vaccine peptides and 6 short peptides to be combined into individual vaccine peptide pools. If the initial peptide prediction or generation yielded fewer than 14 long vaccine peptides, the long-to-short peptide ratio could be altered to accommodate the target of four pools with five peptides each. Similarly, if the vaccine peptide pool consisted of fewer than three (3) long peptides, which extrapolated to a low probability of CD4 engagement, one of the pool peptides was selected to be PADRE.
[0166] Several proposed pool compositions were then communicated to peptide manufacturers and vaccine peptide pool formulations were carried out.
[0167] Groups of peptides were selected and pooled. Up to four (4) pools were prepared with no more than five (5) peptides in each pool (NMT). Each peptide was dissolved in 5.538% (v / v) DMSO, followed by 0.9% NaCl solution at a concentration of 0.4158 mg / mL of each peptide. Peptides that were visibly dissolved were accepted into that group's pool. After all peptides in a group were successfully pooled, the pool was filtered through a 0.2 μm nylon filter. Peptides were shipped in sterile tubes and labeled with the pool name, lot number, ingredients, manufacturing data, quantity / concentration, and storage conditions.
[0168] 1.5. Peptide vaccine formulation To formulate the personalized vaccine, the peptide pool was mixed with poly-ICLC as an adjuvant using the following procedure:
[0169] Remove one peptide vial per pool from the freezer. For each peptide pool, perform the following steps:
[0170] Thaw the peptide pool at room temperature for 20-30 minutes. The peptide solution contains DMSO and may require slight warming (hand warming) and / or stirring for complete thawing. The material is expected to be a clear, colorless solution. If a precipitate forms, vortex the solution to dissolve the precipitate.
[0171] While thawing, prepare and label syringes for steps 4 through 9 (see Table 26): (A) Two (2) mixing syringes per pool labeled "M1-2 [AD]" (indicate pool): (i) Syringe 1: 10 mL (or appropriate size), and (ii) Syringe 2: 1 mL (or appropriate size); and (B) One (1) dosing syringe per pool labeled "Dosing Syringe [AD]" (indicate pool). [Table 26]
[0172] Additional material required (per peptide pool) One (1) guarded female to female Luer adapter Transfer the peptide pool into one (1) compound syringe M1 with a 3" suction needle One (1) non-pyrogenic, sterile, low protein-binding syringe filter, 0.22 μm pore size. (e.g., Pall DMSO-Safe Aerodisc Syringe Filter, #4433, or Millex-GV 0.22 μm PVDF, 33 mm, gamma-sterilized, Millipore, #SLGVM33RS) Three (3) sterile hypodermic needles of appropriate gauge One (1) IM Injection Spiros Closed System Drug Delivery Device (CSTD) [ICU Medical, #SH2000SC-10] Filter-sterilize the contents of each vial: using a 10 mL syringe and Luer-lock (mixing syringe 1, M1) connected to a 0.22 µm sterile filter, slide a 3 inch aspirating needle on the assembly and withdraw 2 mL of the peptide pool from the vial.
[0173] Prepare the poly-ICLC solution. Using a 1 mL syringe (mixing syringe 2, M2) and an appropriately sized needle, aspirate 0.760 mL of poly-ICLC under sterile conditions. Poly-ICLC is a white, translucent solution. Downstream steps may result in a translucent product; this is acceptable.
[0174] Remove the needles from the mixture of syringe 1 (M1) and syringe 2 (M2) and remove the sterile filter from the mixture of syringe 1 (M1).
[0175] Connect Poly-ICLC mixing syringe 2 (M2) to peptide pool mixing syringe 1 (M1) via a female-to-female Luer lock guard connector.
[0176] Transfer Poly-ICLC mixing syringe 2 (M2) to mixing syringe 1 (M1) and mix thoroughly between both syringes. Bubbles may form in the product. Tapping the syringe may help collect the bubbles. The resulting mixture is the personalized vaccine.
Claims
1. 1. A method for selecting one or more tumor-specific neoantigens from a tumor in a subject for a subject-specific immunogenic composition, comprising: a) obtaining tumor sequence data from the tumor; b) obtaining data representing the polypeptide sequence of each of one or more tumor-specific neoantigens from tumor sequence data derived from the tumor; c) inputting the polypeptide sequence of each of the one or more tumor-specific neoantigens and the MHC molecule of the subject into a machine learning platform to generate a probability numerical score that each of the one or more tumor-specific neoantigens will elicit an immune response in the subject; d) quantifying RNA expression of each of the one or more tumor-specific neoantigens in the tumor to identify each of the one or more tumor-specific neoantigens that are expressed in an amount sufficient to elicit an immune response in the subject; e) sequencing the tumor clones of the tumor to identify each one or more tumor-specific neoantigens that represent a sufficient proportion of said tumors; f) calculating a tumor-specific neoantigen score for each of the one or more tumor-specific neoantigens based on the probability numerical score generated in step c) and the RNA expression quantified in step d); g) selecting each of the one or more tumor-specific neoantigens for formulation of a subject-specific immunogenic composition based on the tumor-specific neoantigen score.
2. The method of claim 1, wherein the tumor-specific neoantigen score is calculated based on the probability numerical score generated in step c), the RNA expression quantified in step d) and the sequencing of the tumor clone in step e).
3. 3. The method according to claim 1 or 2, wherein the polypeptide sequence encoding one or more tumor-specific neoantigens of step b) is derived from a short polypeptide, preferably the short polypeptide is presented by an MHC class I molecule, or the polypeptide sequence encoding one or more tumor-specific neoantigens of step b) is derived from a long polypeptide.
4. the immune response in the subject comprises the presentation of one or more tumor-specific neoantigens on the tumor cell surface; the immune response in the subject comprises presentation of one or more tumor-specific neoantigens by one or more MHC molecules on tumor cells; The method of any one of claims 1 to 3, wherein the immune response in the subject comprises one or more tumor-specific neoantigens that can be presented to T cells by antigen-presenting cells and / or the immune response in the subject is a CD4+-mediated response or a CD8+-mediated response.
5. 5. The method of claim 1, wherein a tumor-specific neoantigen having a higher probability numerical score compared to a lower probability numerical score indicates that the tumor-specific neoantigen induces a greater immune response in the subject.
6. 6. The method of any one of claims 1 to 5, wherein a sufficient proportion of the tumor represented by the one or more tumor-specific neoantigens represents at least about 1% of the tumor, and preferably represents at least about 5% of the tumor.
7. The method of any one of claims 1 to 6, wherein the tumor sequence data is nucleotide sequence data and / or polypeptide sequence data.
8. 8. The method of any one of claims 1 to 7, wherein the tumor sequence data is exome, transcriptome, or whole genome nucleotide sequence data, or the tumor sequence data is whole exome sequence data, RNA sequence data, whole genome sequence data, or a combination thereof, preferably the tumor sequence data is a combination of whole exome sequence data, RNA sequence data, and whole genome sequence data.
9. 9. The method according to any one of claims 1 to 8, wherein at least about 10 tumor-specific neoantigens are selected for formulating the subject-specific immunogenic composition, and preferably at least about 20 tumor-specific neoantigens are selected for formulating the subject-specific immunogenic composition.
10. The method of any one of claims 1 to 8, wherein the subject-specific immunogenic composition comprises one or more tumor-specific neoantigens encoded by short polypeptides and / or the subject-specific immunogenic composition comprises one or more tumor-specific neoantigens encoded by long polypeptides.
11. 13. An immunogenic composition for use in a method of treating a subject having a tumor, the composition being formulated by a method performing the steps of claim 1, and further comprising formulating an immunogenic composition comprising a selected one or more of tumor-specific neoantigens.
12. An immunogenic composition for use in the method according to any one of claims 1 to 10 or the method according to claim 11, said method further comprising measuring the ability of one or more tumor-specific neoantigens to induce an autoimmune response against normal tissue, preferably wherein one or more tumor-specific neoantigens that induce an autoimmune response against normal tissue are not selected for said immunogenic composition and / or A method or immunogenic composition, wherein one or more tumor-specific neoantigens that induce an autoimmune response against normal tissue have a lower tumor-specific neoantigen score compared to tumor-specific neoantigens that do not induce an autoimmune response.
13. An immunogenic composition for use in the method of claims 1 to 10 or 12, or the method of claim 11 or 12, wherein the tumor is selected from melanoma, breast cancer, ovarian cancer, prostate cancer, renal cancer, gastric cancer, colon cancer, testicular cancer, head and neck cancer, pancreatic cancer, brain cancer, B-cell lymphoma, acute myeloid leukemia, chronic myeloid leukemia, chronic lymphocytic leukemia, T-cell lymphocytic leukemia, bladder cancer and lung cancer, preferably the tumor is selected from a cancer selected from the group consisting of melanoma, breast cancer, lung cancer, and bladder cancer, more preferably the tumor is melanoma.
14. 11. A method for producing an immunogenic composition, comprising carrying out the steps of claim 1, and further comprising formulating an immunogenic composition comprising the selected one or more tumor-specific neoantigens.
15. 14. An immunogenic composition comprising one or more tumor-specific neoantigens selected by carrying out the method according to any one of claims 1 to 10, 12 and 13, preferably comprising a nucleotide sequence, a polypeptide sequence, RNA, DNA, a cell, a plasmid, a vector, a dendritic cell, or a synthetic long peptide, and optionally further comprising an adjuvant.
