Immunogenic cancer screening test

Abstract

The present disclosure relates to methods for determining the risk that a human subject will develop cancer, the methods comprising quantifying the HLA triplet (HLAT) of a subject, the HLA triplet being capable of binding a T cell epitope in an amino acid sequence of a tumor-associated antigen. The present disclosure also relates to methods of treating a subject determined to have an elevated risk of developing cancer.

Description

Technical Field

[0001] The present disclosure provides methods for determining the risk that a subject will develop cancer based on the subject's HLA class I genotype. The present disclosure further provides methods for treating cancer, particularly prophylactic treatment of subjects determined to have an elevated risk of developing cancer. Background Art

[0002] Possible screening and early diagnosis are crucial for preventing metastatic disease and improving the prognosis of many cancers.

[0003] Hereditary mutations can increase the risk of developing cancer, but known genetic factors do not fully account for the genetic contribution to cancer risk. For example, mutations in BRCA1 and BRCA2 have been identified in 5% of breast cancer cases in the general population, yet nearly 50% of these cases develop breast cancer. In the past decade, efforts to explain the heritability of cancer development have focused on the discovery of high-risk genes and the identification of common genetic variants.

[0004] However, there remains a need in the art for better identification of individuals at elevated genetic risk of developing cancer. Summary of the Invention

[0005] The present disclosure provides methods involving a subject's human leukocyte antigen (HLA) class I genotype as a predictor of cancer development.

[0006] In antigen-presenting cells (APCs), protein antigens, including tumor-associated antigens (TAAs), are processed into peptides. These peptides bind to HLA molecules and are presented as peptide-HLA complexes to T cells on the cell surface. Different individuals express different HLA molecules, and different HLA molecules present different peptides. The TAA epitopes that bind to a single HLA class I allele expressed in an individual are necessary but not sufficient to induce a tumor-specific T cell response. Instead, when the epitopes of TAAs are recognized and presented by HLA molecules encoded by at least three HLA class I genes of an individual (referred to herein as HLA triplets or "HLATs") (PCT / EP2018 / 055231, PCT / EP2018 / 055232, PCT / EP2018 / 055230, EP 3370065, and EP 3369431), the tumor-specific T cell response is optimally activated.

[0007] The inventors have developed a binary classifier capable of separating subjects with cancer from a background population. Using such a classifier, the inventors were able to demonstrate a clear association between HLA genotype and cancer risk. These findings confirm the central role of tumor-specific T cell responses in tumor growth control and imply that HLA genotype analysis can be used to improve diagnostic tests for early identification of subjects at high risk of developing cancer.

[0008] Thus, in a first aspect, the present disclosure provides a method for determining the risk that a human subject will develop cancer, the method comprising quantifying the HLA triplets (HLATs) in the amino acid sequences of the subject that are capable of binding tumor-associated antigens (TAAs), wherein each HLA of the HLAT is capable of binding the same T cell epitope, and determining the risk that the subject will develop cancer, wherein, with respect to the TAA, a smaller number of HLATs that are capable of binding the TAA corresponds to a higher risk that the subject will develop cancer.

[0009] The findings described herein also suggest that the risk of cancer can be reduced by using personalized vaccines that effectively activate the subject's immune system to kill tumor cells.

[0010] Thus, in another aspect, the present disclosure provides a method of treating cancer in a subject, wherein the subject has been determined to have an elevated risk of developing cancer using the method described above, and wherein the treatment method comprises administering to the subject one or more peptides or one or more polynucleic acids or a vector encoding one or more peptides, the peptides comprising an amino acid sequence that (i) is a fragment of a TAA; and (ii) comprises a T cell epitope capable of binding the HLAT of the subject.

[0011] In other aspects, the present invention provides

[0012] - a peptide or polynucleic acid or vector encoding a peptide for use in a method of treating cancer in a specific human subject, wherein the peptide comprises an amino acid sequence that (i) is a fragment of a TAA; and (ii) comprises a T cell epitope capable of binding the HLAT of the subject; and

[0013] - a peptide, or polynucleic acid or vector encoding a peptide, for use in the manufacture of a medicament for treating cancer in a specific human subject, wherein the peptide comprises an amino acid sequence that (i) is a fragment of a TAA; and (ii) comprises a T cell epitope capable of binding the HLAT of the subject.

[0014] In another aspect, the present disclosure provides a system for determining the risk that a human subject will develop cancer, the system comprising:

[0015] (i) A storage module configured to store data including the HLA class I genotype of a subject and the amino acid sequences of TAAs;

[0016] (ii) A calculation module configured to quantify the HLA-T of T cell epitopes in the amino acid sequences of the subject that are capable of binding the TAA, wherein each HLA of the HLA-T is capable of binding the same T cell epitope; and

[0017] (iii) An output module configured to display an indication of the risk that the subject will develop cancer and / or a recommended treatment for the subject.

[0018] (iv)

[0019] The methods and compositions of the present disclosure will now be described in more detail by way of example and not limitation and with reference to the accompanying drawings. When the present disclosure is given, many equivalent modifications and variations will be apparent to those skilled in the art. Accordingly, the exemplary embodiments of the invention set forth are considered to be illustrative and not restrictive. Various changes may be made to the described embodiments without departing from the scope of the present disclosure. All documents cited herein, whether supra or infra, are hereby expressly incorporated herein by reference in their entirety.

[0020] The present disclosure includes combinations of the described aspects and preferred features, except where such combinations are clearly not permitted or are stated to be expressly avoided. As used in this specification and the appended claims, the singular forms "a", "an" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a peptide" includes two or more such peptides.

[0021] The section headings used herein are for convenience only and should not be construed as limiting in any way. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1

[0023] ROC curve of HLA-restricted PEPI biomarkers.

[0024] Figure 2

[0025] ROC curve of ≥1 PEPI3+ tests for determining diagnostic accuracy. An AUC = 0.73 classified a reasonable diagnostic value for the PEPI biomarker.

[0026] Figure 3

[0027] ROC curve (HLAT score) of immunological predictors for classifying melanoma patients from the general population. AUC = 0.645; the black solid line is the ROC curve, and for comparison, the x = y line is indicated in dotted grey.

[0028] Figure 4

[0029] Relative immune risk of melanoma occurrence in five equally sized subgroups. The HLAT score ranges defining the subgroups are presented on the horizontal axis. The black bars indicate 95% confidence intervals. The difference between the first and last subgroups is significant (p = 0.001).

[0030] Figure 5

[0031] Relative immune risk of cancer occurrence in five equally sized subgroups. The HLAT score ranges defining the subgroups are presented on the horizontal axis. The black bars indicate 95% confidence intervals. A. Non-small cell lung cancer; B. Renal cell carcinoma; Colorectal cancer.

[0032] Figure 6

[0033] Relative risk (RR) of melanoma occurrence in five equally sized subgroups. The HLA score ranges defining the subgroups are shown on the x-axis. The black bars indicate 95% confidence intervals. The difference between the first and last subgroups is significant (p < 0.05).

[0034] Figure 7

[0035] Positive correlation between the number of antigens (n = 7) leading to vaccine-specific T cell responses (in 10 patients) and the HLAT score calculated for the group against 48 TSA.

[0036] Figure 8

[0037] Single HLA alleles or incomplete HLA genotypes have limitations in genotype-based separation of the UNPC group from the non-UNPC group. A*02:01 / B*18:01 AUC = 0.556 (not significant).

[0038] Figure 9

[0039] OBERTO trial design (NCT03391232)

[0040] Figure 10

[0041] Antigen expression in the CRC cohort of the OBERTO trial (n = 10). A: Expression frequency of PolyPEPI1018-derived antigens measured on 2391 biopsy tissues. B: The PolyPEPI1018 vaccine design, designated as 3 out of 7 TSAs, is expressed in CRC tumors with a probability higher than 95%. C: On average, 4 out of 10 patients had a pre-existing immune response against each target antigen, which refers to the true expression of TSA in the patient's tumor. D: 7 out of 10 patients had a pre-existing immune response against at least 1 TSA and on average against 3 different TSAs.

[0042] Figure 11

[0043] The immunogenicity of PolyPEPI1018 in CRC patients confirmed the appropriate target antigen and target peptide selection. Upper: Target peptide selection and peptide design of the PolyPEPI1018 vaccine composition. Two 15mers were selected from CRC-specific CTAs (TSAs) that contain 9mer PEPI3+ dominance in a representative model population. Table: During preclinical studies in the CRC cohort, the PolyPEPI1018 vaccine has been retrospectively tested and proven to be immunogenic against at least one antigen in all tested individuals by generating PEPI3+s. Clinical immune responses specific to at least one antigen were measured in 90% of the patients, multi-antigen immune responses against at least 2 antigens were also found in 90% of the patients, and multi-antigen immune responses against at least 3 antigens were also found in 80% of the patients, as tested by IFNγ ELISPOT assays specific to the peptides contained in the vaccine.

[0044] Figure 12

[0045] Clinical responses to PolyPEPI1018 treatment. A: Pooled distribution of clinical responses in the OBERTO trial (NCT03391232). B: Progression-free survival (PFS) and AGP association counts. C: Tumor volume and AGP association counts.

[0046] Figure 13

[0047] Probability of vaccine antigen expression in the tumor cells of Patient A. The probability of 5 out of 13 target antigens in the vaccine regimen being expressed in the patient's tumor exceeded 95%. Therefore, the 13 peptide vaccines together can induce an immune response against at least 5 ovarian cancer antigens (AGP95) with a probability of 95%. The probability of each peptide inducing an immune response in Patient A was 84%. AGP50 is the mean (expected value) = 7.9 (it is a measure of the effectiveness of the vaccine in attacking Patient A's tumor).

[0048] Figure 14

[0049] Treatment plan for Patient A.

[0050] Figure 15

[0051] T cell responses in Patient A. Left: Vaccine peptide-specific T cell responses (20mer). Right: CD8+ cytotoxic T cell responses (9mer). Predicted T cell responses were confirmed by bioassay.

[0052] Figure 16

[0053] MRI findings of Patient A treated with personalized (PIT) vaccine. This patient with advanced, highly pretreated ovarian cancer had an unexpected objective response after PIT vaccine treatment. These MRI findings indicate that the combination of PIT vaccine and chemotherapy can significantly reduce the tumor burden.

[0054] Figure 17

[0055] Probability of vaccine antigen expression in tumor cells of Patient B and treatment plan for Patient B. A: The probability of expression of 4 out of 13 target antigens in the vaccine in the patient's tumor exceeded 95%. B: Therefore, 12 peptide vaccines can together induce an immune response against at least 4 breast cancer antigens (AGP95) with a probability of 95%. The probability of each peptide inducing an immune response in Patient B was 84%. AGP50 = 6.45; it is a measure of the effectiveness of the vaccine against the patient's A tumor. C: Treatment plan for Patient B.

[0056] Figure 18

[0057] T cell responses in Patient A. Left: T cell responses P specific for vaccine peptides (20mer). Right: Kinetics of vaccine-specific CD8+ cytotoxic T cell responses (9mer). Predicted T cell responses were confirmed by bioassay.

[0058] Figure 19

[0059] Treatment plan for Patient C.

[0060] Figure 20

[0061] T cell responses in Patient C. A: Vaccine peptide-specific T cell responses (20mer). B: Vaccine peptide-specific CD8+ T cell responses (9mer). C–D: Kinetics of vaccine-specific CD4+ T cells and CD8+ cytotoxic T cell responses (9mer), respectively. Long-term immune responses specific for CD4 and CD8 T cells were present after 14 months.

[0062] Figure 21

[0063] Treatment plan for Patient D.

[0064] Figure 22

[0065] Immune response of Patient D for PIT treatment. A: CD4+ specific T cell response (20mer) and B: CD8+ T cell specific T cell response (9mer). 0.5 to 4 months refers to the time span after the last vaccination until PBMC sample collection.

[0066] Sequence description

[0067] SEQ ID No: 1 to 13 list the sequences of the personalized vaccines for Patient A and are described in Table 23.

[0068] SEQ ID No: 14 to 25 list the sequences of the personalized vaccines for Patient B and are described in Table 25.

[0069] SEQ ID No: 26 lists the 30 - amino - acid CRC_P3 peptide, Figure 11 . Detailed implementation

[0070] HLA genotype

[0071] HLA is encoded by most polymorphic genes of the human genome. Each person has maternal and paternal alleles of three HLA class I molecules (HLA - A*, HLA - B*, HLA - C*) and four HLA class II molecules (HLA - DP*, HLA - DQ*, HLA - DRB1*, HLA - DRB3* / 4* / 5*). In fact, each person expresses different combinations of 6 HLA class I molecules and 8 HLA class II molecules, which present different epitopes from the same protein antigen.

[0072] The nomenclature for representing the amino acid sequences of HLA molecules is as follows: gene name*allele: protein number. For example, it can look like: HLA - A*02:25. In this instance, "02" refers to the allele. In most cases, alleles are defined by serotype C, which means that the proteins of a given allele do not react with each other in serological assays. The protein number (25 in the above example) is assigned consecutively when the protein is discovered. A new protein number is assigned for any protein with a different amino acid sequence (for example, even a change in one amino acid in the sequence is considered a different protein number). Further information about the nucleic acid sequence of a given locus can be appended to the HLA nomenclature, but such information is not required for the methods described herein.

[0073] An individual's HLA class I genotype or HLA class II genotype can refer to the actual amino acid sequence of each class I or class II HLA of the individual, or can refer to the nomenclature as described above, which minimally specifies the alleles and protein numbers of each HLA gene. In some embodiments, an individual's HLA genotype is obtained or determined by assaying a biological sample from the individual. Biological samples typically contain the subject's DNA. The biological sample can be, for example, a blood, serum, plasma, saliva, urine, breath, cell, or tissue sample. In some embodiments, the biological sample is a saliva sample. In some embodiments, the biological sample is an oral swab sample. Any suitable method can be used to obtain or determine the HLA genotype. For example, methods and protocols known in the art can be used to determine the sequence by sequencing the HLA locus. In some embodiments, sequence-specific primer (SSP) technology is used to determine the HLA genotype. In some embodiments, sequence-specific oligonucleotide (SSO) technology is used to determine the HLA genotype. In some embodiments, sequence-based typing (SBT) technology is used to determine the HLA genotype. In some embodiments, next-generation sequencing is used to determine the HLA genotype. Alternatively, an individual's HLA profile can be stored in a database and accessed using methods known in the art.

[0074] HLA epitope binding

[0075] The HLA of a given subject will present only a limited number of different peptides generated by processing protein antigens in APCs to T cells. As used herein, when used in connection with HLA, "display" or "present" refers to the binding between a peptide (epitope) and HLA. In this regard, "displaying" or "presenting" a peptide is synonymous with "binding" a peptide.

[0076] As used herein, the term "epitope" or "T cell epitope" refers to a continuous amino acid sequence contained within a protein antigen that has a binding affinity for (is capable of binding) one or more HLAs. Epitopes are HLA- and antigen-specific (HLA-epitope pairs, predicted using known methods), but not subject-specific.

[0077] As used herein, the term "personal epitope" or "PEPI" differentiates subject-specific epitopes from HLA-specific epitopes. A "PEPI" is a polypeptide fragment consisting of a contiguous amino acid sequence of a polypeptide that is a T cell epitope capable of binding to one or more HLA class I molecules of a specific human subject. In other words, a "PEPI" is a T cell epitope recognized by the HLA class I set of a particular individual. In contrast to an "epitope", a PEPI is individual-specific because different individuals have different HLA molecules that bind different T cell epitopes. Where appropriate, a PEPI can also refer to a polypeptide fragment consisting of a contiguous amino acid sequence of a polypeptide that is a T cell epitope capable of binding to one or more HLA class II molecules of a specific human subject.

[0078] As used herein, "PEPI1" refers to a peptide or polypeptide fragment that can bind to one HLA class I molecule (or, in certain cases, HLA class II molecule) of an individual. "PEPI1+" refers to a peptide or polypeptide fragment that can bind to one or more HLA class I molecules of an individual.

[0079] "PEPI2" refers to a peptide or polypeptide fragment that can bind to two HLA class I (or class II) molecules of an individual. "PEPI2+" refers to a peptide or polypeptide fragment that can bind to two or more HLA class I (or class II) molecules of an individual, i.e., a fragment identified by the methods of the present disclosure.

[0080] "PEPI3" refers to a peptide or polypeptide fragment that can bind to three HLA class I (or class II) molecules of an individual. "PEPI3+" refers to a peptide or polypeptide fragment that can bind to three or more HLA class I (or class II) molecules of an individual.

[0081] "PEPI4" refers to a peptide or polypeptide fragment that can bind to four HLA class I (or class II) molecules of an individual. "PEPI4+" refers to a peptide or polypeptide fragment that can bind to four or more HLA class I (or class II) molecules of an individual.

[0082] "PEPI5" refers to a peptide or polypeptide fragment that can bind to five HLA class I (or class II) molecules of an individual. "PEPI5+" refers to a peptide or polypeptide fragment that can bind to five or more HLA class I (or class II) molecules of an individual.

[0083] "PEPI6" refers to a peptide or polypeptide fragment that can bind to all six HLA class I (or six HLA class II) molecules of an individual.

[0084] Generally, the length of the epitopes presented by HLA class I molecules is about 9 amino acids. However, for the purposes of the present disclosure, the length of the epitope can be more or less than 9 amino acids as long as the epitope can bind to HLA. For example, the length of an epitope that can be presented (bound to) by one or more HLA class I molecules can be between 7, or 8, or 9 and 9, or 10, or 11 amino acids.

[0085] Generally, the length of the epitopes presented by HLA class I molecules is about 9 amino acids. However, for the purposes of the present disclosure, the length of the epitope can be more or less than 9 amino acids as long as the epitope can bind to HLA. For example, the length of an epitope that can be presented (bound to) by one or more HLA class I molecules can be between 7, or 8, or 9 and 9, or 10, or 11 amino acids.

[0086] Using techniques known in the art, epitopes that will bind to a known HLA can be determined. Any suitable method can be used, provided that the same method is used to determine multiple HLA epitope binding pairs for direct comparison. For example, biochemical assays can be used. Lists of epitopes known to be bound by a given HLA can also be used. Predictive or modeling software can also be used to determine which epitopes can be bound by a given HLA. Examples are provided in Table 1. In some cases, a T cell epitope is capable of binding to a given HLA if the IC50 or predicted IC50 of the T cell epitope is less than 5000 nM, less than 2000 nM, less than 1000 nM or less than 500 nM.

[0087] Table 1 Example software for determining epitope-HLA binding

[0088]

[0089] HLA molecules regulate T cell responses. Until recently, the triggering of an immune response to an individual epitope was thought to be determined by the recognition of the epitope by the product of a single HLA allele, i.e., the HLA-restricted epitope. However, HLA-restricted epitopes induce T cell responses in only a subset of individuals. Peptides that activate T cell responses in one individual are inactive in other individuals despite HLA allele matching. Thus, it was previously unknown how individual HLA molecules present antigen-derived epitopes that are positively activating T cell responses.

[0090] The inventors have found that multiple HLAs expressed by an individual need to present the same peptide to trigger a T cell response. Thus, fragments of polypeptide antigens (epitopes) that are immunogenic (PEPI) for a particular individual are those that can bind to multiple class I (activated cytotoxic T cells) or class II (activated helper T cells) HLAs expressed by that individual. This finding is described in PCT / EP2018 / 055231, PCT / EP2018 / 055232, PCT / EP2018 / 055230, EP 3370065 and EP 3369431.

[0091] As used herein, "HLA triplet" or "HLAT" or "any combination HLAT" is any combination of three of the six HLA class I alleles expressed by a human subject. An HLAT is capable of binding a specific PEPI if all three HLA alleles of the triplet are capable of binding the PEPI. "HLAT number" is the total number of HLATs, consisting of any combination of three HLA alleles of a subject, which are capable of binding one or more defined polypeptides or polypeptide fragments, such as one or more antigens or PEPI. For example, if three of the six HLA class I alleles of a subject are capable of binding a particular PEPI, the HLAT number is one. If four of the six HLA class I alleles of a subject are capable of binding a particular PEPI, the HLAT number is four (the four combinations of any three of the four binding HLA alleles). If five of the six HLA class I alleles of a subject are capable of binding a particular PEPI, the HLAT number is ten (the ten combinations of any three of the five binding HLA alleles). If three of the six HLA class I alleles of a subject are capable of binding a first PEPI in a polypeptide, and the same or a different combination of three of the six HLA class I alleles of the subject is capable of binding a second PEPI in the polypeptide, the HLAT number is two, and so on.

[0092] Some subjects may have two HLA alleles encoding the same HLA molecule (e.g., two copies of HLA-A*02:25 in the case of homozygosity). The HLA molecules encoded by these alleles bind all the same T cell epitopes. For the purposes of this disclosure, HLAs encoded by different alleles are different HLAs, even if the two alleles are identical. "In other words", "binding to at least three HLA molecules of a subject" etc. can alternatively be expressed as "binding to HLA molecules encoded by at least three HLA alleles of a subject".

[0093] Determining cancer risk

[0094] The present disclosure provides methods for determining a subject's risk of developing cancer based on the subject's HLA class I genotype and its ability to recognize tumor-associated antigens. Due to the way HLA-T regulates T cell responses, a subject's class I HLA genotype can represent an inherent genetic cancer risk determinant: some subjects who inherit certain HLA genes from their parents can mount a broad T cell response that effectively kills tumor cells; others with HLA genes that can recognize only a few tumor antigens have a poorer defense against tumor cells. Based on 6 inherited HLA alleles, parents and offspring have different sets of HLA alleles. Since HLA-T-binding PEPI induces a T cell response in a subject, a parent's tumor-specific T cell response is not directly inherited by the offspring.

[0095] According to the present disclosure, the amino acid sequence of a T cell epitope (PEPI) present in a TAA that can bind to a subject's HLA-T indicates that the expression of the TAA in the subject will trigger a T cell response. The greater the number of HLA-Ts that can bind to the TAA epitope, the more effective the subject's T cell response to the TAA expression, and the more effectively the subject will kill cancer cells expressing the TAA. Conversely, the lower the number of HLA-Ts that can bind to the TAA epitope, the less effective the subject's T cell response to the TAA expression, and the less effective the subject is at killing cancer cells expressing the TAA. A tumor appears in a subject only when cancer cells expressing the TAA are not detected and killed by the subject's immune response. Thus, the HLA genotype can represent a genetic risk or protective factor for cancer development in a subject. The greater the number of HLA-Ts that can bind to the T cell epitope of a TAA, the lower the corresponding risk for the subject of developing a tumor (cancer) expressing the TAA. The lower the number of HLA-Ts that can bind to the T cell epitope of a TAA, the higher the risk for the subject of developing a tumor (cancer) that will express the TAA.

[0096] In some cases, the cancer is a cancer of a specific type of cancer or tissue of a specific cell type. In some cases, the cancer is a solid tumor. In some cases, the cancer is carcinoma, sarcoma, lymphoma, leukemia, germ cell tumor or blastoma. The cancer may be hormone-related or -dependent cancer (e.g., estrogen- or androgen-related cancer) or non-hormone-related or -dependent cancer. The tumor can be malignant or benign. The cancer can be metastatic or non-metastatic. The cancer can be related or not related to viral infection or viral oncogenes. In some cases, the cancer is one or more selected from melanoma, lung cancer, renal cell carcinoma, colorectal cancer, bladder cancer, glioma, head and neck cancer, ovarian cancer, non-melanoma skin cancer, prostate cancer, kidney cancer, gastric cancer, liver cancer, cervical cancer, esophageal cancer, non-Hodgkin lymphoma, leukemia, pancreatic cancer, endometrial cancer, lip cancer, oral cancer, thyroid cancer, brain cancer, nervous system cancer, gallbladder cancer, laryngeal cancer, pharyngeal cancer, myeloma, nasopharyngeal cancer, Hodgkin lymphoma, testicular cancer, breast cancer, gastric cancer, bladder cancer, colorectal cancer, renal cell carcinoma, liver cancer, pediatric cancer and Kaposi sarcoma.

[0097] In other cases, the method can be used to determine the risk that a subject will develop any cancer or any combination of cancers disclosed herein.

[0098] In other cases, the method can be used to determine the risk that a subject will develop a cancer that expresses one or more specific TAAs. Suitable TAAs can be selected for use in the methods disclosed herein, as further described below.

[0099] The terms "T cell response" and "immune response" are used interchangeably herein and refer to the activation of T cells and / or the induction of one or more effector functions following the identification of one or more HLA epitope-binding pairs. In some cases, the "immune response" includes an antibody response, since HLA class II molecules stimulate helper responses that are involved in the induction of durable CTL responses and antibody responses. Effector functions include cytotoxicity, cytokine production and proliferation.

[0100] The methods of the present disclosure can be used to determine the immunological risk of developing cancer. Specifically, the methods described herein can be used to determine a subject's ability to recognize and initiate an immune response against a TAA or cancer cells that express those TAAs. Many other factors may contribute to the overall risk of a subject developing cancer. Thus, in some cases, the methods disclosed herein can be combined with or incorporated into broader models for cancer risk prediction that include other risk determinants. For example, in some embodiments, the methods of the present disclosure further include determining other cancer risk factors, such as environmental factors, lifestyle factors, other genetic risk factors, and any other factors that contribute to the overall risk of a subject developing cancer.

[0101] Not all HLATs of all the subjects and / or not all TAAs may play equally important roles in the immunological control of cancer. Thus, in some cases according to the present invention, different weights can be applied to different HLA alleles (e.g., using the “HLA score”-based method described in Examples 7 to 9 herein), different HLATs, and / or HLATs of T cell epitopes capable of binding different TAAs (e.g., using the “HLAT score”-based method described in Examples 5 and 6 herein). Examples The “HLAT score”- and “HLA score”-based methods of the present invention are technically different, but in both cases, if a subject has a better predictive ability to generate an immune response against TSA, then he / she has a greater score. Both methods use statistical learning algorithms. In the case of the HLAT score, the learning algorithm assigns weights to TSAs based on the importance of the immune response against the TSA in combating certain cancers. Then, the final HLAT score is the weighted sum of the HLA triplets that the subject can generate against the TSA. In the case of the HLA score, the learning algorithm assigns scores to individual HLA alleles based on the degree to which HLAT can be generated against the TSA in subjects having the HLA allele. Then, the final HLA score of the subject is the sum of the weights of the HLA alleles he / she has.

[0102] In some cases, the weights to be applied can be determined empirically. For example, in some cases, the method described herein can be used to determine the weight of the T cell epitope of HLAT capable of binding a specific TAA, based on or associated with the ability to separate subjects with (said) cancer from subjects without (said) cancer or a background population including subjects with (said) cancer, for each TAA independently.

[0103] Alternatively or additionally, the weight of the T cell epitope of HLAT capable of binding a specific TAA can be determined by, based on, or associated with the frequency of TAA expression in the cancer or cancer type. The expression frequency of TAAs in different cancers can be determined from published figures and scientific publications.

[0104] In some cases, the weight applied to a specific HLAT can be determined by, based on, or associated with the frequency of the presence of the HLAT in a subset of subjects with cancer or subjects with cancer and / or disease-matched subjects.

[0105] In some cases, the weight of the T cell epitope of HLAT capable of binding to each TAA is defined as or uses the following weight (w(c)):

[0106]

[0107] where t(c) represents the p-value of a one-sided t-test of the HLA T scores of TAA c for groups with and without cancer, and B is the Bonferroni correction (the number of TAAs). This weighting is used in the HLA T score-based method described herein.

[0108] In some cases, the significance score (weight) of an HLA allele (h) is defined as

[0109]

[0110] where u(h) is the p-value of a two-sided u-test that determines whether the number of HLA Ts in two individual subgroups is different: individuals in one subgroup have HLA h and individuals in one subgroup do not have HLA h. B is the Bonferroni correction, and sign(h) is +1 if the average number of HLA Ts in the subgroup with the h allele is greater than the average number of HLA Ts in the subgroup without h, and otherwise -1. This weighting is used in the HLA score-based method described herein.

[0111] In some cases, any suitable method known to those skilled in the art can be used to further optimize the initial weighting. In some cases, the sum of these significance scores is used to determine the risk that a subject will develop cancer in relation to the risk that a subject will develop cancer.

[0112] For example, in some cases, the risk that a subject will develop cancer in relation to the risk that a subject will develop cancer or the risk that a subject will develop cancer is determined using the following HLA T score (s(x)):

[0113]

[0114] where C is the set of TAAs, c is a specific TAA, w(c) is the weight of TAA c, and p(x,c) is the number of HLA Ts of TAA c in subject x.

[0115] The HLA T score-based method and the HLA score-based method described in the examples herein are two examples of methods according to the present invention. By using HLA class I genotype data of an individual, further scoring schemes can be developed. The specific scores to be used depend on the indication and prior data. In some cases, the choice will be made based on the performance of different calculations on the available test dataset. Performance can be evaluated by the AUC value (area under the ROC curve) or by any other performance goodness score known to those skilled in the art.

[0116] Tumor-associated antigen (TAA)

[0117] Cancer or tumor-associated antigens (TAAs) are proteins expressed in cancer or tumor cells. Examples of TAAs include neoantigens (neoantigens, which are expressed during tumorigenesis and altered from similar proteins in normal or healthy cells), products of oncogenes and tumor suppressor genes, overexpressed or abnormally expressed cellular proteins (e.g., HER2, MUC1), antigens produced by oncogenic viruses (e.g., EBV, HPV, HCV, HBV, HTLV), cancer-testis antigens (CTAs, e.g., MAGE family, NY-ESO), cell type-specific differentiation antigens (e.g., MART-1), and tumor-specific antigens (TSAs). TSAs are antigens produced by a specific type of tumor and do not appear on normal cells of the tissue in which the tumor develops. TSAs include shared antigens, neoantigens, and unique antigens. TAA sequences can be found through experiments, or published scientific papers, or publicly available databases, such as the Ludwig Institute for Cancer Research (www.cta.lncc.br / ), the Cancer Immunity Database (cancerimmunity.org / peptide / ), and the TANTIGEN Tumor T Cell Antigen Database (cvc.dfci.harvard.edu / tadb / ). Exemplary TAAs are listed in Tables 2 and 11.

[0118]

[0119]

[0120]

[0121]

[0122]

[0123] Table 2 optionally excludes Ropporin-1A Q9HAT0 and / or WBP2NL Q6ICG8.1.

[0124] In some cases, the methods described herein are used to determine the risk that a subject will develop cancer expressing one or more specific TAAs. In other cases, the method is used to determine the risk that a subject will develop any cancer or a specific type of cancer. In some cases, different TAAs may be associated with different types of cancer, but not every specific type of cancer expresses the same combination of TAAs. Thus, in some cases, binding epitope HLA-T is quantified among multiple TAAs, in some cases, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 30, 35, 40, 45 or more TAAs. Generally, if a TAA is expressed in a higher proportion of cancers or cancer patients or cancers of the selected type, fewer TAAs can be used. If a TAA is expressed in a lower proportion of cancers or cancer patients or cancers of the selected type, more TAAs can be used. In some cases, a set of TAAs can be used that together are expressed or overexpressed in a minimum proportion of cancers, cancer patients or cancers of the selected type, such as 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 98% or more. The expression frequency of TAAs in different cancers can be determined from publicly available figures and scientific publications.

[0125] TAAs selected for use according to the present disclosure are generally TAAs that are expressed or overexpressed in a high proportion of cancers or a specific type of cancer. In some cases, one or more or each TAA can be expressed or overexpressed in at least 1%, 2%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% of cancers or in cancers of a disease- and / or subject-matched population. For example, a subject can be matched according to sex, age, disease type or stage, genotype, expression of one or more biomarkers, etc. or any combination thereof.

[0126] In some cases, one or more or each TAA is a tumor-specific antigen (TSA) or a cancer-testis antigen (CTA). CTA is generally not expressed outside of embryonic development of healthy cells. In healthy adults, CTA expression is limited to male germ cells that do not express HLA and cannot present antigens to T cells. Thus, when expressed in cancer cells, CTA is considered a neoantigen that is expressed. Expression of CTA (i) is specific to tumor cells, (ii) is more frequent in metastases than in primary tumors, and (iii) is conserved in metastatic tumors of the same patient (Gajewski ed. Targeted Therapeutics in Melanoma. Springer New York. 2012).

[0127] In some cases, the method includes the step of selecting and / or identifying a suitable TAA or suitable group of TAAs for use in the methods disclosed herein.

[0128] Methods of treatment

[0129] In some cases, the methods described herein include selecting, preparing, and / or administering a treatment for cancer in a subject. Using the methods described herein, the subject may have been determined to have an elevated risk of developing cancer. As used herein, "treatment" is any action taken to prevent or delay the onset of cancer, to improve one or more symptoms or complications, to induce or prolong remission, to delay recurrence, relapse, or progression, or otherwise to improve or stabilize the subject's disease state or cancer risk. Typically, treatment is prophylactic treatment, intended to delay or prevent the onset of cancer or any symptoms or complications associated with cancer. Treatment can be immunotherapy or vaccination.

[0130] As used herein, the term "treatment" can in some cases include recommendations regarding a subject's behavior, environmental exposures, or lifestyle, which are intended to reduce the risk that the subject will develop cancer or any symptoms or complications associated with cancer. For example, for a subject determined to have an elevated risk of developing melanoma, treatment can include recommending that the subject reduce exposure to UV radiation. For example, this can include avoiding artificial UV sources, reducing sunlight exposure at certain times of the day or avoiding sunlight exposure, using sunscreen that provides adequate protection, wearing protective clothing, avoiding sunburn, and / or taking vitamin D. Methods of treatment may include recommendations related to diet, including use of dietary supplements (such as antioxidant supplements or increasing calcium intake), drug use (including reducing tobacco and / or alcohol intake), exercise, or exposure to potential carcinogens, infectious agents, and / or radiation.

[0131] In other cases, treatment can include additional or increased frequency of screening or examinations to achieve early diagnosis of cancer. In other cases, treatment can include administering anti-inflammatory drugs, such as aspirin or non-steroidal anti-inflammatory drugs, or avoiding or reducing the administration of immunosuppressive drugs. In some cases, treatment can include increased attention to the management of other conditions that are potential risk factors, such as obesity, or conditions associated with chronic inflammation, such as ulcerative colitis and Crohn's disease.

[0132] In other cases, treatment can be any known cancer treatment or prophylactic treatment, such as surgery, chemotherapy, cytotoxic or non-cytotoxic chemotherapy, radiotherapy, targeted therapy, hormone therapy, or administration of a targeted small molecule drug or antibody, such as a monoclonal antibody or co-stimulatory antibody, and includes any cancer treatment described herein.

[0133] Therapies aimed at enhancing the immune response of a subject against cancer cells may be particularly effective in preventing or delaying cancer development in subjects determined to have an elevated cancer risk using the methods described herein. Thus, in some cases, the therapy can be immunotherapy or checkpoint blockade therapy or checkpoint inhibitor therapy. In some cases, the method comprises administering to the subject one or more peptides or one or more polynucleic acids or a vector encoding one or more peptides as described below, the peptide or vector comprising an amino acid sequence that is (i) a fragment of an antigen associated with expression in cancer; and (ii) a T cell epitope capable of binding to the subject's HLA-T.

[0134] Personalized treatment method

[0135] According to the present invention, the ability of a subject's HLA-T to recognize a TAA predicts the risk of cancer in the subject. Thus, by stimulating the immune response of the subject with a peptide corresponding to the TAA epitope recognized by the subject's HLA-T, the risk of cancer in the subject can be reduced.

[0136] Thus, in some cases, the present disclosure relates to a method for prophylactic treatment of cancer, wherein the method comprises administering to the subject one or more peptides, or one or more polynucleic acids or a vector encoding one or more peptides, the peptide comprising an amino acid sequence that is (i) a fragment of a TAA; and (ii) a T cell epitope capable of binding to the subject's HLA-T (i.e., PEPI3+). In some cases, it has been determined using the methods described herein that the subject is at an elevated risk of developing cancer.

[0137] One or more suitable TAAs and suitable epitopes in the TAAs that bind to the subject's HLA-T can be selected as described herein. In certain cases, the method can include the step of identifying and / or selecting suitable TAAs, epitopes, and / or peptides. Generally, one or more TAAs are TAAs that are frequently expressed in cancer cells.

[0138] In some cases, it has been determined that the subject is at an elevated risk of developing cancer in which cancer cells express a specific TAA. This may be the case if the TAA contains a small number of epitopes that are PEPI3+ for a particular subject, or the epitopes of the TAA are recognized by a small number of the subject's HLA-Ts. Treatment of the subject can include administering a peptide comprising an amino acid sequence that is (i) a fragment of the TAA and (ii) comprises a T cell epitope capable of binding to one or more of the subject's HLA-Ts.

[0139] In other cases, it is determined that the subject is at an elevated risk of developing one or more specific types of cancer, such as any type of cancer disclosed herein. Treatment of the subject can include administering a peptide comprising an amino acid sequence that (i) is a fragment of a TAA associated with expression in that cancer type and (ii) contains a T cell epitope capable of binding one or more HLAs of the subject.

[0140] In some cases, the TAA is a TAA recognized by a minority of the subject's HLAs. Such treatment will enhance the T cell response to the TAA. In other cases, the TAA can be a TAA recognized by multiple HLAs. The subject will generally already be able to mount a broad T cell response against such a TAA. This may be particularly helpful in killing cancer cells that often co-express the target TAA with other TAAs that may not be well recognized by the subject's HLAs.

[0141] The peptide can be engineered or non-naturally occurring. The length of the fragment and / or peptide can be up to 50, 45, 40, 35, 30, 25, 20, 15, 14, 13, 12, 11, 10, or 9 amino acids. Generally, the length of the peptide can be 15 or 20 to 30 or 35 amino acids. In some cases, the amino acid sequence corresponding to the fragment of the TAA is flanked at the N and / or C terminus by other amino acids that are part of a non-TAA contiguous sequence. In some cases, the sequence is flanked at the N and / or C terminus by up to 41, or 35, or 30, or 25, or 20, or 15, or 10, or 9, or 8, or 7, or 6, or 5, or 4, or 3, or 2, or 1 additional amino acids. In other cases, each peptide can consist of a fragment of a TAA, or of two or more such fragments that are linked end-to-end (sequentially arranged end-to-end in the peptide) or that overlap within a single peptide.

[0142] In some cases, the treatment method includes administering to the subject one or more peptides, or one or more nucleic acids or vectors encoding one or more peptides, the peptides comprising at least 2, or 3, or 4, or 5, or 6, or 7, or 8, or 9, or 10, or 11, or 12, or 13, or 14, or 15, or 20, or 25, or 30, or 35, or 40, or 45, or 50 or more different T cell epitopes (PEPIs), each of the T cell epitopes (i) being contained within a fragment of a TAA and (ii) being capable of binding an HLA of the subject. In some cases, two or more PEPIs are contained within fragments of at least 2, or 3, or 4, or 5, or 6, or 7, or 8, or 9, or 10, or 11, or 12 or more different TAAs. In some cases, one or more or each TAA is a TSA and / or a CTA.

[0143] In some cases, one or more of the peptide fragments comprise an amino acid sequence that is a T cell epitope capable of binding to at least three or at least four HLA class II alleles of a subject. Such treatment can elicit a CD8+ T cell response and a CD4+ T cell response in the treated subject.

[0144] In some cases, the treatment method includes administering to a subject any one or more of the peptides, or one or more nucleic acids or vectors encoding one or more of the peptides, or administering any pharmaceutical composition as described in any one of PCT / EP2018 / 055231, PCT / EP2018 / 055232, PCT / EP2018 / 055230, EP 3370065, and EP 3369431. In some specific cases, the treatment is for preventing breast cancer, ovarian cancer, or colorectal cancer, and includes administering the composition described in PCT / EP2018 / 055230 and / or EP 3369431.

[0145] As used herein, the term "polypeptide" refers to a full-length protein, a part of a protein, or a peptide characterized as a string of amino acids. The term "peptide" refers to a short polypeptide. As used herein, the term "fragment" or "fragment of a polypeptide" refers to a string of amino acids or an amino acid sequence that generally has a reduced length relative to the above or a reference polypeptide and contains the same amino acid sequence as the reference polypeptide in a common portion. In appropriate cases, such fragments according to the present disclosure can be included in a larger polypeptide of which they are a component. In some cases, the fragment can comprise the full length of the polypeptide, for example, the entire polypeptide, for example, a 9-amino acid peptide, which is a single T cell epitope. In some cases, the length of the peptide or polypeptide fragment can be between 7, or 8, or 9, or 10, or 11, or 12, or 13, or 14, or 15 and 10, or 11, or 12, or 13, or 14, or 15, or 20, or 25, or 30, or 35, or 40, or 45, or 50 amino acids.

[0146] Pharmaceutical Compositions and Modes of Administration

[0147] In some cases, the present disclosure relates to a treatment method that includes administering to a subject one or more of the peptides as described herein. The one or more peptides can be administered to the subject together or sequentially. For example, the treatment can include administering a number of peptides over a period of time, for example, up to one year. In some cases, the treatment cycle can also be repeated to enhance the immune response.

[0148] In addition to one or more peptides, a pharmaceutical composition for administration to a subject can comprise a pharmaceutically acceptable excipient, carrier, diluent, buffer, stabilizer, preservative, adjuvant, or other materials known to those of skill in the art. Such substances are preferably non-toxic and preferably do not interfere with the pharmaceutical activity of the active ingredient. The pharmaceutical carrier or diluent can be, for example, an aqueous solution. The exact nature of the carrier or other substance can depend on the route of administration, such as oral, intravenous, dermal or subcutaneous, intranasal, intramuscular, intradermal, and intraperitoneal routes.

[0149] To increase the immunogenicity of the composition, the pharmaceutical composition can comprise one or more adjuvants and / or cytokines.

[0150] Suitable adjuvants include aluminum salts such as aluminum hydroxide or aluminum phosphate, but can also be salts of calcium, iron or zinc, or can be an insoluble suspension of acylated tyrosine or acylated sugars, or can be cationically or anionic - derived sugars, polyphosphazenes, biodegradable microspheres, monophosphoryl lipid A (MPL), lipid A derivatives (e.g., reduced toxicity), 3 - O - deacylated monophosphoryl lipid A (3D - MPL), Quil A, saponins, QS21, Freund's incomplete adjuvant (Difco Laboratories, Detroit, Michigan), Merck adjuvant 65 (Merck & Co., Inc., Rahway, New Jersey), AS - 2 (SmithKline Beecham, Philadelphia, Pennsylvania), CpG oligonucleotides, bioadhesives and mucoadhesives, microparticles, liposomes, polyoxyethylene ether formulations, polyoxyethylene ester formulations, muramyl peptides or imidazoquinolone compounds (e.g., imiquimod and its homologs). Human immunomodulators suitable for use as adjuvants in the present disclosure include cytokines, such as interleukins (e.g., IL - 1, IL - 2, IL - 4, IL - 5, IL - 6, IL - 7, IL - 12, etc.), macrophage colony - stimulating factor (M - CSF), tumor necrosis factor (TNF), granulocyte - macrophage colony - stimulating factor (GM - CSF) can also be used as adjuvants.

[0151] In some embodiments, the composition comprises an adjuvant selected from the group consisting of Montanide ISA-51 (Seppic, Fairfield, NJ, USA), QS-21 (Aquila Biopharmaceuticals, Lexington, MA, USA), granulocyte-macrophage colony stimulating factor (GM-CSF), cyclophosphamide, Bacillus Calmette-Guerin (BCG), corynbacterium parvum, levamisole, azimezone, isopyrazam, dinitrochlorobenzene (DNCB), keyhole limpet hemocyanin (KLH), Freund's adjuvant (complete and incomplete), mineral gels, aluminum hydroxide, lysophosphatidylcholine, pluronic polyols, polyanions, oil emulsions, dinitrophenol, diphtheria toxin (DT).

[0152] Examples of suitable compositions and methods of administration of polypeptide fragments are provided in Esseku & Adeye (2011) and Van den Mooter G (2006). The preparation of vaccine and immunotherapy compositions is generally described in "Vaccine Design" ("The subunit and adjuvant approach" (eds Powell M.F. & Newman M.J. (1995) Plenum Press New York)). Encapsulation within liposomes, as described by Fullerton in U.S. Patent 4,235,877, is also contemplated.

[0153] The method of treatment may include administering to a subject a pharmaceutical composition comprising one or more peptides as described herein as an active ingredient. The term "active ingredient" as used herein refers to a peptide that is intended to induce an immune response in a subject to whom the pharmaceutical composition can be administered. In some cases, the active ingredient peptide may be a peptide product of a vaccine or immunotherapy composition that is produced in vivo after administration to the subject. For DNA or RNA immunotherapy compositions, the peptide may be produced in vivo by the cells of the subject to whom the composition is administered. For cell-based compositions, the polypeptide may be processed and / or presented by the cells of the composition, such as autologous dendritic cells or antigen-presenting cells pulsed with the polypeptide or containing an expression construct encoding the polypeptide.

[0154] In some embodiments, the compositions disclosed herein can be formulated as nucleic acid vaccines. In some embodiments, the nucleic acid vaccine is a DNA vaccine. In some embodiments, the DNA vaccine or gene vaccine comprises a plasmid having a promoter and appropriate transcriptional and translational control elements, and a nucleic acid sequence encoding one or more polypeptides of the present disclosure. In some embodiments, the plasmid further includes sequences that enhance, for example, expression levels, intracellular targeting, or proteasomal processing. In some embodiments, the DNA vaccine comprises a viral vector containing a nucleic acid sequence encoding one or more polypeptides of the present disclosure. In a further aspect, the compositions disclosed herein comprise one or more nucleic acids encoding peptides that have been determined to be immunoreactive with a biological sample. For example, in some embodiments, the composition comprises one or more nucleotide sequences encoding 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or more peptides, the peptides comprising fragments that are T cell epitopes capable of binding to at least three HLA class I molecules of a patient. In some embodiments, the DNA or gene vaccine also encodes an immunomodulatory molecule to manipulate the immune response generated, such as enhancing vaccine potency, stimulating the immune system, or reducing immunosuppression. Strategies for enhancing the immunogenicity of DNA or gene vaccines include encoding of xenoantigens, fusion of antigens with T cell activating or co-stimulatory molecules, priming with a DNA vector and then boosting with a viral vector, and utilization of immunomodulatory molecules. In some embodiments, the DNA vaccine is introduced by needle, gene gun, aerosol syringe, patch, microneedle, abrasion, and other forms. In some forms, the DNA vaccine is incorporated into liposomes or other forms of nanobodies. In some embodiments, the DNA vaccine includes a delivery system selected from the group consisting of transfection agents, protamine, protamine liposomes, polysaccharide particles, cationic nanoemulsions, cationic polymers, cationic polymer liposomes, cationic nanoparticles, cationic lipids and cholesterol nanoparticles, cationic lipids, cholesterol and PEG nanoparticles, dendrimer nanoparticles. In some embodiments, the DNA vaccine is administered by inhalation or ingestion. In some embodiments, the DNA vaccine is introduced into the blood, thymus, pancreas, skin, muscle, tumor, or other sites.

[0155] In some embodiments, the compositions disclosed herein are prepared as RNA vaccines. In some embodiments, the RNA is non-replicating mRNA or self-amplifying RNA of viral origin. In some embodiments, the non-replicating mRNA encodes a peptide disclosed herein and contains 5′ and 3′ untranslated regions (UTRs). In some embodiments, the self-amplifying RNA of viral origin not only encodes a peptide disclosed herein, but also encodes a viral replication mechanism capable of intracellular RNA amplification and high levels of protein expression. In some embodiments, the RNA is directly introduced into an individual. In some embodiments, the RNA is chemically synthesized or transcribed in vitro. In some embodiments, mRNA is generated from a linear DNA template using T7, T3, or Sp6 phage RNA polymerase, and the resulting product contains an open reading frame encoding a peptide disclosed herein, flanking UTRs, a 5′ cap, and a poly(A) tail. In some embodiments, various forms of 5′ caps are added during or after the transcription reaction using a vaccinia virus capping enzyme or by incorporating a synthetic cap or an anti-reverse transcriptase cap analog. In some embodiments, the optimal length of the poly(A) tail is added directly from the encoding DNA template or by using a poly(A) polymerase to the mRNA. The RNA encodes one or more peptides that comprise fragments of T cell epitopes capable of binding at least three HLA class I molecules of a patient. In some embodiments, the RNA includes signals that enhance stability and translation. In some embodiments, the RNA further includes unnatural nucleotides that increase the half-life or modified nucleosides that alter the immunostimulatory profile. In some embodiments, the RNA is introduced by needles, gene guns, aerosol syringes, patches, microneedles, abrasion, and other forms. In some forms, the RNA vaccine is incorporated into liposomes or other forms of nanobodies that facilitate cellular uptake of the RNA and protect it from degradation.In some embodiments, the RNA vaccine includes a delivery system selected from the group consisting of transfection agents, protamine, protamine liposomes, polysaccharide particles, cationic nanoemulsions, cationic polymers, cationic polymer liposomes, cationic nanoparticles, cationic lipids and cholesterol nanoparticles, cationic lipids, cholesterol and PEG nanoparticles, dendrimer nanoparticles, and / or naked mRNA, in vivo electroporated naked mRNA, protamine complexed mRNA, mRNA associated with a positively charged oil-in-water cationic nanoemulsion, mRNA associated with a chemically modified dendrimer and complexed with polyethylene glycol (PEG)-lipid, protamine complexed mRNA in PEG-lipid nanoparticles, mRNA associated with a cationic polymer such as polyethyleneimine (PEI), mRNA associated with a cationic polymer such as PEI and a lipid component, mRNA associated with a polysaccharide (e.g., chitosan) particle or gel, mRNA in cationic lipid nanoparticles (e.g., 1,2-dioleoyl-3-trimethylammonium propane (DOTAP) or dioleoyl phosphatidylethanolamine (DOPE) lipids), mRNA complexed with cationic lipid and cholesterol, or mRNA complexed with cationic lipid / cholesterol and PEG-lipid. In some embodiments, the RNA vaccine is administered by inhalation or ingestion. In some embodiments, the RNA is introduced into the blood, thymus, pancreas, skin, muscle, tumor, or other sites, and / or by intradermal, intramuscular, subcutaneous, intranasal, intranodular, intravenous, intrasplenic, intratumoral, or other delivery routes.

[0156] The polynucleotide or oligonucleotide component can be a naked nucleotide sequence or combined with a cationic lipid, polymer, or targeting system. They can be delivered by any available technique. For example, the polynucleotide or oligonucleotide is introduced by needle injection, preferably intradermal, subcutaneous, or intramuscular injection. Alternatively, a delivery device such as particle-mediated gene delivery is used to directly deliver the polynucleotide or oligonucleotide through the skin. The polynucleotide or oligonucleotide can be administered locally to the skin or mucosal surface, for example, by intranasal, oral, or rectal administration.

[0157] Uptake of the polynucleotide or oligonucleotide construct can be enhanced by several known transfection techniques, such as those that include the use of transfection agents. Examples of these reagents include cationic reagents such as calcium phosphate and DEAE-dextran, and lipid transfection agents such as lipofectam and transfectam. The dose of the polynucleotide or oligonucleotide administered can be varied.

[0158] Administration is typically at a “prophylactically effective amount” or “therapeutically effective amount” (depending on the situation, although prophylaxis can be considered treatment), which is sufficient to result in a clinical response or show clinical benefit to an individual, such as preventing or delaying the onset of a disease or disorder, improving one or more symptoms, inducing or prolonging remission or delaying recurrence or relapse. In some cases, the treatment methods according to the present disclosure can be carried out in individuals with a primary cancer disease that has been cured to prevent cancer recurrence or metastasis.

[0159] The dose can be determined according to various parameters, particularly according to the substance used, the age, weight and condition of the individual to be treated, the route of administration, and the desired regimen. The amount of antigen in each dose is selected as the amount that induces an immune response. A physician is able to determine the route of administration and dose required for any particular individual. The dose can be provided as a single dose or can be provided as multiple doses, for example, taken at regular intervals, such as administering 2, 3, or 4 doses per hour. Generally, peptides, polynucleotides or oligonucleotides are typically administered in the range of 1 pg to 1 mg, more typically 1 pg to 10 μg for particle-mediated delivery, and in the range of 1 μg to 1 mg, more typically 1 μg to 100 μg, more typically 5 μg to 50 μg for other routes. Generally, each dose is expected to contain 0.01 mg to 3 mg of antigen. The optimal amount of a particular vaccine can be determined by studies involving observing the immune response in subjects.

[0160] Examples of the above techniques and regimens can be found in Remington’s Pharmaceutical Sciences, 20 th Edition, 2000, pub. Lippincott, Williams & Wilkins.

[0161] Routes of administration include, but are not limited to, intranasal, oral, subcutaneous, intradermal and intramuscular. Generally, administration is subcutaneous. Subcutaneous administration can be, for example, by injection into the abdomen, the side and front of the upper arm or thigh, the scapular region of the back or the gluteal region of the upper abdomen.

[0162] Those skilled in the art will recognize that the composition can also be administered in one or more doses and by other routes of administration. For example, these other routes include intradermal, intravenous, intra-vascular, intra-arterial, intraperitoneal, intrathecal, intratracheal, intracardiac, intralobular, intramedullary, intrapulmonary and intravaginal. Depending on the desired duration of treatment, the composition according to the present disclosure can be administered once or multiple times and can also be administered intermittently, such as monthly for several months or years and at different doses.

[0163] The treatment methods according to the present disclosure can be carried out alone or in combination with other pharmacological compositions or treatments, such as behavioral or lifestyle modifications, chemotherapy, immunotherapy, and / or vaccines. Other therapeutic compositions or treatments can be, for example, one or more of those discussed herein and can be administered simultaneously or sequentially (before or after) with the compositions or treatments of the present disclosure.

[0164] In some cases, the treatment can be administered in combination with surgery, chemotherapy, cytotoxic or non-cytotoxic chemotherapy, radiotherapy, targeted therapy, hormonal therapy or the administration of targeted small molecule drugs or antibodies (e.g., monoclonal antibodies) or co-stimulatory antibodies. Chemotherapy has been shown to sensitize tumors to killing by specifically cytotoxic T cells induced by vaccination (Ramakrishnan et al. J Clin Invest. 2010;120(4):1111-1124). Examples of chemotherapeutic agents include alkylating agents, including nitrogen mustards such as dichloromethyldiethylamine (HN2), cyclophosphamide, ifosfamide, melphalan (L-sarcolysin protease), and chlorambucil; anthracyclines; epothilones; nitrosoureas such as carmustine (BCNU), lomustine (CCNU), semustine (methyl CCNU), and streptozocin; triazenes such as dacarbazine (DTIC); dimethyltriazenoimidazole carboxamide; ethyleneimines / methylmelamines such as hexamethylmelamine, thiotepa; alkyl sulfonates such as busulfan; antimetabolites, including folic acid analogs such as methotrexate; alkylating agents, antimetabolites, pyrimidine analogs such as fluorouracil (5-fluorouracil; 5-FU), floxuridine (fluorodeoxyuridine; FUdR), and cytarabine (cytosine arabinoside); purine analogs and related inhibitors such as mercaptopurine (6-mercaptopurine; 6-MP), thioguanine (6-thioguanine; TG), and pentostatin (2′-deoxycoformycin); epipodophyllotoxins; enzymes such as L-asparaginase; biological response modifiers such as IFNα, IL-2, G-CSF, and GM-CSF; platinum coordination complexes such as cisplatin (cis-DDP), oxaliplatin, and carboplatin; anthracenediones such as mitoxantrone and anthramycin; substituted ureas such as hydroxyurea; methylhydrazine derivatives, including procarbazine (N-methylhydrazine, MIH) and methylbenzylhydrazine; adrenocortical inhibitors such as mitotane (o,p′-DDD) and aminoglutethimide; paclitaxel and analogs / derivatives; hormones / hormonal therapies and agonists / antagonists, including adrenocortical steroid antagonists such as prednisone and its equivalents, dexamethasone, and aminoglutethimide, progesterones such as hydroxyprogesterone caproate, medroxyprogesterone acetate, and megestrol acetate, estrogens such as diethylstilbestrol and ethinylestradiol equivalents, antiestrogens such as tamoxifen, androgens including testosterone propionate and fluoxymesterone / equivalents, antiandrogens such as flutamide, gonadotropin-releasing hormone analogs, and leuprolide, and non-steroidal antiandrogens such as flutamide; natural products including vinca alkaloids such as vinblastine (VLB) and vincristine, epipodophyllotoxins such as etoposide and teniposide, antibiotics such as dactinomycin (actinomycin D), daunorubicin (daunomycin; rubidomycin), doxorubicin, bleomycin, plicamycin (mithramycin), and mitomycin (mitomycin C), enzymes such as L-asparaginase, and biological response modifiers such as interferon alphenome.

[0165] System

[0166] The present disclosure provides a system. The system may include a storage module configured to store data including the HLA class I genotype and the amino acid sequence of the TAA of a subject. The system may include a computing module configured to quantify the HLA-T of a subject that can bind to the T cell epitopes in the amino acid sequence that can bind to the TAA, wherein each HLA of the HLA-T can bind to the same T cell epitope. The system may include a module for receiving at least one sample from at least one subject. The system may include an HLA genotyping module for determining the class I and / or class II HLA genotype of a subject. The storage module may be configured to store the data output from the genotyping module. The HLA genotyping module may receive a biological sample obtained from the subject and determine the class I and / or class II HLA genotype of the subject. The sample generally contains the subject's DNA. The sample may be, for example, a blood, serum, plasma, saliva, urine, exhaled breath, cell or tissue sample. The system may further include an output module configured to display an indication of the risk that the subject will develop cancer and / or the recommended treatment for the subject as described herein.

[0167] Other embodiments of the present disclosure

[0168] 1. A method for treating a human subject at risk of developing cancer, the method comprising

[0169] a. quantifying the HLA triplet (HLA-T) of the subject that can bind to the T cell epitopes in the amino acid sequence that can bind to the tumor-associated antigen (TAA), wherein each HLA of the HLA-T can bind to the same T cell epitope;

[0170] b. determining the risk that the subject will develop cancer, wherein, for the TAA, the lower the number of HLA-Ts that can bind to the T cell epitopes of the TAA, the higher the risk that the subject will develop cancer; and

[0171] c. administering to the subject a peptide comprising an amino acid sequence, or a polynucleotide or vector encoding the peptide

[0172] i. that is a fragment of the TAA; and

[0173] ii. that comprises a T cell epitope that can bind to the HLA-T of the subject.

[0174] 2. The method according to item 1, wherein the N and / or C-terminal flanks of the fragment of the TAA are other amino acids that are part of a non-TAA sequence.

[0175] 3. The method according to any one of items 1 to 2, wherein the cancer is selected from melanoma, lung cancer, renal cell carcinoma, colorectal cancer, bladder cancer, glioma, head and neck cancer, ovarian cancer, non-melanoma skin cancer, prostate cancer, kidney cancer, gastric cancer, liver cancer, cervical cancer, esophageal cancer, non-Hodgkin lymphoma, leukemia, pancreatic cancer, uterine body cancer, lip cancer, oral cancer, thyroid cancer, brain cancer, nervous system cancer, gallbladder cancer, laryngeal cancer, pharyngeal cancer, myeloma, nasopharyngeal cancer, Hodgkin lymphoma, testicular cancer, breast cancer, and Kaposi sarcoma.

[0176] 4. The method according to item 1, wherein the TAA is selected from any one listed in Table 2 or Table 11.

[0177] 5. A method for treating cancer in an individual in need thereof with cancer therapy, comprising:

[0178] determining whether the individual is at a higher risk of developing cancer by the following steps:

[0179] performing a quantitative determination on a biological sample from the individual to determine the individual's HLA triplet (HLAT), which is capable of binding a T cell epitope in the amino acid sequence of a tumor-associated antigen (TAA), wherein each HLA of the HLAT is capable of binding the same T cell epitope; and

[0180] administering the cancer therapy to the individual if the individual has a lower number of HLATs capable of binding TAA than a threshold derived from a control group of individuals.

[0181] 6. The method according to item 5, further comprising obtaining the biological sample from the individual.

[0182] 7. The method according to item 5, wherein the cancer is selected from melanoma, lung cancer, renal cell carcinoma, colorectal cancer, bladder cancer, glioma, head and neck cancer, ovarian cancer, non-melanoma skin cancer, prostate cancer, kidney cancer, gastric cancer, liver cancer, cervical cancer, esophageal cancer, non-Hodgkin lymphoma, leukemia, pancreatic cancer, uterine body cancer, lip cancer, oral cancer, thyroid cancer, brain cancer, nervous system cancer, gallbladder cancer, laryngeal cancer, pharyngeal cancer, myeloma, nasopharyngeal cancer, Hodgkin lymphoma, testicular cancer, breast cancer, and Kaposi sarcoma.

[0183] 8. The method according to item 5, wherein the cancer therapy comprises administering to the individual a peptide comprising an amino acid sequence, or a polynucleotide or vector encoding the peptide, the amino acid sequence

[0184] (i) is a fragment of the TAA; and

[0185] (ii) comprises a T cell epitope capable of binding the individual's HLAT.

[0186] 9. The method according to item 8, wherein the N- and / or C-terminal flanks of the fragment of the TAA are additional amino acids that are part of a non-TAA sequence.

[0187] 10. The method according to item 5, wherein the TAA is selected from any of those listed in Table 2 or Table 11.

[0188] 11. The method according to item 5, wherein the biological sample comprises blood, serum, plasma, saliva, urine, exhaled breath, cells or tissue.

[0189] 12. A method for treating cancer in an individual in need thereof, comprising:

[0190] administering a cancer treatment to an individual having a lower number of HLA triplets (HLAT) than a threshold derived from a control population of individuals, wherein the HLA triplets are capable of binding a T cell epitope of a tumor-associated antigen (TAA).

[0191] 13. The method according to item 12, wherein the cancer treatment comprises administering to the individual a peptide comprising an amino acid sequence, or a polynucleotide or vector encoding the peptide, the amino acid sequence

[0192] (i) is a fragment of a TAA; and

[0193] (ii) comprises a T cell epitope capable of binding the HLAT of the individual;

[0194] optionally wherein the fragment of the TAA is flanked at the N- and / or C-terminus by additional amino acids that are not part of the TAA sequence.

[0195] 14. The method according to item 12, wherein the TAA is selected from any of those listed in Table 2 or Table 11.

[0196] 15. The method according to item 5, wherein the cancer is selected from melanoma, lung cancer, renal cell carcinoma, colorectal cancer, bladder cancer, glioma, head and neck cancer, ovarian cancer, non-melanoma skin cancer, prostate cancer, kidney cancer, gastric cancer, liver cancer, cervical cancer, esophageal cancer, non-Hodgkin lymphoma, leukemia, pancreatic cancer, uterine body cancer, lip cancer, oral cancer, thyroid cancer, brain cancer, nervous system cancer, gallbladder cancer, laryngeal cancer, pharyngeal cancer, myeloma, nasopharyngeal cancer, Hodgkin lymphoma, testicular cancer, breast cancer and Kaposi sarcoma.

[0197] 16. A system for determining the risk that a human subject will develop cancer, the system comprising:

[0198] (i) a storage module configured to store data comprising the HLA class I genotype of the subject and the amino acid sequence of a TAA;

[0199] (ii) A calculation module configured to quantify the HLAT of the T cell epitopes in the amino acid sequences of the subject that can bind to the TAA, wherein each HLA of the HLAT can bind to the same T cell epitope; and

[0200] (iii) An output module configured to display an indication of the risk that the subject will develop cancer and / or a recommended treatment for the subject.

[0201] Example

[0202] Example 1 HLA epitope binding prediction method and validation

[0203] The prediction of the binding between a specific HLA and an epitope (9mer peptide) is based on the Immune Epitope Database tool (www.iedb.org) for epitope prediction.

[0204] The HLA I epitope binding prediction process was verified by comparison with HLA I epitope pairs determined by laboratory experiments. A dataset of HLA I epitope pairs reported in peer-reviewed publications or public immunology databases was compiled.

[0205] The coincidence rate with the experimentally determined dataset (Table 3) was determined. The binding HLA I epitope pairs of the dataset were correctly predicted with a probability of 93%. Coincidentally, non-binding HLA I epitope pairs were also correctly predicted with a probability of 93%.

[0206] Table 3 Analytical specificity and sensitivity of the HLA epitope binding prediction method

[0207]

[0208] The accuracy of predicting multiple HLA binding epitopes was also determined (Table 4). Based on the analytical specificity and sensitivity with a 93% probability of true positive and true negative predictions and a 7% (= 100% - 93%) probability of false positive and false negative predictions, the probability of the presence of multiple HLA binding epitopes in the human body can be calculated. The probability of multiple HLA binding to epitopes shows the relationship between the number of HLA binding epitopes and the expected minimum actual binding number. According to the PEPI definition, 3 is the expected minimum numerical value (in bold) of the HLA that binds to the epitope.

[0209] Table 4 Accuracy of multiple HLA binding epitope prediction

[0210]

[0211] The verified HLA epitope binding prediction method was used to determine all HLA epitope binding pairs described in the following examples.

[0212] Example 2 Prediction of cytotoxic T lymphocyte (CTL) responses for epitope presentation of multiple HLAs

[0213] This study investigated whether the presentation of one or more epitopes of a polypeptide antigen by one or more HLA class I molecules of an individual predicts a CTL response.

[0214] The study was conducted by a retrospective analysis of six clinical trials in 71 cancer patients and 9 HIV-infected patients (Table 5). Patients from these studies were treated with an HPV vaccine, three different NY-ESO-1-specific cancer vaccines, an HIV-1 vaccine, and an anti-CTLA-4 monoclonal antibody ipilimumab, which has been shown to reactivate CTLs against the NY-ESO-1 antigen in melanoma patients. All of these clinical trials measured antigen-specific CD8+ CTL responses (immunogenicity) in the study subjects after vaccination. In some cases, the correlation between CTL response and clinical response was reported.

[0215] No patients were excluded from the retrospective study for any reason other than data availability. A 157-patient dataset (Table 5) was randomized using a standard random number generator to create two independent cohorts for training and evaluation of the study. In some cases, the cohort contained multiple datasets from the same patient, resulting in a training cohort of 76 datasets from 48 patients and a test / validation cohort of 81 datasets from 51 patients.

[0216] Table 5 Summary of patient datasets

[0217]

[0218]

[0219] The reported CD8+ T responses of the training datasets were compared with the HLA class I restriction profiles of the vaccine antigen epitopes (9mers). Antigen sequences and HLA class I genotypes for each patient were obtained from publicly available protein sequence databases or peer-reviewed publications. The HLA epitope-binding prediction method was blinded to the patient's clinical CD8+ T cell response data, where the CD8+ T cells were IFN-γ-producing CTLs (9mers) specific for the vaccine peptide. The number of epitopes of each antigen predicted to bind to at least 1 (PEPI1+), or at least 2 (PEPI2+), or at least 3 (PEPI3+), or at least 4 (PEPI4+), or at least 5 (PEPI5+), or all 6 (PEPI6+) HLA class I molecules was determined, and the number of HLA bindings was used as a classifier of the reported CTL response. The true positive rate (sensitivity) and true negative rate (specificity) of each classifier (number of HLA bindings) were determined separately from the training dataset.

[0220] ROC analysis was performed for each classifier. In the ROC curve, the true positive rate (sensitivity) at different cut-off points was plotted as a function of the false positive rate (1 - specificity) Figure 1 ). Each point on the ROC curve represents a sensitivity / specificity pair corresponding to a specific decision threshold (epitope (PEPI) count). The area under the ROC curve (AUC) is a measure of the degree to which the classifier can distinguish between two diagnostic groups (CTL responders or non-responders).

[0221] This analysis unexpectedly revealed that epitope presentation predicted by multiple class I HLAs of the subject (PEPI2+, PEPI3+, PEPI4+, PEPI5+ or PEPI6) predicted CD8+ T cell responses or CTL responses better than epitope presentation by just one or more HLA class I molecules in each case (PEPI1+, AUC = 0.48, Table 6).

[0222] Table 6 Diagnostic value of PEPI biomarkers determined by ROC analysis

[0223]

[0224] By considering the antigenic epitopes that can be presented by at least 3 HLA class I alleles of an individual, the CTL response of the individual was best predicted (PEPI3+, AUC = 0.65, Table 7). The threshold count for PEPI3+ (the number of antigen-specific epitopes presented by 3 or more HLAs of an individual) that best predicted a positive CTL response was 1 (Table 7). In other words, at least 3 HLA class I molecules of the subject (≥1 PEPI3+) present at least one antigen-derived epitope, and then that antigen can trigger at least one CTL clone, and the subject is a possible CTL responder. Using the ≥1 PEPI3+ threshold to predict possible CTL responders ("≥1 PEPI3+ test") provided a true positive rate (diagnostic sensitivity) of 76% (Table 7).

[0225] Table 7 Determination of the ≥1 PEPI3+ threshold to predict possible CTL responders in the training dataset

[0226]

[0227] Example 3 Retrospective validation of ≥1 PEPI3+ threshold as a new biomarker for PEPI test

[0228] In a retrospective analysis, a test cohort of 81 data sets from 51 patients was used to validate the ≥1PEPI3+ threshold to predict antigen-specific CD8+ T cell or CTL responses. For each data set in the test cohort, it was determined whether the ≥1PEPI3+ threshold was met (at least one antigen-derived epitope presented by at least three class I HLAs of an individual). This was compared with the experimentally determined CD8+ T cell response (CTL response) reported in the clinical trial (Table 8).

[0229] Retrospective validation demonstrated that the PEPI3+ peptide induced CD8+ T cell responses (CTL responses) in individuals with an 84% probability. 84% is the same value determined in the analytical validation of the PEPI3+ prediction, where the epitope binds to at least 3 HLAs of an individual (Table 4). These data provide strong evidence for the induction of immune responses by PEPI in individuals.

[0230] Table 8 Diagnostic performance characteristics of ≥1PEPI3+ test (n = 81)

[0231]

[0232] ROC analysis using the PEPI3+ count as the cut-off determined the diagnostic accuracy ( Figure 2 ). The AUC value = 0.73. For ROC analysis, an AUC of 0.7 to 0.8 is generally considered a reasonable diagnostic value.

[0233] A PEPI3+ count of at least 1 (≥1PEPI3+) best predicted the CTL responses in the test data set (Table 9). This result confirmed the threshold determined during training (Table 6).

[0234] Table 9 Confirmation of the ≥1PEPI3+ threshold to predict possible CTL responders in the test / validation data set.

[0235]

[0236] Example 4 Clinical validation of PEPI3+ threshold as a new biomarker for PEPI test

[0237] Vaccine design based on the PEPI3+ biomarker has been tested for the first time in a phase I clinical trial in patients with metastatic colorectal cancer (mCRC) in the OBERTO I / II clinical trial (NCT03391232). In this study, we evaluated the safety, tolerability, and immunogenicity of single-dose or multi-dose PolyPEPI1018 as an additional measure for maintenance therapy in mCRC patients. PolyPEPI1018 is a peptide vaccine containing 12 unique epitopes derived from 7 conserved TSAs (WO2018158455A1) that are frequently expressed in mCRC. These epitopes were designed to bind to at least three autologous HLA alleles, which are more likely to induce T cell responses than epitopes presented by a single HLA (see Examples 2 and 3). mCRC patients in the first-line setting received the vaccine (dose: 0.2 mg / peptide) immediately after transitioning to maintenance therapy with fluoropyrimidine and bevacizumab. Vaccine-specific T cell responses were first predicted by in silico identification of PEPI3+ (using the patient's complete HLA genotype and CRC-specific antigen expression rates) and then measured by ELISpot after one vaccination cycle (phase I part of the trial).

[0238] Seventy data sets from 10 patients (phase I cohort and OBERTO trial data set) were used for prospective validation of the PEPI3+ biomarker to predict antigen-specific CTL responses. For each data set, the predicted PEPI3+ was determined in silico and compared with the vaccine-specific immune response measured from the patient's blood by the ELISPOT assay. The diagnostic characteristics (positive predictive value, negative predictive value, overall percentage agreement) determined in this way were then compared with the retrospective validation results described in Example 3.

[0239] The overall percentage agreement was 64%, with a high positive predictive value of 79%, indicating a 79% probability that patients with predicted PEPI3+ would generate CD8 T cell-specific immune responses against the antigens analyzed. The clinical trial data were significantly correlated with the retrospective trial results (p = 0.01) and provided evidence for the calculation of PEPI3+ together with the PEPI assay to predict antigen-specific T cell responses based on the patient's complete HLA genotype (Table 10).

[0240] Table 10 Prospective validation of ≥1 PEPI3+ and the PEPI assay

[0241]

[0242]

[0243] *51 patients; 6 clinical trials; 81 datasets **10 patients; Treos Phase I clinical trial (OBERTO); 70 datasets Example 5 Prediction of melanoma risk by HLA class I genotype (HLAT score-based method)

[0244] Select putative immunoprotective tumor antigens

[0245] Assume that tumor-specific antigens (TSAs) are immunoprotective antigens because cancer patients with spontaneous TSA-specific T cell responses have a favorable clinical course. Forty-eight TSAs expressed in different tumor types were selected to study protective tumor-specific T cell responses (Table 11). These TSAs have been studied in melanoma and other cancers and have been shown to induce spontaneous T cell responses.

[0246] Table 11 TSAs selected for risk analysis

[0247]

[0248]

[0249] CRC: Colorectal cancer, NSCLC: Non-small cell lung cancer, HNSCC: Head and neck squamous cell carcinoma, RCC: Renal cell carcinoma

[0250] The HLAT score of the subject is a risk factor for developing melanoma related to the HLA genotype

[0251] The HLAT number predicts the breadth of T cell responses to the 48 selected TSAs. Assume that not all subjects' HLAT plays an equally important role in the immune control of melanoma. Therefore, HLAT (for the 48 TSAs) is weighted based on the ability to separate melanoma patients from the general population. Generally, the greater the weight, the more important the corresponding TSA. In fact, using the initial weight (truncated log p-value), the AUC is already higher than 0.6.

[0252] Performance of the binary classifier in separating melanoma patients from the background

[0253] This study used binary classification (see Methods) to compare a US subset (n = 1400) from the dbMHC dataset (7,189 patient cohort) with melanoma subjects also of US origin (n = 513). Figure 3 The ROC curve obtained using the HLAT score as a binary classifier is shown. The HLAT score predicts which of two possible groups a subject belongs to: the melanoma cancer group or the background population. The ROC curve is presented by plotting the true positive rate (TPR) against the false positive rate (FPR) at various HLAT score thresholds.

[0254] The obtained AUC value was 0.645. This value indicates a significant separation between the two groups, especially since in the case of melanoma / carcinogenesis, there is not only a single differentiating cause (such as HLAT). The most significant sunlight and indoor tanning exposures are important determinants of melanoma risk, phenotypes such as blond or red hair, blue eyes, and freckles, and genetic factors such as the high-penetrance, 3 moderate-penetrance, and 16 low-penetrance genes associated with melanoma described by Read et al. (J. Med. Genet. 2016; 53(1): 1-14). In fact, the transformed z-score of 10.065 achieved in this study was highly significant (p < 0.001).

[0255] The HLAT score of the subject is a risk factor for developing melanoma associated with the HLA genotype

[0256] The total test population (background population mixed with the cancer population) was divided into five equal-sized groups based on the HLAT score. The relative immune risk (RiR) in each group was determined compared to the risk of the average US population ( Figure 4 ). For example, the risk of developing melanoma in the first subgroup was 4.4%, while the US average was 2.6%. Therefore, this subgroup had a relative immune risk of 1.7. The group with the lowest HLAT score represents the population with the highest immunological risk of developing cancer. The group with the highest HLAT score represents the population with the lowest immune risk of developing cancer. The most at-risk subgroup consists of those subjects with an HLAT score less than 26, and the HLAT score varies between 29 and 51 in the second-risk subgroup. In the middle 20% are those subjects with an HLAT score greater than or equal to 51 and less than 88 and RiR < 1, indicating that certain HLAT scores are associated with a reduced melanoma risk. Interestingly, this HLAT score range of 51–88 is similar to the HLAT score (75) that can separate populations with low and high melanoma incidences ( Figure 4 ). In the second most protected subgroup, the HLAT score is between 88 and 164. Finally, in the most protected subgroup, each subject has an HLAT score of at least 164. In these subgroups, as Figure 4 shown, the relative immune risk of developing melanoma decreases monotonically, and although there is no significant change between consecutive groups, the difference between the first and last groups is significant (p = 0.001).

[0257] Example 6 Prediction of risks of different types of cancers by HLA class I genotype (HLAT score-based method)

[0258] Similar analyses were performed for six other cancer indications. The results are summarized in Table 12, where the AUC values were significant for melanoma, lung cancer, renal cell carcinoma, colorectal cancer, and bladder cancer. The p-value was not significant for head and neck cancer. However, head and neck cancer is associated with viral HPV infection. Only TSA was used in this study, excluding viral proteins. The risk of developing certain cancers such as head and neck cancer that may be associated with viral infection can be better determined by including viral antigens in the analysis.

[0259] Table 12 Summary of immunological risk prediction for different types of cancer compared to the average population

[0260]

[0261] By dividing the test population (background population mixed with the cancer population) into five equally large subgroups based on HLA T scores, we were able to calculate the relative immune risks associated with certain HLA T scores in the case of non-small cell lung cancer, renal cell carcinoma, and colorectal cancer ( Figure 5 A - C). For other indications, the number of cancer subjects in the subgroups was too small to perform a similar analysis.

[0262] The relative immune risk ratio between the risk subgroup (the 20% of the test population with the lowest HLA T scores) and the protected subgroup (the 20% of the test population with the highest HLA T scores) was calculated and compared to the risk of the average US population. For example, the risk of developing melanoma in the most at-risk subgroup characterized was 4.4%. The US average was 2.4%, so the risk group had a relative immune risk of 1.7. The risk of developing melanoma in the protected group was 0.7%. That is, the relative immune risk of the most protected group was 0.31. In other words, compared to the average population, this group had less than three times the risk of developing melanoma. The risk ratio achieved for melanoma was 5.53 (Table 12).

[0263] Methods of Examples 5, 6, and 10

[0264] HLA genotype data of individuals in the general population

[0265] 7,189 eligible subjects with complete 4-digit HLA genotypes were identified from the dbMHC database.

[0266] HLA genotype data of cancer patients

[0267] Eligible patients had complete 4-digit HLA class I genotypes. Data from 513 melanoma patients were obtained from the following sources:

[0268] The complete four-digit HLA class I genotypes of 429 melanoma subjects were obtained from three peer-reviewed publications (Snyder et al. N Engl J Med. 2014;371(23):2189-99, Van Allen et al. Science. 2015;350(6257):207-11, Chowell et al. Science. 2018;359(375):582-7). At Memorial Sloan Kettering Cancer Center, patients were treated with anti-CTLA-4 and / or PD-1 / PD-L1 inhibitors at Memorial Sloan Kettering Cancer Center (MSKCC), New York. High-resolution HLA class I genotyping from normal DNA was performed using DNA sequencing data or clinically validated HLA typing assays performed by LabCorp. HLA genotypes of 17 stage III / IV melanoma patients were provided by MSKCC. These patients were treated with ipilimumab at MSKCC, New York (Yuan et al. Proc Natl Acad Sci U S A. 2011;108(40):16723-8). Sixty-five melanoma patients, who were from a phase 3 randomized double-blind multicenter study (CA184007, NCT00135408) and a phase 2 (CA184002, NCT00094653) study, had patients who had unresectable stage III or IV malignant melanoma and previously treated unresectable stage III or IV melanoma, respectively. These 65 patients treated at MSKCC, New York had available samples for HLA testing provided by Bristol-Myers-Squibb. Samples were retrospectively tested with NGS group G resolution, and HLA allele interpretation was based on the IMGT / HLA database version 3.15. HLA results were obtained using sequence-based typing (SBT), sequence-specific oligonucleotide probes (SSOP), and / or sequence-specific primers (SSP) to obtain the required resolution. HLA testing was performed by LabCorp, USA.

[0269] HLA genotype data for 370 non-small cell lung cancer, 129 renal cell carcinoma, 87 bladder cancer, 82 glioma, and 58 head and neck cancer subjects were collected from a peer-reviewed publication (Chowell et al.).

[0270] Data on HLA genotypes from 37 colorectal cancer (CRC) patients were obtained from the National Center for Biotechnology (NCBI) Sequence Read Archive, Encyclopedia of deoxyribonucleic acid elements (Boegel et al. Oncoimmunology. 2014; 3(8): e954893). Blood samples from CRC patients were obtained from Asterand Bioscience, and HLA genotypes were identified by LabCorp (Burlington NC).

[0271] TSA sequence data

[0272] Forty-eight TSAs were selected. Amino acid sequence data for these antigens were obtained from UniProt.

[0273] Incidence

[0274] Incidence rates were obtained from http: / / globocan.iarc.fr / Pages / online.aspx.

[0275] Human leukocyte antigen triplet (HLAT)

[0276] HLA class I genes are expressed in most cells and bind to epitopes recognized by T cell receptors. Epitopes that bind to at least three HLAs (HLA triplets or HLATs) out of the six HLA alleles in humans can generate a T cell response. For each j = 1, 2,... 6, we established a scoring system to score the subject's immune system based on the degree to which the subject's immune system can bind to epitopes. Based on combinatorics, there are possible sets of HLA alleles j for a specific epitope, where k is the number of autologous HLA alleles that can bind to the epitope. When we are interested in HLA triplets, j = 3. Thus, the number of antigen HLATs for a subject is defined as the sum of HLATs.

[0277] The subject's HLATs were identified using the PEPI test, which was validated to identify HLA-binding epitopes with 93% accuracy.

[0278] Immunogenetic predictor: HLAT score

[0279] The HLAT score for subject x is defined as:

[0280]

[0281] Where C is the set of TSAs, c is a specific TSA, w(c) is the weight of TSA c, and p(x,c) is the number of HLATs of TSA c in subject x.

[0282] HLAT score weight optimization

[0283] For each TSA for which the HLAT score does not significantly separate cancer patients from the background population, the initial weight is 0. Since we assume that having HLAT does not increase the chance of developing cancer, only non - negative weights are considered. The initial weight is defined as

[0284]

[0285] where t(c) represents the p - value of the one - sided t - test of the HLAT score of TSA c for cancer and background groups, and 48 is the Bonferroni correction.

[0286] The initial weights are further optimized using parallel annealing. Six parallel Markov chains have been applied, with temperatures RT = 0.001, 0.01, 0.02, 0.04, 0.1, 0.2. The energy is assumed to be defined as - 1 times the sum of RiRR (Relative Immune Risk Ratio, see below) and AUC. The weights that provide the maximum relative risk ratio have been reported.

[0287] Relative immune risk (RiR)

[0288] RiR is calculated as the risk ratio between subgroups and the total test population (cancer population and background population), with a 95% confidence interval (CI). For this purpose, considering the lifetime risk, the general population is combined in a way similar to the percentages of different cancer patients in the general US population. The lifetime risks of developing different types of cancer were obtained from the website of the American Cancer Society. Generally, the lifetime risks are different for men and women, so we take their (harmonic) mean. The risks so obtained are: 1:38 for melanoma, 1:16 for lung cancer, 1:61 for renal cell carcinoma, 1:23 for colorectal cancer, 1:41 for bladder cancer, 1:55 for head and neck cancer, and 1:161 for glioma. RiR>1 indicates that the subject has a higher risk of developing a specific cancer compared to subjects in the average population.

[0289] RiR ratio (RiRR)

[0290] The RiR ratio is calculated as the ratio between the groups with the highest and lowest HLAT scores.

[0291] Example 7 HLA score based on HLA triplet provides the best separation between cancer and background subjects

[0292] When developing the screening assay, we considered several scoring schemes. The potential scoring schemes differed in the minimum size of the HLA allele group binding to a specific epitope, which was considered to contribute to the subject's score. For each size j = 1, 2, ..., 6 of the HLA allele subset, we calculated the significance score for each allele based on the frequency with which each allele participated in training subjects' HLA j-tuples binding to the specific epitope. Briefly, if subjects with a given HLA allele had significantly more epitopes with HLA j-mers than subjects without the given HLA allele, then we considered the significance score to be positive. If subjects with a given HLA allele had significantly fewer HLA j-mer epitopes than subjects without the given HLA allele, then the significance score was negative. Then for each subject, we summed the significance scores of his / her HLA alleles. Next, we tested how well these total scores could distinguish melanoma and background subjects by calculating the area under the receiver operating characteristic curve (ROC-AUC, AUC). According to Table 13, j = 2 and j = 3 equally achieved the best separation of the melanoma and background populations, based on the significant difference between the AUC values of the 1-group and j-group different scores, j > 1, indicating that the presentation of epitopes by multiple HLA alleles plays an important role in generating an effective anti-tumor immune response. Additionally, these results suggest that it would be challenging to separate cancer and background (healthy) subjects based on a single allele of their HLA genotype. The decrease in the AUC value when j = 6 can be explained by the fact that there are only a very limited number of epitope-HLA allele combinations in which all 6 HLA alleles of a subject can bind to the epitope.

[0293] Table 13 Calculated AUC values for melanoma with different HLA j-groups

[0294] j AUC j AUC 1-group 0.60 4-group 0.68 2-group 0.69 5-group 0.68 3-group 0.69 6-group 0.61

[0295] Example 8 HLA score is a risk or protective indicator for melanoma, interpreted as RiR and RiRR

[0296] Comparison of the AUC values for melanoma and background subjects in the United States (0.69) indicates a significant separation between the two groups using the HLA score. In fact, the transformed z-value was 12.57, which is highly significant (p < 0.001). These results demonstrate that the HLA genotype of the subjects affects the genetic risk of developing melanoma.

[0297] Based on the HLA score, the background and melanoma populations were divided into five equal-sized subgroups according to their HLA scores (s); s < 34, 34 ≤ s < 55, 55 ≤ s < 76, 76 ≤ s < 96, and 96 < s. The relative risk (RR) for each subgroup was calculated ( Figure 6). We found that the subjects with the highest immune risk of developing melanoma (6.1%) were in the lowest HLA score subgroup (s<34). Since the average risk of melanoma in the United States is 2.6%, subjects in the s<34 subgroup had a 2.3-fold higher risk of melanoma than the average American subject. Conversely, subjects with the highest HLA scores (96 <s)的亚组代表患黑素瘤免疫风险最低(1.1%)的受试者。该亚组中的受试者比美国的平均受试者具有低0.42倍的风险。第一和最后亚组之间的差异是显著的(p<0.05)。

[0298] We calculated the most protected group and the group at risk (RR extremities ). We found that the RR of melanoma extremities The HLA score was 5.69, indicating that subjects with an HLA score below 34 had an approximately 6-fold higher risk of developing melanoma compared to subjects with an HLA score above 96 (Table 14).

[0299] Example 9 Performance of HLA score as a predictor of risks of developing different types of cancers

[0300] In some cases, the significance score of an HLA allele (h) is defined as

[0301]

[0302] Wherein u(h) is the p value of the double-sided u test of allele h, which determines whether the number of HLAT in two individual subgroups is different: the individual in a subgroup has HLA h, and the individual in a subgroup does not have HLA h. B is the Bonferroni correction, if the mean number of HLAT in the subgroup with h allele is greater than the mean number of HLAT in the subgroup without h, then sign(h) is +1, otherwise -1. In some cases, any suitable method known to those skilled in the art can be used to further optimize this initial score. In some cases, the summation of these significance scores is used to determine that the risk of a subject developing cancer is related to the risk of a subject developing cancer.

[0303] The specific score to be used depends on the indication and the prior data. In some cases, the selection will be made based on the performance of the different calculations of the available test data sets. Performance can be evaluated by AUC value (area under the ROC curve) or by any other performance goodness score known to those skilled in the art.

[0304] We determined ROC curves, RR and RR for non-small cell lung cancer, renal cancer, colorectal cancer, bladder cancer, head and neck cancer, and glioma using the same methods as described for melanoma. extremities (Table 14). The ROC-AUC values ​​were significant for all cancer types except colorectal cancer.

[0305] We obtained the RR for the cancer indications studied extremities ranging from 2.35 - 5.69, indicating different levels of immune protection against different types of cancer (Table 14). However, for all cancer indications, the RR extremities > 2 indicates that the HLA genotype represents a substantial genetic risk for developing cancer.

[0306] Table 14 Immune risk prediction in different cancer types

[0307]

[0308] * Risk group, the lowest 20% of the general population in terms of HLA score; Protected group, 20% of the general population with the highest HLA score. Each cancer indication was classified against the same background population. RR extremities is the risk ratio of the most at - risk and most protected groups; AUC, area under the ROC curve. A Bonferroni - corrected p - value less than 0.007 indicates significance.

[0309] Example 10 Risk screening and vaccine design for patient D with CRC

[0310] This example shows how to calculate the HLAT score for Patient D described in Example 19, who has been diagnosed with metastatic colorectal cancer. Using the HLA genotype of Patient D, the predicted numbers of the PEPI3, PEPI4, PEPI5, and PEPI6 epitopes on 48 selected TSAs were determined (Table 15). Based on statistics, the total HLAT for each TSA (rows 6, 14, and 22 of Table 15) and the weighted score for each TSA (rows 8, 16, and 24 of Table 15) were calculated. This weighted score is simply the product of the total HLAT and the weight of the TSA (rows 7, 15, and 23 of Table 15). The weights were obtained using the method described in the "HLAT score weight optimization" section of Example 6. The total weighted score (as described in Equation (1)) is 43.09. Based on the comparison between US CRC and the US background population, the risk of Patient D developing colorectal cancer is 1.26 times that of the average person in the US. Since the risk of CRC occurring in the US is 4.2%, based on our results, the risk for Patient D is 5.3%.

[0311] Table 15

[0312]

[0313]

[0314] Example 11 Results of CRC phase I trial: PEPI vs HLAT for immunogenicity

[0315] In the OBERTO trial, we predicted the immune responses of 7 antigens and 11 subjects and also measured the immune responses of 10 patient samples. The 7 antigens of the vaccine were part of 48 TSAs. The prediction and measurement are summarized in Table 16, and the overall percentage agreement was 64%.

[0316] Table 16 Measurement and PEPI test predicted the immune responses to the vaccine-containing peptides specific to the listed TSAs.

[0317]

[0318] We compared the HLAT scores and the number of antigens with the measured immune responses ( Figure 7 ). We found a positive correlation between the HLAT scores and the number of antigens and the immune responses. However, we did not expect a significant correlation with such a small number of measurements (n = 10), and since the HLAT score takes into account the epitope binding of the predicted 48 antigens and the immune responses were only measured for 7 of the 48 antigens, this analysis was able to show a correlation but provided low statistical power.)

[0319] Example 12 Comparison between classification based on HLAT score and classification based on HLA score

[0320] Table 17 Classification based on HLAT scores:

[0321]

[0322]

[0323] Table 18 Classification based on HLA scores:

[0324]

[0325] It can be seen that the classification based on HLAT scores is better in the case of colorectal cancer, while the classification based on HLA scores is better in the case of head and neck cancer.

[0326] Example 13 HLA score of CLL related to HLA

[0327] A*02:01, C*05:01, and C*07:01 are HLA alleles associated with CLL (chronic lymphocytic leukemia) (Gragert et al., 2014), i.e., subjects with any of these HLA class I alleles have an increased risk of developing CLL. During HLA score training, we observed that subjects in the training population with any of these HLAs had significantly fewer HLATs for the 48 TSAs analyzed than subjects without these HLAs. Table 19 shows the average number of HLATs for the 48 TSAs in the case of the 9 most common HLA alleles. However, these few HLA alleles can only be found in a small fraction of the population, and thus, the information obtained from the association between cancer and these few alleles cannot be used for subjects without any of these alleles. On the other hand, the HLA score method assigns informative scores to all subjects and can thus be used to classify the entire population. Therefore, the HLA score method provides a better classification than methods that only use information on the association between individual HLA alleles and cancer.

[0328] Table 19 presents an HLAT analysis of individuals with one of the HLA A*02:01 or C*05:01 or C*07:01 alleles, which are associated with an increased risk of CLL.

[0329]

[0330] Example 14 One allele or incomplete HLA genotype is not suitable for determining genetic risk

[0331] It is known that Epstein-Barr virus (EBV) infection can induce undifferentiated nasopharyngeal carcinoma (UNPC). Pasini et al. analyzed 82 Italian UNPC patients and 286 bone marrow donors from the same population and observed that some conserved alleles, A*0201, B*1801, and B*3501 HLA, which are capable of binding some EBV epitopes in a given region, were not fully represented in UNPC subjects (Pasini E et al. Int. J. Cancer: 125, 1358–1364 (2009)). However, the study of frequent alleles in a population is a completely different approach from the study of the true target HLA combinations that induce immune responses, such as the analysis of an individual's HLA T repertoire. Because the latter suggests the potential of a person to generate diseased cells that kill all components of T cells, which is the mechanism for explaining the "progress" or risk of immunogenetics. In addition, they found an additive effect on protective HLA alleles. However, they did not infer whether these HLA alleles could bind the same epitopes or different epitopes on different EBV antigens. They also found HLA alleles that were positively correlated with UNPC. However, they could not measure the reduced ability of these HLA alleles to bind EBV epitopes. They only considered antigens from EBV, so their method could not be generalized to other cancers. Since even the most common HLA alleles only cover a limited part of the entire population, a diagnostic device cannot be constructed based only on them. For example, in a device based only on the A*02:01 allele, the AUC value was only 0.573( Figure 8 ). The combined haplotype A*02:01 / B*18:01 is even rarer, and although the OR value is high, a device based on the analysis of a single "haplotype" would only have an AUC value of 0.556. This means that it could not significantly separate the population consisting of 82 UNPC patients from the background of 286 subjects, with a transformed Z value of 1.65 and a corresponding p value (for a one-sided test) of 0.06. Example 15 Study design and Preliminary safety data of the I / II phase clinical trial of OBERTO

[0332] The OBERTO trial is a phase I / II trial of the PolyPEPI1018 vaccine and CDx for the treatment of metastatic colorectal cancer (NCT03391232). The study design is as Figure 9 shown.

[0333] Inclusion criteria

[0334] · Histologically confirmed metastatic adenocarcinoma originating from the colon or rectum

[0335] · Presence of at least 1 measurable reference lesion according to RECIST 1.1

[0336] · Partial response or stable disease during first-line treatment with a systemic chemotherapy regimen and 1 biological treatment regimen

[0337] Maintenance therapy with a fluoropyrimidine (5-fluorouracil or capecitabine) plus the same biologic agent (bevacizumab, cetuximab, or panitumumab) during induction, planned to start before the first day of treatment with the investigational drug

[0338] · Last CT scan 3 weeks or less before the first day of treatment

[0339] Subject withdrawal and discontinuation

[0340] · In the initial study phase (12W), if a patient experiences disease progression and needs to start second-line treatment, the patient will withdraw from the study.

[0341] · During the second part of the study (after the second dose), if a patient experiences disease progression and needs to start second-line treatment, the patient will remain in the study, receive the third vaccination as planned, and complete the follow-up.

[0342] · As expected, transient local erythema and edema at the injection site, as well as a flu-like syndrome with mild fever and fatigue, were observed. These reactions are known for peptide vaccination and are generally related to the mechanism of action, as fever and flu-like syndrome may be consequences and signs of induction of an immune response (this is called the typical vaccine response for childhood vaccination).

[0343] · Only one serious adverse event (SAE) "possibly related" to the vaccine was recorded (Table 20).

[0344] · A single-dose limiting toxicity (DLT) (syncope) unrelated to the vaccine occurred.

[0345] Safety results are summarized in Table 19.

[0346] Table 20 Serious adverse events reported in the OBERTO clinical trial. No related SAEs occurred (only 1 "possibly related").

[0347]

[0348]

[0349] Example 16 Target antigen selection based on expression frequency during vaccine design and its clinical validation for mCRC

[0350] Shared tumor antigens can precisely target all tumor types - including those with a low mutational burden. Population expression data previously collected from 2,391 CRC biopsies represent the variability of antigen expression among CRC patients worldwide ( Figure 10 A).

[0351] PolyPEPI1018 is a peptide vaccine that we designed to contain 12 unique epitopes from 7 conserved testis-specific antigens (TSAs) that are frequently expressed in mCRC. In our model, we hypothesized that by selecting TSAs that are frequently expressed in CRC, target identification would be correct and the need for tumor biopsies would be eliminated. We have calculated that the probability of 3 out of 7 TSAs being expressed in each tumor is greater than 95% ( Figure 10 B).

[0352] In a phase I study, we evaluated the safety, tolerability, and immunogenicity of PolyPEPI1018 as an adjunct to maintenance therapy in patients with metastatic colorectal cancer (mCRC) (NCT03391232) (see also Example 4).

[0353] Immunogenicity measurements demonstrated pre-existing immune responses and indirectly confirmed the expression of target antigens in patients. Immunogenicity was measured in PBMC samples isolated before vaccination and at different time points following a single immunization with PolyPEPI1018 using an enriched fluorescence spot assay (ELISPOT) to confirm vaccine-induced T cell responses; PBMC samples were stimulated in vitro with vaccine-specific peptides (9mer and 30mer) to determine vaccine-induced T cell responses above baseline. On average, 4 and at least 2 patients had pre-existing CD8 T cell responses against each target antigen ( Figure 10 C). Seven out of 10 patients had pre-existing immune responses against at least 1 antigen (average 3) ( Figure 10 D). These results provide evidence of correct target selection, as the responses of CD8+ T cells to CRC-specific target TSAs before vaccination with the PolyPEPI1018 vaccine confirmed the expression of that target antigen in the patients analyzed. Targeting true (expressed) TSAs is a prerequisite for an effective tumor vaccine.

[0354] Example 17 Preclinical and clinical immunogenicity of PolyPEPI1018 vaccine demonstrate appropriate peptide selection

[0355] The PolyPEPI1018 vaccine contains six 30mer peptides, each designed by linking two immunogenic 15mer fragments derived from 7 TSAs (each fragment involves 9mer PEPI and thus, by design, there are 2 PEPI in each 30mer) ( Figure 11 ). Based on the analysis of 2,391 biopsies, these antigens are frequently expressed in CRC tumors ( Figure 10 ).

[0356] Preclinical immunogenicity results calculated for the model cohort (n = 433) and CRC cohort (n = 37) based on PEPI test predictions led to 98% and 100% predicted immunogenicity, and this was clinically demonstrated in the OBERTO trial (n = 10), where immune responses to at least one antigen were measured in 90% of patients. More interestingly, 90% of patients had vaccine peptide-specific immune responses against at least 2 antigens, and 80% had CD8+ T cell responses against 3 or more different vaccine antigens, showing evidence of appropriate target antigen selection during the PolyPEPI1018 design. CD4+ T cell specificity and CD8+ T cell specificity clinical immunogenicity are detailed in Table 21. High immune response rates were found for both effector and memory effector T cells, CD4+ and CD8+ T cells, and 9 out of 10 patients were boosted or re-induced by the vaccine. Additionally, the proportions of CRC-reactive, multifunctional CD8+ and CD4+ T cells increased 2.5-fold and 13-fold, respectively, in the patients' PBMCs after vaccination.

[0357] Table 21 Clinical immunogenicity results of PolyPEPI1018 in mCRC.

[0358]

[0359] Example 18 Clinical response to PolyPEPI1018 treatment

[0360] Analyze the preliminary objective tumor response rate (RECIST 1.1) (Figure 12) of the OBERTO clinical trial (NCT03391232) further described in Examples 4, 15, 16, and 17. Among eleven vaccinated patients receiving maintenance therapy, 5 had stable disease (SD) at the time point of the preliminary analysis (12 weeks), 3 had unexpected tumor responses (partial responses, PR) during treatment (maintenance therapy + vaccination), and 3 had progressed to disease (PD) according to the RECIST 1.1 criteria. Stable disease was achieved as the best response in 69% of patients on maintenance therapy (capecitabine and bevacizumab). The efficacy of Patient 020004 was durable 12 weeks later, and the efficacy of Patient 010004 was durable and suitable for radical surgery. After the third vaccination, this patient had no signs of disease and was thus a complete responder, as shown in the waterfall plot in Figure 12.

[0361] After one vaccination, the ORR was 27% and the DCR was 63%. Among patients receiving at least 2 doses (out of 3 doses), 2 out of 5 had an ORR (40%), and the DCR was as high as 80% (4 out of 5 patients had SD + PR + CR) (Table 22).

[0362] Table 22 Clinical responses for POEPI1018 treatment after ≥1 and ≥2 vaccine doses

[0363]

[0364] Based on data from 5 patients who received multiple doses of the PolyPEPI1018 vaccine in the OBERTO-101 clinical trial, preliminary data indicate that higher AGP counts (>2) are associated with longer PFS and increased tumor size reduction ( Figure 10 B and C).

[0365] Example 19 Personalized immunotherapy (PIT) design and treatment for ovarian cancer, breast cancer and colorectal cancer

[0366] This example provides evidence of concept data from 4 patients with metastatic cancer who were treated with a personalized immunotherapy vaccine composition to support the principle of inducing a cytotoxic T cell response with multiple HLA-binding epitopes in the subject, and this disclosure is in part based on this principle.

[0367] Composition for treating ovarian cancer with POC01–PIT (patient A)

[0368] This example describes treating a patient with ovarian cancer with a personalized immunotherapy composition, where the composition is specifically designed for the patient based on their HLA genotype, based on the disclosure described herein.

[0369] The HLA class I and class II genotypes of a patient with metastatic ovarian adenocarcinoma (Patient A) were determined from a saliva sample.

[0370] To prepare a personalized pharmaceutical composition for Patient A, 13 peptides were selected, each of which met the following two criteria: (i) derived from antigens expressed in ovarian cancer, as reported in peer-reviewed scientific publications; and (ii) contained fragments of T cell epitopes capable of binding at least three HLA class I molecules of Patient A (Table 23). In addition, each peptide was optimized to bind the maximum number of HLA class II of the patient.

[0371] Table 23 Personalized vaccine for ovarian cancer patient A.

[0372] Vaccine of POC01 for patient A Target antigen Antigen expression 20mer peptide HLA class I maximum HLA class II maximum SEQ ID NO POC01_P1 AKAP4 89％ NSLQKQLQAVLQWIAASQFN 3 5 1 POC01_P2 BORIS 82％ SGDERSDEIVLTVSNSNVEE 4 2 2 POC01_P3 SPAG9 76％ VQKEDGRVQAFGWSLPQKYK 3 3 3 POC01_P4 OY-TES-1 75％ EVESTPMIMENIQELIRSAQ 3 4 4 POC01_P5 SP17 69％ AYFESLLEKREKTNFDPAEW 3 1 5 POC01_P6 WT1 63％ PSQASSGQARMFPNAPYLPS 4 1 6 POC01_P7 HIWI 63％ RRSIAGFVASINEGMTRWFS 3 4 7 POC01_P8 PRAME 60％ MQDIKMILKMVQLDSIEDLE 3 4 8 POC01_P9 AKAP-3 58％ ANSVVSDMMVSIMKTLKIQV 3 4 9 POC01_P10 MAGE-A4 37％ REALSNKVDELAHFLLRKYR 3 2 10 POC01_P11 MAGE-A9 37％ ETSYEKVINYLVMLNAREPI 3 4 11 POC01_P12a MAGE-A10 52％ DVKEVDPTGHSFVLVTSLGL 3 4 12 POC01_P12b BAGE 30％ SAQLLQARLMKEESPVVSWR 3 2 13

[0373] Based on the validation of the PEPI test shown in Table 4, 11 PEPI3 peptides in the immunotherapy composition can induce T cell responses in Patient A with an 84% probability, and two PEPI4 peptides (POC01-P2 and POC01-P5) can induce T cell responses with a 98% probability. The T cell responses target 13 antigens expressed in ovarian cancer. The expression of these cancer antigens in Patient A was not tested. Instead, the probability of successfully killing cancer cells was determined based on the probability of antigen expression in the patient's cancer cells and the positive predictive value of ≥1 PEPI3+ test (AGP count). The AGP count predicts the effectiveness of the vaccine in a subject: the number of vaccine antigens expressed by PEPI in the patient's tumor (ovarian adenocarcinoma). The AGP count represents the number of tumor antigens that the vaccine recognizes and induces T cell responses against the patient's tumor (hits the target). The AGP count depends on the vaccine-antigen expression rate in the subject's tumor and the HLA genotype of the subject. The correct value must be between 0 (no antigen-presenting PEPI expressed) and the maximum number of antigens (all antigens are expressed and present PEPI).

[0374] The probability that Patient A will express one or more of the 13 antigens is as Figure 13 shown. AGP95 (AGP with 95% probability) = 5, AGP50 (average expected value - discrete probability distribution) = 7.9, mAGP (probability that AGP is at least 2) = 100%, AP = 13.

[0375] The pharmaceutical composition for Patient A can consist of at least 2 of the 13 peptides (Table 23), because it is determined that presenting at least two polypeptide fragments (epitopes) in the vaccine or immunotherapy composition that can bind to at least three HLAs of an individual (≥2 PEPI3+) predicts a clinical response. The synthetic peptides are dissolved in a pharmaceutically acceptable solvent and mixed with an adjuvant before injection. It is desirable for the patient to receive personalized immunotherapy with at least two peptide vaccines, but it is more preferable to increase the likelihood of killing cancer cells and reduce the chance of recurrence.

[0376] To treat Patient A, the 13 peptides are formulated into 4×3 or 4 peptides (POC01 / 1, POC01 / 2, POC01 / 3, POC01 / 4). One treatment cycle is defined as administering all 13 peptides within 30 days.

[0377] Patient history:

[0378] Diagnosis: Metastatic ovarian adenocarcinoma

[0379] Age: 51 years old

[0380] Family history: Colon cancer and ovarian cancer (mother); breast cancer (grandmother)

[0381] Tumor pathology:

[0382] 2011: First diagnosis of ovarian adenocarcinoma; Wertheim operation and chemotherapy; lymph node dissection

[0383] 2015: Pericardial adipose tissue metastasis, resection

[0384] 2016: Liver metastases

[0385] 2017: Retroperitoneal and mesenteric lymph node enlargement; initial peritoneal cancer with a small amount of ascites

[0386] Previous treatments:

[0387] 2012: Paclitaxel - carboplatin (6×)

[0388] 2014: Caelyx - carboplatin (1×)

[0389] 2016 - 2017 (9 months): Lymparza (olaparib) 2×400 mg / day, orally

[0390] 2017: Hycamtin infusion 5×2.5 mg (3× one series / month)

[0391] PIT vaccine treatment started on April 21, 2017. Figure 14 。

[0392] 2017 - 2018: Patient A received 8 vaccination cycles as adjuvant treatment and survived 17 months (528 days) after the start of treatment. During this time interval, she experienced partial response as the best response after the 3rd and 4th vaccine treatments. She died in October 2018.

[0393] Interferon (IFN)-γ ELISPOT bioassay confirmed the predicted T cell responses of patient A to 13 peptides. Positive T cell responses (defined as, >5 - fold higher than the control, or >3 - fold higher than the control and >50 points) were detected in all 13 20mer peptides and all 13 9mer peptides (each with the sequence of the PEPI of the peptide that can bind to the most HLA class I alleles of patient A) ( Figure 15 )。

[0394] Results of the patient's tumor MRI (baseline: April 15, 2016) (BL: Figure 16 baseline for tumor response evaluation

[0395] The disease was mainly limited to the liver and lymph nodes. The use of MRI limited the detection of metastases in the lungs (pulmonary)

[0396] May 2016 to January 2017: Olaparib treatment (FU1: Figure 16 follow - up 1)

[0397] December 25, 2016 (before PIT vaccination), in (FU2: Figure 16 in follow-up 2) confirmed a significant reduction in tumor burden

[0398] January to March 2017 - TOPO regimen (topoisomerase)

[0399] April 6, 2017: ( Figure 16 in follow-up 3) showed regrowth of existing lesions and the appearance of new lesions leading to disease progression. Ascites-increased carcinomatosis peritonei. Progressive liver tumors and lymph nodes

[0400] PIT started on April 21, 2017

[0401] June 26, 2017 (after the 2nd PIT cycle): ( Figure 16 in follow-up 4) progression / pseudoprogression

[0402] Rapid progression in lymph nodes, liver, retroperitoneal and thoracic regions, with large amounts of pleural fluid and ascites. Initiated carboplatin, gemcitabine, bevacizumab.

[0403] September 20, 2017 (after the 3rd PIT cycle): ( Figure 16 in follow-up 5) partial response

[0404] Complete remission of pleura / fluid and ascites

[0405] Remission in liver, retroperitoneal region and lymph nodes

[0406] These findings suggest pseudoprogression.

[0407] November 28, 2017 (after the 4th PIT cycle): ( Figure 16 in follow-up 6) partial response

[0408] Complete remission in the chest. Remission in liver, retroperitoneal region and lymph nodes

[0409] April 13, 2018: Progression

[0410] Complete remission in the chest and retroperitoneal region. Progression in the liver center and lymph nodes

[0411] June 12, 2018: Stable disease

[0412] Complete remission in the chest and retroperitoneal region. Minimal regression in the liver center and lymph nodes

[0413] July 2018: Progression

[0414] October 2018: Patient died

[0415] Partial MRI data of Patient A are shown in Table 24 and Figure 16 .

[0416] Table 24 Summary table of injury responses

[0417]

[0418] Design, safety, and immunogenicity of a personalized immunotherapy composition (PBRC01) for the treatment of metastatic breast cancer (Patient B) PT9 Safety and immunogenicity

[0419] The HLA class I and class II genotypes of metastatic breast cancer Patient B were determined from saliva samples. To prepare a personalized pharmaceutical composition for Patient B, 12 peptides were selected, each of which met the following two criteria: (i) derived from antigens expressed in breast cancer, as reported in peer-reviewed scientific publications; and (ii) contained fragments of T cell epitopes capable of binding at least three HLA class I molecules of Patient B (Table 25). In addition, each peptide was optimized to bind the maximum number of HLA class II of the patient. These 12 peptides target 12 breast cancer antigens. The probability that Patient B will express one or more of the 12 antigens is as Figure 17 shown.

[0420] Table 25 Twelve peptides for treating breast cancer Patient B

[0421]

[0422]

[0423] Predicted efficacy: AGP95 = 4; The probability that the PIT vaccine induces a CTL response against 4 TSAs expressed in BRC09 breast cancer cells is 95%. Other efficacy parameters: AGP50 = 6.45, mAGP = 100%, AP = 12.

[0424] To treat Patient B, the 12 peptides were formulated into 4 × 3 peptides (PBR01 / 1, PBR 01 / 2, PBR 01 / 3, PBR 01 / 4). One treatment cycle was defined as administering all 12 different peptide vaccines within 30 days ( Figure 17 C).

[0425] Patient history:

[0426] 2013: Diagnosis: Breast cancer diagnosis; CT scan and bone scan excluded metastatic disease.

[0427] 2014: Bilateral mastectomy, postoperative chemotherapy

[0428] 2016: Extensive metastatic disease with supra- and infra-diaphragmatic lymph node involvement. Multiple liver and lung metastases.

[0429] Treatment:

[0430] 2013–2014: Doxorubicin - Cyclophosphamide and Paclitaxel

[0431] 2017: Letrozole, Palbociclib, Gosorelin and PIT vaccine

[0432] 2018: The condition deteriorated and the patient died in January

[0433] The PIT vaccine treatment started on April 7, 2017. The treatment regimen and main characteristics of the disease of Patient B are shown in Table 26

[0434] Table 26 Treatment and response of Patient B

[0435]

[0436] No data

[0437] It was predicted with 95% confidence that 8 to 12 vaccine peptides would induce T - cell responses in Patient B. Peptide - specific T - cell responses were measured in all available PBMC samples using the interferon (IFN)-γ ELISPOT bioassay( Figure 18 ). The results confirmed the prediction: 9 peptides were positive in the reaction, indicating that T cells could recognize the tumor cells of Patient B, and the tumor cells expressed FISP1, BORIS, MAGE - A11, HOM - TES - 85, NY - BR - 1, MAGE - A9, SCP1, MAGE - A1 and MAGE - C2 antigens. Some tumor - specific T cells were present after the first vaccination and were boosted with additional treatments (such as MAGE - A1), and others were induced after boosting (such as MAGE - A9). This extensive tumor - specific T - cell response is remarkable in patients with advanced cancer

[0438] Medical history and outcomes of Patient B

[0439] March 7, 2017: Previous PIT vaccine treatment

[0440] Multiple metastatic diseases in the liver, with true external compression originating from the common bile duct and extensive dilation of intrahepatic bile ducts. Abdominal cavity, porta hepatis and retroperitoneal adenopathy

[0441] March 2017: Treatment initiation - Letrozole, Palbociclib, Gosorelin and PIT vaccine

[0442] May 2017: Drug interruption

[0443] May 26, 2017: After 1 PIT cycle

[0444] Tumor metabolic activity (PET CT) decreased by 83% in the liver, lung lymph nodes, and other metastases.

[0445] June 2017: As confirmed by the patient, the standardized neutrophil value showed interruption of palbociclib

[0446] 4-month delayed rebound of tumor markers

[0447] March to May 2017: CEA and CA remained elevated, consistent with the outcome of her anticancer treatment (Ban, Future Oncol 2018)

[0448] June to September 2017: CEA and CA decreased consistently with the delayed response to immunotherapy

[0449] Quality of life

[0450] February to March 2017: Poor, hospitalized due to jaundice

[0451] April to October 2017: Excellent

[0452] November 2017: (Deterioration (tumor escape?))

[0453] January 2018: Patient B died.

[0454] Summary of immunogenicity results in Figure 18 .

[0455] Clinical outcome measurement of the patient: One month before the start of PIT vaccine treatment, PET CT demonstrated extensive DFG avid disease with nodules involved both above and below the diaphragm (Table 26). She had progressive multiple liver, multifocal bone, and lung metastases and retroperitoneal adenopathy. HER liver enzymes were elevated, consistent with the damage caused by HER liver metastases with elevated bilirubin and jaundice. She received letrozole, palbociclib, and goserelin as anticancer treatment. Two months after the start of PIT vaccination, the patient felt very well and her quality of life normalized. In fact, her PET CT showed significant morphological and metabolic regression in liver, lung, bone, and lymph node metastases. No metabolic adenopathy was identified in the supra-diaphragmatic stage.

[0456] The combination of palbociclib and the individualized vaccine may be the reason for the significant early response observed after vaccination. It has been demonstrated that palbociclib improves the activity of immunotherapy by increasing the CTA presentation of HLA and reducing the proliferation of Tregs (Goelet al. Nature. 2017:471 - 475). The outcome of patient B's treatment suggests that the PIT vaccine can be used as an adjunct to existing therapies to achieve maximum efficacy.

[0457] Tracking the tumor biomarkers of Patient B to separate the effects of existing - technology therapies from those of the PIT vaccine. The tumor markers did not change during the first 2 - 3 months of treatment and then decreased sharply, indicating a delayed effect, which is typical of immunotherapy (Table 26). In addition, when the tumor biomarkers decreased, the patient had voluntarily interrupted the treatment and it was confirmed by an increase in neutrophil count.

[0458] After the 5th PIT treatment, the patient experienced symptoms. The tumor markers and liver enzyme levels increased again. 33 days after the last PIT vaccination, her PET - CT showed significant metabolic processes in the liver, peritoneum, bone, and left adrenal sites, confirming the laboratory findings. The discontinuous recurrence in distant metastases may be due to potential immune resistance; it may be caused by the down - regulation of HLA expression that impairs the recognition of tumors by PIT - induced T cells. However, the PET - CT detected a complete regression of the metabolic activity of all axillary and supra - diaphragmatic mediastinal axillary targets (Table 26). These local tumor responses can be interpreted as known delayed and durable responses to immunotherapy, because these tumor sites are unlikely to recur after the interruption of anticancer drug treatment.

[0459] Personalized immunotherapy composition for the treatment of a patient with metastatic breast cancer (Patient C)

[0460] A PIT vaccine was prepared that was designed to be similar to the described Patients A and B for the treatment of patients with metastatic breast cancer (Patient C). The PIT vaccine contained 12 PEPI. The predicted efficacy of the PIT vaccine was AGP = 4. The treatment regimen of the patient is shown in Figure 19 。

[0461] Tumor Pathology

[0462] 2011 Primary tumor: HER2 -, ER +, sentinel lymph node negative

[0463] 2017 Multiple bone metastases: ER +, cytokeratin 7 +, cytokeratin 20 -, CA125 -, TTF1 -, CDX2 -

[0464] Treatment

[0465] 2011 Wide local excision, sentinel lymph node negative; radiotherapy

[0466] 2017 - Anticancer therapy (Tx): Letrozole (2.5 mg / day), Denosumab;

[0467] Radiation (Rx): One bone

[0468] Additional PIT vaccine (3 cycles) as the standard of care

[0469] Biometric verification confirmed that positive T cell responses (defined as >5-fold higher than the control, or >3-fold higher than the control and >50 points) were detected in 11 out of 12 20-mer peptides of 12 PTI vaccines and 12 9-mer peptides (each peptide with the PEPI sequence capable of binding to the most HLA class I alleles of patient A). Figure 20 )

[0470] Persistent memory T cell responses were detected 14 months after the last vaccination ( Figure 20 C–D).

[0471] Treatment outcomes

[0472] The clinical treatment outcomes of patient C are shown in Table 35. Patient C had a partial response and signs of cured bone metastasis.

[0473] Table 27 Clinical treatment outcomes of breast cancer patient C

[0474] Before PIT +70 days * (10 weeks) +150 days * (21 weeks) +388 days * (55 weeks) Bone biopsy Metastatic breast cancer DCIS Not performed RIB5 negative Not performed PET CT Multiple metastases Only RIB5 is DFG-affin Not performed Not performed CT Multiple metastases Not performed Not performed Healed bone metastases (sclerotic foci) CA-15-3 87 50 32 24

[0475] *After 3 cycles of PIT vaccination

[0476] The immune response was as Figure 20 shown, predicted immunogenicity, PEPI = 12 (CI 95% [8,12]

[0477] Detected immunogenicity: 11 (20-mer) and 11 (9-mer) antigen-specific T cell responses after 3 PIT vaccinations ( Figure 20 A, B). Specific immune responses to the PIT vaccine could still be detected 4.5, 11, or 14 months after the last vaccination ( Figure 20 C, D).

[0478] Personalized immunotherapy composition for the treatment of a patient with metastatic colorectal cancer (Patient D)

[0479] Tumor pathology

[0480] In February 2017, a surgically induced tumor (in the sigmoid colon) with liver metastasis. pT3 pN2b (8 / 16) M1. KRAS G12D, TP53-C135Y, KDR-Q472H, MET-T1010I mutations. SATB2 expression. EGFR wt, PIK3CA-I391M (non-driver).

[0481] In June 2017, partial hepatectomy: KRAS-G12D (35G>A) NRAS wt,

[0482] In May 2018, 2 resections: SATB2 expression, lung metastases 3→21

[0483] Treatment

[0484] FOLFOX-4 (oxaliplatin, Ca-folinate, 5-FU) allergic reaction during the second treatment process in 2017

[0485] DeGramont (5-FU + Ca-folinate)

[0486] FOLFIRI plus ramucirumab in 2018 (June), every two weeks; chemoembolization

[0487] PIT vaccination (13 patient-specific peptides, 4 doses) as an adjunct to standard care in 2018 (October).

[0488] The treatment plan of the patient is shown in Figure 21 .

[0489] Treatment outcome

[0490] The patient was in good general condition, and disease progression in the lungs was confirmed by CT after 8 months.

[0491] Both PIT-induced and pre-existing T cell responses were measured by enriched fluorescent spots from PBMCs, using 9mer and 20mer peptides for stimulation ( Figure 22 ).

[0492] The summary of the immune response rate and immunogenicity results demonstrated a rational design for target antigen selection and induction of polypeptide-targeted immune responses (CD4+ and CD8+ specific responses).

[0493] Table 28 Summary table of immunological analysis of patients A–D

[0494]

[0495]

[0496] *After vaccination for 1 to 3 cycles. Sequence Listing <110> Treos Bio Zrt <120> Immunogenic cancer screening test <130> N409652WO <150> 1814361.0 <151> 2018-09-04 <160> 26 <170> PatentIn version 3.5 <210> 1 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P1 AKAP4 <400> 1 Asn Ser Leu Gln Lys Gln Leu Gln Ala Val Leu Gln Trp Ile Ala Ala 1 5 10 15 Ser Gln Phe Asn 20 <210> 2 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P2 BORIS <400> 2 Ser Gly Asp Glu Arg Ser Asp Glu Ile Val Leu Thr Val Ser Asn Ser 1 5 10 15 Asn Val Glu Glu 20 <210> 3 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P3 SPAG9 <400> 3 Val Gln Lys Glu Asp Gly Arg Val Gln Ala Phe Gly Trp Ser Leu Pro 1 5 10 15 Gln Lys Tyr Lys 20 <210> 4 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P4 OY-TES-1 <400> 4 Glu Val Glu Ser Thr Pro Met Ile Met Glu Asn Ile Gln Glu Leu Ile 1 5 10 15 Arg Ser Ala Gln 20 <210> 5 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P5 SP17 <400> 5 Ala Tyr Phe Glu Ser Leu Leu Glu Lys Arg Glu Lys Thr Asn Phe Asp 1 5 10 15 Pro Ala Glu Trp 20 <210> 6 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P6 WT1 <400> 6 Pro Ser Gln Ala Ser Ser Gly Gln Ala Arg Met Phe Pro Asn Ala Pro 1 5 10 15 Tyr Leu Pro Ser 20 <210> 7 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P7 HIWI <400> 7 Arg Arg Ser Ile Ala Gly Phe Val Ala Ser Ile Asn Glu Gly Met Thr 1 5 10 15 Arg Trp Phe Ser 20 <210> 8 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P8 PRAME <400> 8 Met Gln Asp Ile Lys Met Ile Leu Lys Met Val Gln Leu Asp Ser Ile 1 5 10 15 Glu Asp Leu Glu 20 <210> 9 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P9 AKAP-3 <400> 9 Ala Asn Ser Val Val Ser Asp Met Met Val Ser Ile Met Lys Thr Leu 1 5 10 15 Lys Ile Gln Val 20 <210> 10 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> POC01_P10 MAGE-A4 <400> 10 Arg Glu Ala Leu Ser Asn Lys Val Asp Glu Leu Ala His Phe Leu Leu 1 5 10 15 Arg Lys Tyr Arg 20 <210> 11 <211> 20 <212> PRT <213> Artificial sequence <220> <223> POC01_P11 MAGE-A9 <400> 11 Glu Thr Ser Tyr Glu Lys Val Ile Asn Tyr Leu Val Met Leu Asn Ala 1 5 10 15 Arg Glu Pro Ile 20 <210> 12 <211> 20 <212> PRT <213> Artificial sequence <220> <223> POC01_P12a MAGE-A10 <400> 12 Asp Val Lys Glu Val Asp Pro Thr Gly His Ser Phe Val Leu Val Thr 1 5 10 15 Ser Leu Gly Leu 20 <210> 13 <211> 20 <212> PRT <213> Artificial sequence <220> <223> POC01_P12b BAGE <400> 13 Ser Ala Gln Leu Leu Gln Ala Arg Leu Met Lys Glu Glu Ser Pro Val 1 5 10 15 Val Ser Trp Arg 20 <210> 14 <211> 20 <212> PRT <213> Artificial sequence <220> <223> PBRC01_cP1 FSIP1 <400> 14 Ile Ser Asp Thr Lys Asp Tyr Phe Met Ser Lys Thr Leu Gly Ile Gly 1 5 10 15 Arg Leu Lys Arg 20 <210> 15 <211> 20 <212> PRT <213> Artificial sequence <220> <223> PBRC01_cP2 SPAG9 <400> 15 Phe Asp Arg Asn Thr Glu Ser Leu Phe Glu Glu Leu Ser Ser Ala Gly 1 5 10 15 Ser Gly Leu Ile 20 <210> 16 <211> 20 <212> PRT <213> Artificial sequence <220> <223> PBRC01_cP3 AKAP4 <400> 16 Ser Gln Lys Met Asp Met Ser Asn Ile Val Leu Met Leu Ile Gln Lys 1 5 10 15 Leu Leu Asn Glu 20 <210> 17 <211> 20 <212> PRT <213> Artificial sequence <220> <223> PBRC01_cP4 BORIS <400> 317 Ser Ala Val Phe His Glu Arg Tyr Ala Leu Ile Gln His Gln Lys Thr 1 5 10 15 His Lys Asn Glu 20 <210> 18 <211> 20 <212> PRT <213> Artificial sequence <220> <223> PBRC01_cP5 MAGE-A11 <400> 18 Asp Val Lys Glu Val Asp Pro Thr Ser His Ser Tyr Val Leu Val Thr 1 5 10 15 Ser Leu Asn Leu 20 <210> 19 <211> 20 <212> PRT <213> Artificial sequence <220> <223> PBRC01_cP6 NY-SAR-35 <400> 19 Glu Asn Ala His Gly Gln Ser Leu Glu Glu Asp Ser Ala Leu Glu Ala 1 5 10 15 Leu Leu Asn Phe 20 <210> 20 <211> 20 <212> PRT <213> artificial sequence <220> <223> PBRC01_cP7 HOM-TES-85 <400> 20 Met Ala Ser Phe Arg Lys Leu Thr Leu Ser Glu Lys Val Pro Pro Asn 1 5 10 15 His Pro Ser Arg 20 <210> 21 <211> 20 <212> PRT <213> artificial sequence <220> <223> PBRC01_cP8 NY-BR-1 <400> 21 Lys Arg Ala Ser Gln Tyr Ser Gly Gln Leu Lys Val Leu Ile Ala Glu 1 5 10 15 Asn Thr Met Leu 20 <210> 22 <211> 20 <212> PRT <213> artificial sequence <220> <223> PBRC01_cP9 MAGE-A9 <400> 22 Val Asp Pro Ala Gln Leu Glu Phe Met Phe Gln Glu Ala Leu Lys Leu 1 5 10 15 Lys Val Ala Glu 20 <210> 23 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> PBRC01_cP10 SCP-1 <400> 23 Glu Tyr Glu Arg Glu Glu Thr Arg Gln Val Tyr Met Asp Leu Asn Asn 1 5 10 15 Asn Ile Glu Lys 20 <210> 24 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> PBRC01_cP11 MAGE-A1 <400> 24 Pro Glu Ile Phe Gly Lys Ala Ser Glu Ser Leu Gln Leu Val Phe Gly 1 5 10 15 Ile Asp Val Lys 20 <210> 25 <211> 20 <212> PRT <213> Artificial Sequence <220> <223> PBRC01_cP12 MAGE-C2 <400> 25 Asp Ser Glu Ser Ser Phe Thr Tyr Thr Leu Asp Glu Lys Val Ala Glu 1 5 10 15 Leu Val Glu Phe 20 <210> 26 <211> 30 <212> PRT <213> artificial sequence <220> <223> crc_P3 peptide 30aa <400> 26 Tyr Val Asp Glu Lys Ala Pro Glu Phe Ser Met Gln Gly Leu Lys Asp 1 5 10 15 Glu Lys Val Ala Glu Leu Val Arg Phe Leu Leu Arg Lys Tyr 20 25 30

Claims

1. A system for determining the risk that a human subject will develop cancer, the system comprising a computing module configured to (i) quantify HLA triplets of T cell epitopes in amino acid sequences of the subject that are capable of binding tumor-associated antigens (TAAs), wherein each HLA of the HLA triplet is capable of binding the same T cell epitope, and wherein the HLA triplet is any combination of three of the HLA class I alleles expressed by the subject; and (ii) Determine the risk that the subject will develop cancer, wherein, with respect to the TAA, a smaller number of HLA triplets of T cell epitopes capable of binding the TAA corresponds to a higher risk that the subject will develop cancer.

2. The system according to claim 1, wherein the cancer is selected from melanoma, lung cancer, renal cell carcinoma, colorectal cancer, bladder cancer, glioma, and head and neck cancer.

3. The system according to claim 1, wherein the cancer is a pediatric cancer.

4. The system according to claim 1, wherein the cancer is Kaposi's sarcoma.

5. The system according to any one of claims 1 to 4, wherein the computing module is further configured to select a treatment for cancer for the subject when the subject is determined to have an elevated risk of developing cancer.

6. The system according to claim 5, wherein the selected treatment comprises administering to the subject a peptide comprising an amino acid sequence, or a polynucleic acid or a vector encoding the peptide, the amino acid sequence (a) being a fragment of the TAA; and (b) comprising a T cell epitope capable of binding the HLA triplet of the subject.

7. The system according to claim 5 or 6, wherein the TAA is selected from (i) ACRBP (Q8NEB7.1), ACTL8 (Q9H568.1), ADAM2 (Q99965.1), ADAM29 (Q9UKF5.1), AKAP-3 (O75969.1), AKAP-4 (Q5JQC9.1), ANKRD45 (Q5TZF3.1), ARMC3 (B4DXS3.1), ARX (Q96QS3.1), BAGE-1 (Q13072.1), BAGE-2 (Q86Y30.1), BAGE-3 (Q86Y29.1), BAGE-4 (Q86Y28.1), BAGE-5 (Q86Y27.1), BRDT (Q58F21.1), C15orf60 (Q7Z4M0.1), CABYR (O75952.1), CAGE1 (Q8CT20.1), CASC5 (Q8NG31.1), CCDC110 (Q8TBZ0.1), CCDC33 (Q8N5R6.1), CCDC36 (Q8IYA8.1), CCDC62 (Q6P9F0.1), CCDC83 (Q8IWF9.1), CCNA1 (P78396.1), CDCA1 (Q9BZD4.1), CEP290 (O15078.1), CEP55 (Q53EZ4.1), COX6B2 (Q6YFQ2.1), CPXCR1 (Q8N123.1), CT45 (Q5HYN5.1), CT45A2 (Q5DJT8.1), CT45A3 (Q8NHU0.1), CT45A4 (Q8N7B7.1), CT45A5 (Q6NSH3.1), CT45A6 (P0DMU7.1), CT46 (Q86X24.1), CT47 (Q5JQC4.1), CT47B1 (P0C2P7.1), CTAGE2 (Q96RT6.1), cTAGE5 (O15320.1), CTCFL (Q8NI51.1), CTNNA2 (P26232.1), CTSP1 (A0RZH4.1), CXorf48 (Q8WUE5.1), CXorf61 (Q5H943.1), CSAG1 (Q6PB30.1), DCAF12 (Q5T6F0.1), DKKL1 (Q9UK85.1), DMRT1 (Q9Y5R6.1), DNAJB8 (Q8NHS0.1), DPPA2 (Q7Z7J5.1), DRG1 (Q9Y295.1), EDAG (Q9BXL5.1), ELOVL4 (Q9GZR5.1), FAM133A (Q8N9E0.1), FAM46D (Q8NEK8.1), FATE1 (Q969F0.1), FBXO39 (Q8N4B4.1), FMR1NB (Q8N0W7.1), FTHL17 (Q9BXU8.1), GAGE-1 (Q13065.1), GAGE12B / C / D / E (A1L429.1), GAGE12F (P0CL80.1), GAGE12G (P0CL81.1), GAGE12H (A6NDE8.1), GAGE12I (P0CL82.1), GAGE12J (A6NER3.1), GAGE-2 (Q6NT46.1), GAGE-3 (Q13067.1), GAGE-4 (Q13068.1), GAGE-5 (Q13069.1), GAGE-6 (Q13070.1), GAGE-7 (O76087.1), GAGE-8 (Q9UEU5.1), GPAT2 (Q6NUI2.1), GPATCH2 (Q9NW75.1), HAGE (Q9NXZ2.1), HOM-TES-85 (Q9P127.1), HORMAD1 (Q86X24.1), HORMAD2 (Q8N7B1.1), HSPB9 (Q9BQS6.1), IGFS11 (Q5DX21.1), IL13RA2 (Q14627.1), IMP-3 (Q9NV31.1), JARID1B (Q9UGL1.1), KIAA0100 (Q14667.1), Lage-1 (O75638.1), LDHC (P07864.1), LEMD1 (Q68G75.1), LIPI (Q6XZB0.1), LOC647107 (Q8TAI5.1), LY6K (Q17RY6.1), LYPD6B (Q8NI32.1), MAEL (Q96JY0.1), MAGE-A1 (P43355.1), MAGE-A10 (P43363.1), MAGE-A11 (P43364.1), MAGE-A12 (P43365.1), MAGE-A2 (P43356.1), MAGE-A2B (Q6P448.1), MAGE-A3 (P43357.1), MAGE-A4 (P43358.1), MAGE-A5 (P43359.1), MAGE-A6 (P43360.1), MAGE-A8 (P43361.1), MAGE-A9 (P43362.1), MAGE-B1 (P43366.1), MAGE-B2 (O15479.1), MAGE-B3 (O15480.1), MAGE-B4 (O15481.1), MAGE-B5 (Q9BZ81.1), MAGE-B6 (Q8N7X4.1), MAGE-C1 (O60732.1), MAGE-C2 (Q9UBF1.1), MAGE-C3 (Q8TD91.1), MCAK (Q99661.1), MORC1 (Q86VD1.1), MPHOSPH1 (Q96Q89.1), NA88-A (P0C5K6.1), NLRP4 (Q96MN2.1), NOL4 (O94818.1), NR6A1 (Q15406.1), NXF2 (Q9GZY0.1), NXF2B (Q5JRM6.1), NY-ESO-1 (P78358.1), ODF1 ((Q14990.1), ODF2 (Q5BJF6.1), ODF3 (Q96PU9.1), ODF4 (Q2M2E3.1), OIP5 (O43482.1), OTOA (Q05BM7.1), PAGE1 (O75459.1), PAGE2 (Q7Z2X2.1), PAGE2B (Q5JRK9.1), PAGE3 (Q5JUK9.1), PAGE4 (O60829.1), PAGE5 (Q96GU1.1), PASD1 (Q8IV76.1), PBK (Q96KB5.1), PEPP2 (Q9HAU0.1), PIWIL1 (Q96J94.1), PIWIL2 (Q8TC59.1), PLAC1 (Q9HBJ0.1), POTEA (Q6S8J7.1), POTEB (Q6S5H4.1), POTEC (B2RU33.1), POTED (Q86YR6.1), POTEE (Q6S8J3.1), POTEG (Q6S5H5.1), POTEH (Q6S545.1), PRAME (P78395.1), PRM1 (P04553.1), PRM2 (P04554.1), PRSS54 (Q6PEW0.1), PRSS55 (Q6UWB4.1), PTPN20A (Q4JDL3.1), RBM46 (Q8TBY0.1), RGS22 (Q8NE09.1), ROPN1A (Q9HAT0.1), RQCD1 (Q92600.1), SAGE1 (Q9NXZ1.1), SEMG1 (P04279.1), SLCO6A1 (Q86UG4.1), SPA17 (Q15506.1), SPACA3 (Q8IXA5.1), SPAG1 (Q07617.1), SPAG17 (Q6Q759.1), SPAG4 (Q9NPE6.1), SPAG6 (O75602.1), SPAG8 (Q99932.1), SPAG9 (O60271.1), SPANXA1 (Q9NS26.1), SPANXB (Q9NS25.1), SPANXC (Q9NY87.1), SPANXD (Q9BXN6.1), SPANXE (Q8TAD1.1), SPANXN1 (Q5VSR9.1), SPANXN2 (Q5MJ10.1), SPANXN3 (Q5MJ09.1), SPANXN4 (Q5MJ08.1), SPANXN5 (Q5MJ07.1), SPATA19 (Q7Z5L4.1), SPEF2 (Q9C093.1), SPINLW1 (O95925.1), SPO11 (Q9Y5K1.1), SSX-1 (Q16384.1), SSX-2 (Q16385.1), SSX-3 (Q99909.1), SSX-4 (O60224.1), SSX-5 (O60225.1), SSX-6 (Q7RTT6.1), SSX-7 (Q7RTT5.1), SSX-9 (Q7RTT3.1), SYCE1 (Q8N0S2.1), SYCP1 (Q15431.1), TAF7L (Q5H9L4.1), TAG-1 (Q02246.1), TDRD1 (Q9BXT4.1), TDRD4 (Q9BXT8.1), TDRD6 (O60522.1), TEKT5 (Q96M29.1), TEX101 (Q9BY14.1), TEX14 (Q8IWB6.1), TEX15 (Q9BXT5.1), TEX38 (Q6PEX7.1), TFDP3 (Q5H9I0.1), THEG (Q9P2T0.1), TMEFF1 (Q8IYR6.1), TMEFF2 (Q9UIK5.1), TMEM108 (Q6UXF1.1), TMPRSS12 (Q86WS5.1), TPPP2 (P59282.1), TPTE (P56180.1), TRAG-3 (Q9Y5P2.1), TSGA10 (Q9BZW7.1), TSP50 (Q9UI38.1), TSPY1 (Q01534.1), TSPY2 (A6NKD2.1), TSPY3 (Q6B019.1), TSSK6 (Q9BXA6.1), TTK (P33981.1), TULP2 (O00295.1), XAGE-1 (Q9HD64.1), XAGE-2 (Q96GT9.1), XAGE-3 (Q8WTP9.1), XAGE-4 (Q8WWM0.1), XAGE-5 (Q8WWM1.1), ZNF165 (P49910.1) and ZNF645 (Q8N7E2.1); or. (ii) SPAG9, AKAP4, BORIS, Survivin, MAGE-A1, PRAME, CT45, NY-SAR-35, FSIP1, HOM-TES-85, NY-BR-1, MAGE-A9, SCP-1, MAGE-A12, MAGE-A10, GATA-3, GAGE-7, SSX-4, SPANXC, CT46, MAGE-A3, MAGE-C2, TSP50, EpCAM, CAGE, MAGE-A8, FBXO39, PAGE-4, MAGE-A6, BAGE-4, MAGE-C1, NY-ESO-1, MAGE-A2, XAGE-1, MAGE-A11, SSX-2, LAGE-1, MAGE-A4, MAGE-A5, MAGE-B2, MAGE-B1, HAGE, SSX-1, NXF2, SAGE, LEMD1, and OY-TES-1.

8. The system according to any one of claims 1 to 7, wherein the system comprises: (a) a storage module configured to store data including the HLA class I genotype of the subject and the amino acid sequences of the TAAs; (b) A calculation module configured to quantify the HLA triplets of the T cell epitopes in the amino acid sequences of the subject that can bind to the TAA, wherein each HLA of the HLA triplet can bind to the same T cell epitope, and wherein the HLA triplet is any combination of three of the HLA class I alleles expressed by the subject; and (c) An output module configured to display an indication of the risk that the subject will develop cancer and / or a recommended treatment for the subject.

Images