Composition for diagnosing or determining prognosis of papillary thyroid cancer and method for providing information for diagnosing or determining prognosis of papillary thyroid cancer

Abstract

The present invention relates to a composition for cancer diagnosis for predicting the prognosis of papillary thyroid cancer and providing criteria for classifying subtypes thereof. In addition, the present invention relates to a method for providing cancer diagnosis information using the composition. According to the present invention, the accuracy of papillary thyroid cancer prediction can be improved on the basis of a specific biomarker. In addition, thyroid cancer types can be divided into two subgroups to contribute to prognosis prediction and treatment strategies.

Description

Composition for diagnosing or determining the prognosis of papillary thyroid cancer and method for providing information for diagnosing or determining the prognosis of papillary thyroid cancer

[0001] The present invention relates to a composition for determining the prognosis of thyroid cancer, which is used to predict the prognosis of papillary thyroid cancer and to provide criteria for classifying its subtypes.

[0002] In addition, the present invention relates to a method for providing information for the diagnosis or prognosis prediction of thyroid cancer using the above composition.

[0003] Traditionally, ultrasound examinations and certain molecular markers have been primarily utilized for the diagnosis and prognosis prediction of thyroid cancer. However, these methods have limitations in that they fail to adequately reflect tumor heterogeneity or subtle molecular and biological changes. In particular, existing molecular-based prognosis prediction technologies have been developed mainly for cancer types with poor prognosis, which limits their ability to provide high predictive accuracy for thyroid cancer, which generally has a good prognosis. Consequently, it is difficult to precisely evaluate individual patient tumor characteristics to determine the optimal timing and scope of surgery.

[0004] Papillary thyroid carcinoma (PTC) is the most common form of thyroid cancer and is generally known to have a good prognosis. However, some patients may exhibit local invasion, lymph node metastasis, recurrence, or progression to more aggressive subtypes, requiring a detailed risk assessment when establishing treatment strategies. In particular, determining whether to perform surgery immediately or apply active surveillance at the initial diagnosis stage is a clinically important issue, but current cytological examination results alone have limitations in accurately predicting tumor aggressiveness.

[0005] Accordingly, there is a continuously growing need for new biomarkers capable of reflecting molecular characteristics closely associated with the prognosis of thyroid cancer. In particular, research is required on indicators that can be measured using samples obtained through minimally invasive methods, such as FNAB, while precisely reflecting differences in tumor biological characteristics and prognosis. From this perspective, there is a need to develop a new system capable of precisely evaluating the diagnosis or prognosis of thyroid cancer based on more stable and highly prognostic molecular indicators, such as epigenetic changes as well as changes in gene expression.

[0006]

[0007] The objective of the present invention is to provide a composition for cancer diagnosis or prognosis determination for predicting the prognosis of papillary thyroid cancer.

[0008] Another objective of the present invention is to provide a method for providing information for cancer diagnosis or prognosis determination using the above composition.

[0009] However, the technical problems that the present invention aims to solve are not limited to those mentioned above, and other unmentioned problems will be clearly understood by those skilled in the art from the description below.

[0010]

[0011] Various embodiments of the present invention are described with reference to the drawings. In the following description, for a complete understanding of the present invention, various specific details, such as specific forms, compositions, and processes, are described. However, specific embodiments may be practiced without one or more of these specific details, or in combination with other known methods and forms. In other examples, known processes and manufacturing techniques are not described as specific details so as not to make the present invention unnecessary or obscure. Reference throughout this specification to one embodiment implies that a particular feature, form, composition, or characteristic described in association with the embodiment is included in one or more embodiments of the present invention. Accordingly, the circumstances of an embodiment expressed at various locations throughout this specification do not necessarily represent the same embodiment of the present invention. Additionally, a particular feature, form, composition, or characteristic may be combined in any suitable way in one or more embodiments.

[0012]

[0013] The present invention provides a minimally invasive diagnostic tool that can contribute to the decision regarding whether to pursue aggressive treatment, such as surgery, for thyroid cancer. For example, the present invention can predict diagnosis or prognosis by evaluating the aggressiveness of thyroid cancer based on epigenetic biomarkers. In the present invention, the diagnosis of thyroid cancer includes determining the stage of the thyroid cancer, and the prognosis of the thyroid cancer includes the survival rate, metastasis rate, and recurrence rate.

[0014] According to one embodiment, regions of CpG islands with altered methylation levels in thyroid cancer tissue can be identified by comprehensively analyzing methylomics data. Candidate regions identified by this method can be further validated from a cohort of patients with papillary thyroid cancer (PTC) to develop methylation-specific primers suitable for fine needle aspiration biopsy (FNAB) samples. These primers refer to short nucleic acid sequences having a short free 3'-terminal OH group, capable of forming base pairs with a complementary template, and functioning as starting points for template strand replication.

[0015] In the present invention, the primer has 12 or more consecutive nucleotides that hybridize to a bisulfite-converted sequence or a sequence complementary thereof.

[0016] In addition, the present invention can effectively distinguish subgroups of thyroid cancer by precisely evaluating the DNA methylation level of a specific biomarker.

[0017] The biomarker composition for diagnosing papillary thyroid cancer according to the present invention comprises, as an active ingredient, one or more genes selected from AGAP2, EHBP1L1, PRDM8, and CD37, or proteins encoded by one or more of these genes. In particular, since distinct methylation changes are observed in AGAP2, the biomarker composition for diagnosing papillary thyroid cancer according to the present invention may include AGAP2 as a representative candidate biomarker. In one embodiment, EHBP1L1 and PRDM8 exhibit consistent changes in methylation in two or more cell lines. In one embodiment, CD37 significantly enhances the characteristics that promote tumor formation and progression in terms of proliferation, invasion, wound healing, and colony formation when overexpressed.

[0018] In addition, the composition of the present invention may further include one or more genes among RIN3, CARMIL2, and GPR84, or a protein encoded by one or more genes as an active ingredient.

[0019] According to one embodiment, the composition of the present invention may include as an active ingredient a protein encoded by the genes AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2 and GPR84 or one or more of the genes.

[0020] RIN3 and AGAP2 play important roles in the Ras pathway and are critically involved in the progression of thyroid cancer. RIN3 appears to be involved in the Ras signaling pathway as it contains a Ras-associated domain, while AGAP2 exerts an anti-apoptotic effect by activating phosphoinositide 3-kinase (PI3K), a process frequently initiated by Ras activation. AGAP2, EHBP1L1, and CARMIL2 play important roles in cancer metastasis by contributing to cell migration and invasion. On the other hand, CD37 and GPR84 are associated with the immune system, which is consistent with the immune-related transcriptome characteristics of the first subgroup of papillary thyroid cancer (PTC1), suggesting a potential role in regulating immune responses within thyroid cancer.

[0021] The seven biomarkers according to the present invention—AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84—represent an area that has not yet been sufficiently explored in thyroid cancer research. Their discovery deepens the understanding of the molecular complexity of thyroid cancer (THCA) and demonstrates their potential as therapeutic targets or diagnostic markers when combined with new technologies. In particular, their diverse signaling pathways, ranging from their interactions with the immune system, provide new insights into thyroid cancer research.

[0022]

[0023] According to another embodiment of the present invention, a method for providing information for diagnosing papillary thyroid cancer is provided, comprising: S1) measuring the methylation level of one or more genes among AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, GPR84 or proteins encoded by one or more genes from a biological sample of papillary thyroid cancer; and S2) comparing the methylation level measured in step S1) with the methylation level of the same genes or proteins encoded by said genes measured in a biological sample of normal tissue.

[0024] According to one embodiment, the biological sample of the present invention may be thyroid cancer tissue or adjacent normal tissue. The biological sample may be obtained from tissue excised by surgical methods, fine needle aspiration biopsy (FNAB), etc., but is not limited thereto. Furthermore, the biological sample may be one or more selected from the group consisting of cells, tissues, biopsies, paraffin tissues, blood, serum, plasma, fine needle aspiration specimens, urine, and combinations thereof derived from the target individual, but is not limited thereto. Additionally, the biological sample may be derived from the target individual.

[0025] According to one embodiment, in the present invention, if the methylation level measured according to step S1) is higher than that measured according to step S2), it can be diagnosed as papillary thyroid cancer.

[0026] According to one embodiment, the present invention can distinguish patients into two prognostic groups based on the methylation pattern of a biological sample, one of which may be a group with a relatively low survival rate.

[0027] Specifically, the present invention may further include S3) a step of calculating a sum of methylation signals of one or more genes among AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, GPR84 or proteins encoded by one or more genes; and S4) a step of distinguishing a first subgroup (PTC1) and a first subgroup (PTC2) of papillary thyroid cancer from the sum of methylation signals.

[0028] According to one embodiment, the sum of the methylation signals in step S3) may be 0.4 or more for PTC1 and less than 0.4 for PTC2.

[0029] According to another embodiment of the present invention, a kit for diagnosing papillary thyroid cancer comprising the biomarker composition may be provided. The kit of the present invention may further include a sample necessary for diagnosing thyroid cancer. Specifically, for example, the kit may include a solid support, a substrate for immunological detection of an antibody or antigen-binding fragment, a suitable buffer solution, a chromogenic enzyme, a secondary antibody labeled with a fluorescent substance, or a chromogenic substrate. Additionally, for nucleic acid detection, it may include a polymerase, a buffer, a nucleic acid, a coenzyme, a fluorescent substance, or a combination thereof. The polymerase may be, for example, Taq polymerase.

[0030] In addition, the term "diagnostic kit" in the present invention refers to a kit comprising a composition and components necessary for analysis, which is used for the diagnosis, prognosis prediction, or classification of cancer by analyzing the methylation level of a methylation region containing a CpG island included in any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84. The diagnostic kit targets genomic DNA isolated from cells, tissues, biopsy samples, blood, serum, plasma, urine, or combinations thereof derived from a patient suspected of having cancer or a subject for diagnosis, and is configured to detect whether the genes are methylated after being treated with a reagent that modifies methylated DNA and non-methylated DNA differently. To this end, the kit may include, as active ingredients, one or more of the following: a primer pair capable of amplifying or detecting fragments containing methylated CpG islands; a probe capable of hybridizing with methylated CpG island sites; a methylation-specific binding protein; a methylation-specific binding antibody or aptamer; a methylation-sensitive restriction endonuclease; and a sequencing primer. Additionally, the diagnostic kit may include a pretreatment agent using a reagent selected from the group consisting of bisulfite, hydrogen sulfite, disulfite, or combinations thereof, and components for performing one or more of the following methods: PCR, methylation-specific PCR, real-time methylation-specific PCR, quantitative PCR, nucleic acid chip, sequencing, sequencing biosynthesis, or sequencing bioligation. Accordingly, the diagnostic kit of the present invention can be usefully employed for the diagnosis or prognosis prediction of various cancers, including thyroid cancer, using the methylation patterns of the genes as indicators.

[0031] In the present invention, the term “AGAP2” refers to the ArfGAP with GTPase domain, ankyrin repeat, and PH domain 2 gene, which is a gene located on human chromosome 12 (chr12) and includes an epigenetic biomarker in which changes in DNA methylation levels, gene expression, and chromatin accessibility are observed integrally in papillary thyroid cancer. It has been reported that significant changes in methylation of specific CpG island regions are observed in AGAP2 in subtypes of papillary thyroid cancer with poor prognosis, and that such methylation changes are associated with changes in gene expression and chromatin accessibility. Therefore, in the present invention, AGAP2 is a concept that includes a gene that can be utilized as a DNA methylation-based prognostic prediction indicator related to the aggressiveness, prognostic differences, or high-risk group classification of papillary thyroid cancer.

[0032] In the present invention, “chr12:58132478-58132734” of AGAP2 refers to the genomic region from position 58132478 to position 58132734 of chromosome 12 (chr12), “chr12:58130870-58132047” of AGAP2 refers to the genomic region from position 58130870 to position 58132047 of chromosome 12 (chr12), and “chr12:58119909-58121551” of AGAP2 refers to the genomic region from position 58119909 to position 58121551 of chromosome 12 (chr12). In addition, the beta value of each CpG location is calculated based on coordinates defined in the human reference genome (NCBI GRCh37; hg19).

[0033] In the present invention, the term “EHBP1L1” ​​refers to the EH domain binding protein 1-like 1 gene, which is a gene located on human chromosome 11 (chr11) and includes an epigenetic biomarker in which changes in DNA methylation levels, gene expression, and chromatin accessibility are observed integrally in papillary thyroid cancer. It has been confirmed that EHBP1L1 shows significant differences in DNA methylation levels in specific CpG island regions when compared among subgroups of papillary thyroid cancer, and that such methylation changes are associated with increased RNA expression levels and increased chromatin accessibility. Therefore, in the present invention, EHBP1L1 is a concept that includes a gene that can be utilized as a DNA methylation-based prognostic prediction indicator associated with prognostic differences or aggressiveness in papillary thyroid cancer.

[0034] In the present invention, “chr11:65352231-65353134” of EHBP1L1 refers to the genomic region from position 65352231 to position 65353134 of chromosome 11 (chr11), and “chr11:65359292-65360328” of EHBP1L1 refers to the genomic region from position 65359292 to position 65360328 of chromosome 11 (chr11). In addition, the beta value of each CpG position is calculated based on coordinates defined based on the human reference genome (NCBI GRCh37; hg19).

[0035] In the present invention, the term “PRDM8” refers to the PR / SET domain 8 gene, which is a gene located on a human chromosome and includes an epigenetic biomarker in which changes in DNA methylation levels, gene expression, and chromatin accessibility are observed integrally in papillary thyroid cancer. In comparisons between subgroups of papillary thyroid cancer, the DNA methylation levels of specific CpG island regions of PRDM8 showed significant differences, and it was confirmed that these methylation changes were associated with increased RNA expression levels and increased chromatin accessibility. Furthermore, PRDM8 was selected as a gene in which changes in chromatin accessibility were observed after treatment with a demethylating agent, and it includes a regulatory region that responds sensitively to changes in DNA methylation status. Therefore, in the present invention, PRDM8 is a concept that includes a gene that can be utilized as a DNA methylation-based prognostic indicator associated with differences in prognosis or aggressiveness in papillary thyroid cancer.

[0036] In the present invention, “chr4:81128229-81128691” of PRDM8 refers to the genomic region from position 81128229 to position 81128691 of chromosome 4 (chr4), and the beta value of each CpG position is calculated based on coordinates defined based on the human reference genome (NCBI GRCh37; hg19).

[0037] In the present invention, the term “CD37” refers to the cluster of differentiation 37 gene, which is a gene encoding a protein primarily expressed on the surface of immune cells, and includes an epigenetic biomarker in which changes in DNA methylation levels, gene expression, and chromatin accessibility are observed integrally in papillary thyroid cancer. It was confirmed that CD37 shows significant differences in DNA methylation levels in specific CpG islands or adjacent genomic regions during comparisons between subgroups of papillary thyroid cancer, and that such methylation changes are associated with increased RNA expression levels and changes in chromatin accessibility. Furthermore, CD37 is a gene in which phenomena associated with tumorigenic characteristics were observed through functional verification experiments in papillary thyroid cancer cell lines, and the concept includes a gene that can be utilized as a DNA methylation-based prognostic prediction indicator related to prognostic differences or aggressiveness in papillary thyroid cancer.

[0038] In the present invention, “chr19:49841187-49841628” of CD37 refers to the genomic region from position 49841187 to position 49841628 of chromosome 19 (chr19), and “chr19:49842654-49843628” of CD37 refers to the genomic region from position 49842654 to position 49843628 of chromosome 19 (chr19). In addition, the beta value of each CpG position is calculated based on coordinates defined based on the human reference genome (NCBI GRCh37; hg19).

[0039] In the present invention, the term “RIN3” refers to the Ras and Rab interactor 3 gene, which is a gene encoding a protein associated with the Ras-related signaling pathway and includes an epigenetic biomarker in which changes in DNA methylation levels, gene expression, and chromatin accessibility are observed integrally in papillary thyroid cancer. It has been confirmed that RIN3 shows significant differences in DNA methylation levels in specific CpG islands or adjacent genomic regions in comparisons between subgroups of papillary thyroid cancer, and that such methylation changes are associated with increased RNA expression levels and changes in chromatin accessibility. Therefore, in the present invention, RIN3 is a concept that includes a gene that can be utilized as a DNA methylation-based prognostic indicator related to prognostic differences or aggressiveness in papillary thyroid cancer.

[0040] In the present invention, “chr14:93153278-93154759” of RIN3 refers to the genomic region from position 93153278 to position 93154759 of chromosome 14 (chr14), and the beta value of each CpG position is calculated based on coordinates defined based on the human reference genome (NCBI GRCh37; hg19).

[0041] In the present invention, the term “CARMIL2” refers to the capping protein regulator and myosin 1 linker 2 gene, which encodes a protein involved in the regulation of the actin cytoskeleton and cell migration, and includes an epigenetic biomarker in which changes in DNA methylation levels, gene expression, and chromatin accessibility are observed integrally in papillary thyroid cancer. It was confirmed that CARMIL2 shows significant differences in DNA methylation levels in specific CpG islands or adjacent genomic regions in comparisons between subgroups of papillary thyroid cancer, and that such methylation changes are associated with increased RNA expression levels and changes in chromatin accessibility. Therefore, in the present invention, CARMIL2 is a concept that includes a gene that can be utilized as a DNA methylation-based prognostic indicator reflecting prognostic differences related to the invasiveness, mobility, or aggressiveness of papillary thyroid cancer.

[0042] In the present invention, “chr16:67686860-67687674” of CARMIL2 refers to the genomic region from position 67686860 to position 67687674 of chromosome 16 (chr16), and “chr16:67681975-67683924” of CARMIL2 refers to the genomic region from position 67681975 to position 67683924 of chromosome 16 (chr16). In addition, the beta value of each CpG position is calculated based on coordinates defined based on the human reference genome (NCBI GRCh37; hg19).

[0043] In the present invention, the term “GPR84” refers to the G protein-coupled receptor 84 gene. As a gene encoding a G protein-coupled receptor associated with an immune response, it includes an epigenetic biomarker in which changes in DNA methylation levels, gene expression, and chromatin accessibility are observed integrally in papillary thyroid cancer. In comparisons between subgroups of papillary thyroid cancer, GPR84 was found to show significant differences in DNA methylation levels in specific CpG islands or adjacent genomic regions, and these methylation changes were identified as being associated with increased RNA expression levels and changes in chromatin accessibility. Furthermore, GPR84 was selected as a gene in which changes in chromatin accessibility and gene expression were observed after treatment with a demethylating agent, and it includes a regulatory region that responds sensitively to changes in DNA methylation status. Therefore, in the present invention, GPR84 is a concept that includes a gene capable of being utilized as a DNA methylation-based prognostic indicator associated with prognostic differences or aggressiveness in papillary thyroid cancer.

[0044] In the present invention, “chr12:54764065-54764510” of GRP84 refers to the genomic region from position 54764065 to position 54764510 of chromosome 12 (chr12), and the beta value of each CpG position is calculated based on coordinates defined based on the human reference genome (NCBI GRCh37; hg19).

[0045] In the present invention, the term “agent for analyzing methylation levels” refers to a agent used to qualitatively or quantitatively detect, compare, or evaluate the DNA methylation status or degree of methylation in a specific CpG island or adjacent genomic region of DNA derived from a biological sample. The agent for analyzing methylation levels may include, but is not limited to, one or more selected from the group consisting of a primer pair capable of amplifying a fragment containing a methylated CpG island site, a probe capable of hybridizing with a methylated CpG island site, a methylation-specific binding protein capable of binding with a methylated CpG island site, a methylation-specific binding antibody or aptamer, a methylation-sensitive restriction endonuclases, sequencing primers, sequencing by synthesis primers, and sequencing by ligation primers.

[0046] In the present invention, the term “primer” refers to a short nucleic acid sequence having a short free 3’ end hydroxyl group, capable of forming base pairs with a complementary template, and functioning as a starting point for template strand replication. The primer can initiate DNA synthesis in the presence of a reagent for a polymerization reaction (i.e., DNA polymerase or reverse transcriptase) and four different nucleoside triphosphates at an appropriate buffer solution and temperature. Additionally, the primer may incorporate additional features that do not alter the basic properties of the primer acting as a starting point for DNA synthesis, such as sense and antisense nucleic acids having sequences of 7 to 50 nucleotides.

[0047] In addition, the primers of the present invention may be designed according to the sequence of a specific CpG site to be analyzed for methylation status, and preferably, they may be one or more selected from the group consisting of a primer pair capable of specifically amplifying cytosine that is methylated and not modified by bisulfite treatment, a primer pair capable of specifically amplifying cytosine that is not methylated and modified by bisulfite treatment, a primer pair capable of specifically amplifying cytosine that is methylated and modified by TET-family proteins, and a primer pair capable of specifically amplifying cytosine that is not methylated and not modified by TET-family proteins.

[0048] In the present invention, the term “methylation” refers to a chemical modification in which a methyl group is covalently bonded to a cytosine residue among the bases of nucleic acids, particularly DNA, to form 5-methylcytosine. The methylation occurs mainly within CpG dinucleotide sequences and includes epigenetic modifications involved in gene expression regulation, chromatin structure changes, and cell function regulation without changes in the genomic base sequence.

[0049] In the present invention, the term “genomic DNA” means DNA comprising the entire genome sequence isolated from a biological sample selected from the group consisting of cells, tissues, biopsy samples, paraffin-embedded tissues, blood, serum, plasma, fine-needle aspiration samples, urine, and combinations thereof derived from a patient suspected of having cancer or a subject of diagnosis. The genomic DNA comprises nuclear chromosomal DNA and may additionally include mitochondrial DNA as needed.

[0050] In the present invention, "cancer" refers to or indicates a physiological state characterized by uncontrolled cell growth typically found in mammals. The cancer subject to prevention, improvement, or treatment in the present invention may be a solid tumor consisting of a mass formed by abnormal cell growth in a solid organ, and depending on the location of the solid organ, it may be gastric cancer, liver cancer, glioblastoma, ovarian cancer, colorectal cancer, head and neck cancer, bladder cancer, renal cell carcinoma, breast cancer, metastatic cancer, prostate cancer, pancreatic cancer, melanoma, or lung cancer, but is not limited thereto. The cancer may include, but is not limited to, thyroid cancer, breast cancer, pancreatic cancer, gastric cancer, blood cancer, lung cancer, liver cancer, colorectal cancer, rectal cancer, esophageal cancer, biliary tract cancer, kidney cancer, bladder cancer, prostate cancer, ovarian cancer, uterine cancer, cervical cancer, head and neck cancer, skin cancer, and brain tumor.

[0051] In the present invention, the term “papillary thyroid carcinoma (PTC)” refers to the most common form of thyroid cancer, which is a malignant tumor originating from thyroid follicular epithelial cells and characterized by a papillary structure upon microscopic observation. Although papillary thyroid carcinoma is generally known to have a relatively good prognosis, some patients may exhibit increased tumor invasiveness, lymph node metastasis, recurrence, or a more aggressive clinical course.

[0052] In the present invention, the term “CpG island” refers to a region on genomic DNA where the frequency of CpG dinucleotides, in which cytosine and guanine are arranged continuously through phosphate bonds, is relatively high. The CpG island generally has a high GC content, may be located near a gene promoter or a transcription start site, and may be associated with the regulation of gene expression depending on the DNA methylation status. Furthermore, the CpG island may include forms in which CpG dinucleotides are distributed continuously or discontinuously with other base sequences interposed between the CpG dinucleotides. In the present invention, a CpG island is a genomic region comprising one or more CpG sites, and one or more CpG sites included in the entire or part thereof of the CpG island can be used as targets for methylation analysis and can be specifically identified by genomic coordinates defined based on the human reference genome, specific CpG locations, or corresponding sequences.

[0053] Specifically, in the present invention, the term “methylated CpG island” means a state in which a methyl group is added to one or more cytosine residues within the CpG island, and the term “non-methylated CpG island” means a state in which a methyl group is not added to the cytosine residues within the CpG island.

[0054] In the present invention, the term “methylation level of a CpG island” means the ratio, frequency, or relative degree of cytosine methylated in a specific CpG island or part thereof, which can be measured, compared, or evaluated qualitatively or quantitatively.

[0055] In the present invention, "diagnostic nucleic acid chip" refers to an analytical means comprising a substrate on which a probe designed to hybridize with a fragment containing a methylated CpG island included in one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 is immobilized. The diagnostic nucleic acid chip is used to detect or analyze the methylation of CpG islands in the corresponding genes after treating genomic DNA isolated from a biological sample derived from a patient suspected of having cancer or a subject for diagnosis with a reagent that modifies methylated DNA and unmethylated DNA differently. This configuration enables the simultaneous analysis of the methylation patterns of specific genes. The substrate used in the nucleic acid chip may be made of glass, silicon, plastic, or an equivalent material, and the probe may be immobilized in an array form to specifically bind to a nucleic acid sequence containing the methylated CpG island. Such an array structure enables the parallel analysis of multiple genes or multiple methylation sites. Therefore, the diagnostic nucleic acid chip of the present invention can be usefully utilized for the diagnosis of cancer by using the CpG island methylation information of the genes as an indicator.

[0056] In the present invention, the terms “cancer diagnostic device” or “cancer prognosis prediction device” refer to a device configured to provide information for diagnosing cancer, assessing the prognosis or risk of cancer by analyzing DNA isolated from a biological sample. The cancer diagnostic device or cancer prognosis prediction device is configured to detect DNA methylation levels for one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84, normalize the detected methylation levels to calculate a score, and then integrate the scores to classify patients into a high-risk group or a low-risk group.

[0057] The above detection unit refers to a module configured to qualitatively or quantitatively detect the DNA methylation level of CpG island sites corresponding to one or more of the genes AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 from DNA isolated from a biological sample.

[0058] The above calculation unit refers to a module configured to normalize the DNA methylation levels of each gene detected by the above detection unit to calculate individual methylation scores, and to integrate the calculated scores to derive a comprehensive methylation score at the patient level.

[0059] The evaluation unit above refers to a module configured to classify patients into a high-risk group if the comprehensive methylation score derived from the calculation unit above is 0.4 or higher, and into a low-risk group if the comprehensive methylation score is less than 0.4, by comparing the comprehensive methylation score with a preset reference value.

[0060] In addition, the aforementioned high-risk group is configured to be evaluated as having a lower survival rate and a higher likelihood of undergoing total thyroidectomy compared to the low-risk group.

[0061] Specifically, the cancer diagnostic device may be implemented as a single device or in a form where multiple modules are logically or physically combined, and may include a processor and a storage medium to perform the detection, calculation, and evaluation functions automatically or semi-automatically.

[0062] In the present invention, “high-risk group” refers to a group of patients in which the comprehensive methylation score, calculated by normalizing and integrating the DNA methylation levels of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84, is greater than or equal to a preset threshold value. The high-risk group is evaluated as a patient group that is more likely to show a poor prognosis compared to the low-risk group and requires consideration of more aggressive surgical strategies, such as total thyroidectomy.

[0063] In the present invention, “low-risk group” refers to a group of patients whose comprehensive methylation score is below a preset threshold value. The low-risk group is likely to show a relatively favorable prognosis compared to the high-risk group and may be evaluated as a patient group for which limited surgery or active surveillance strategies may be considered depending on the clinical situation.

[0064] In the present invention, the term “reference value” refers to a classification reference value set based on the distribution of comprehensive methylation scores calculated from a plurality of papillary thyroid cancer patient cohorts. The reference value is a value according to one embodiment and may be set to a different value depending on the analysis method, the combination of genes used, the patient group, or the statistical method.

[0065] According to one embodiment of the present invention, the reference value is a value set based on the distribution of comprehensive normalized methylation scores measured in surgical tissues of a plurality of papillary thyroid cancer patients, and can function as a classification reference value to effectively distinguish between molecular subgroups PTC1 (high-risk group) and PTC2 (low-risk group) at 0.4 (Fig. 38).

[0066] In addition, the above reference value may be 0.1 to 0.5, 0.2 to 0.45, or 0.3 to 0.4.

[0067] According to another embodiment of the present invention, the reference value may be set to 0 or a value near it in the qMSP results of fine needle aspiration biopsy (FNAB) samples, and the normalized methylation score calculated through qMSP in the FNAB samples shows a statistically significant difference between PTC1 and PTC2, and in particular, it is confirmed that the two subgroups are clearly separated based on 0 or a value near it (Fig. 40). This supports the validity of the methylation-based classification criteria of the present invention even in minimally invasive FNAB samples. The value near 0 may be -0.5 to +0.5, -0.45 to +0.45, -0.3 to +0.3, or -0.25 to +0.25.

[0068] In the present invention, "prevention" or "diagnosis" refers to any act that obtains information regarding the occurrence, likelihood of occurrence, progression status, or prognosis of cancer, or enables clinical judgment based thereon. Specifically, "prevention" includes pre-evaluating the risk of cancer development, the likelihood of recurrence, or whether a subject is in a high-risk group by analyzing the methylation levels of specific genes or CpG islands for subjects who have not yet been confirmed to have cancer, and utilizing the results for follow-up observation, additional testing, lifestyle management, or clinical decision-making. Furthermore, "diagnosis" includes detecting and analyzing the methylation levels of specific genes in DNA isolated from biological samples to determine or classify the presence, type, pathological characteristics, stage of progression, or prognostic risk of cancer, and is understood as a concept that encompasses not only definitive diagnosis but also auxiliary diagnosis, differential diagnosis, early diagnosis, prognosis prediction, and recurrence monitoring. "Prevention or diagnosis" in the present invention may be performed alone or in conjunction with imaging diagnosis, histological examination, hematological examination, or clinical information.

[0069] As used in this specification, “improvement or treatment” refers to any medical or clinical intervention aimed at changing a clinical condition related to cancer, biological indicators, or the patient’s prognosis in a desirable direction. Specifically, “improvement” includes changes in condition that contribute to reducing the rate of tumor progression, mitigating malignancy, reducing the risk of recurrence, lowering prognostic risk, improving the potential for increased survival, or optimizing treatment strategies such as surgery or chemotherapy, regardless of whether the cancer itself is completely eradicated. Furthermore, “treatment” includes all actions performed for the purpose of inhibiting cancer growth, delaying progression, shrinking or removing lesions, extending survival time, or reducing the clinical burden caused by cancer; it is a concept that encompasses not only curative treatment but also adjuvant therapy, combination therapy, maintenance therapy, and the establishment of prognosis-based treatment strategies. Accordingly, “improvement or treatment” in the present invention includes cases where treatment guidelines are determined based on the results of analyzing the methylation levels of specific genes, or where the scope of surgery is selected, treatment intensity is adjusted, or follow-up strategies are established, and is not necessarily limited to the act of administering a specific therapeutic agent.

[0070] In one embodiment of the present invention, a method for diagnosing and treating cancer is provided, comprising the step of analyzing the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

[0071] In the above embodiment, a method for diagnosing and treating cancer is provided, further comprising the step of administering a cancer treatment agent to a target individual who has been diagnosed with cancer, suspected of having cancer, or classified as a high-risk group by analyzing the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

[0072] The above "intended subject" refers to a subject whose onset of disease is uncertain and who has a high probability of developing the disease; specifically, it refers to a patient suspected of having cancer or a subject for diagnosis.

[0073] Cancer therapeutic agents administered to target subjects diagnosed with cancer or evaluated as having a high risk of cancer by the above methylation level analysis may be selected according to the type of cancer, stage of progression, pathological characteristics and the patient's condition, and may include, but are not limited to, the following, for example.

[0074] Examples of the above-mentioned cancer therapeutic agents may include alkylating agents, antimetabolites, microtubule inhibitors, topoisomerase inhibitors, platinum compounds, anthracycline compounds, taxane compounds, anticancer antibiotics, hormone therapies, hormone antagonists, selective estrogen receptor modulators, aromatase inhibitors, targeted anticancer agents, tyrosine kinase inhibitors, monoclonal antibodies, immune checkpoint inhibitors, antibody-drug conjugates, cytotoxic T-cell activators, anti-angiogenesis inhibitors, platinum-based anticancer agents, DNA damage response modulators, PARP inhibitors, immunomodulators, and combination therapies thereof.Specifically, cisplatin, carboplatin, oxaliplatin, doxorubicin, epirubicin, daunorubicin, paclitaxel, docetaxel, vincristine, vinblastine, etoposide, irinotecan, 5-fluorouracil, capecitabine, gemcitabine, methotrexate, cyclophosphamide, ifosfamide, dacarbazine, temozolomide, paraplatin, bleomycin, In addition to mitomycin, targeted anticancer drugs may include imatinib, erlotinib, gefitinib, sorafenib, sunitinib, lapatinib, olaparib, niraparib, bevacizumab, trastuzumab, pertuzumab, cetuximab, and panitumumab; immunotherapies may include nivolumab, pembrolizumab, atezolizumab, durvalumab, and ipilimumab; and hormone therapies may include tamoxifen and letrozole. Drugs such as anastrozole and leuprorelin may be included, but are not limited to.

[0075] In addition, the above-mentioned cancer treatment drug may be administered alone or in combination with surgery, radiation therapy, immunotherapy, chemotherapy, or a combination of these.

[0076] The administration of the above-mentioned cancer treatment agent may be performed as part of various cancer treatment methods, and such treatment methods may be applied alone or in combination. For example, the treatment methods may include surgical treatment methods including surgical resection, partial resection, total resection, or tumor removal surgery. In addition, radiation therapy that inhibits the growth or kills cancer cells using radiation may be performed alone or as adjuvant therapy before or after surgery. Furthermore, the treatment methods may include chemotherapy involving the systemic or local administration of anticancer agents, and may include targeted therapy or immunotherapy using targeted anticancer agents or immunotherapies. Along with this, in the case of hormone-dependent cancer, hormone therapy or hormone deprivation therapy may be performed. Moreover, the treatment methods may be performed as combination therapy, neoadjuvant therapy, adjuvant therapy, or maintenance therapy combining two or more of surgery, radiation therapy, chemotherapy, targeted therapy, and immunotherapy, and may be utilized to determine treatment strategies or adjust treatment intensity based on the results of methylation level analysis.

[0077]

[0078] In one embodiment of the present invention, the first aspect provides a composition for cancer diagnosis or prognosis prediction comprising an active ingredient that analyzes the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

[0079] In the first embodiment, the second embodiment provides a composition for cancer diagnosis or prognosis prediction, wherein the cancer is any one selected from the group comprising thyroid cancer, breast cancer, pancreatic cancer, stomach cancer, blood cancer, lung cancer, liver cancer, colorectal cancer, rectal cancer, esophageal cancer, bile duct cancer, kidney cancer, bladder cancer, prostate cancer, ovarian cancer, uterine cancer, cervical cancer, head and neck cancer, skin cancer, and brain tumor.

[0080] In the first or second embodiment, the third embodiment provides a composition for cancer diagnosis or prognosis prediction, wherein the thyroid cancer is papillary thyroid cancer.

[0081] In any one of the first to third embodiments, the fourth embodiment provides a composition for cancer diagnosis or prognosis prediction, wherein the AGAP2 comprises chr12:58132478-58132734, EHBP1L1 comprises chr11:65352231-65353134, PRDM8 comprises chr4:81109887-81110460, CD37 comprises chr19:49841187-49841628, RIN3 comprises chr14:93153278-93154759, CARMIL2 comprises chr16:67686860-67687674, and GPR84 comprises chr12:54764065-54764510 CpG sites.

[0082] In any one of the first to fourth embodiments, the fifth embodiment provides a composition for cancer diagnosis or prognosis prediction, wherein the agent for analyzing the methylation level is one or more selected from the group consisting of a primer pair capable of amplifying a fragment containing a methylated CpG island site, a probe capable of hybridizing with a methylated CpG island site, a methylation-specific binding protein capable of binding with a methylated CpG island site, a methylation-specific binding antibody or aptamer, a methylation-sensitive restriction endonuclease, a sequencing primer, a sequencing biosynthesis primer, and a sequencing bioligation primer.

[0083] In any one of the first to fifth embodiments, the sixth embodiment provides a composition for cancer diagnosis or prognosis prediction in which the methylation is confirmed by contacting genomic DNA treated with a reagent that differently modifies methylated DNA and non-methylated DNA with a substance capable of detecting whether the CpG island region of any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 is methylated.

[0084] In any one of the first to sixth embodiments, the seventh embodiment provides a composition for cancer diagnosis or prognosis prediction in which the genomic DNA is isolated from a biological sample selected from the group consisting of cells, tissues, biopsies, paraffin tissues, blood, serum, plasma, fine needle aspiration specimens, urine, and combinations thereof derived from a patient suspected of having cancer or a subject of diagnosis.

[0085] In any one of the first to seventh embodiments, the eighth embodiment provides a composition for cancer diagnosis or prognosis prediction in which the reagent is one or more selected from the group consisting of bisulfite, hydrogen sulfite, disulfite, or combinations thereof.

[0086] In any one of the first to eighth embodiments, the ninth embodiment provides a composition for cancer diagnosis or prognosis prediction in which the methylation is detected by a method selected from the group consisting of PCR, methylation specific PCR, real-time methylation specific PCR, PCR using a methylation DNA specific binding protein, PCR using a methylation DNA specific binding antibody or aptamer, quantitative PCR, nucleic acid chip, sequencing, sequencing biosynthesis, and sequencing bioligation.

[0087] In any one of the first to ninth embodiments, the tenth embodiment provides a cancer diagnostic kit comprising the above composition.

[0088] In any one of the first to ten embodiments, the eleventh embodiment provides a cancer diagnostic kit in which the cancer is papillary thyroid cancer.

[0089] In one embodiment of the present invention, the 12th aspect provides a nucleic acid chip for cancer diagnosis comprising a probe capable of hybridizing with a fragment comprising a CpG island methylated for any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

[0090] In the 12th embodiment, the 13th embodiment provides a nucleic acid chip for cancer diagnosis in which the cancer is papillary thyroid cancer.

[0091] In the 12th or 13th embodiment, the 14th embodiment provides a nucleic acid chip for cancer diagnosis, wherein the AGAP2 comprises chr12:58132478-58132734, EHBP1L1 comprises chr11:65352231-65353134, PRDM8 comprises chr4:81109887-81110460, CD37 comprises chr19:49841187-49841628, RIN3 comprises chr14:93153278-93154759, CARMIL2 comprises chr16:67686860-67687674, and GPR84 comprises chr12:54764065-54764510.

[0092] In one embodiment of the present invention, the 15th aspect provides a method for providing information necessary for predicting the prognosis of cancer, comprising: (a) a step of isolating DNA from a biological sample; and (b) a step of detecting the methylation level for one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 in the DNA isolated in step (a).

[0093] In the 15th embodiment, the 16th embodiment provides a method for providing information necessary for predicting the prognosis of cancer, wherein the cancer is any one selected from the group comprising thyroid cancer, breast cancer, pancreatic cancer, stomach cancer, blood cancer, lung cancer, liver cancer, colorectal cancer, rectal cancer, esophageal cancer, bile duct cancer, kidney cancer, bladder cancer, prostate cancer, ovarian cancer, uterine cancer, cervical cancer, head and neck cancer, skin cancer, and brain tumor.

[0094] In the 15th or 16th embodiment, the 17th embodiment provides a method for providing information necessary for predicting the prognosis of cancer, wherein the thyroid cancer is papillary thyroid cancer.

[0095] In any one of the 15th to 17th embodiments, the 18th embodiment provides a method for providing information necessary for predicting the prognosis of cancer, wherein the AGAP2 comprises chr12:58132478-58132734, EHBP1L1 comprises chr11:65352231-65353134, PRDM8 comprises chr4:81109887-81110460, CD37 comprises chr19:49841187-49841628, RIN3 comprises chr14:93153278-93154759, CARMIL2 comprises chr16:67686860-67687674, and GPR84 comprises chr12:54764065-54764510.

[0096] In any one of the 15th to 18th embodiments, the 19th embodiment further comprises: step (b) a step of treating the genomic DNA separated in step (a) with a reagent that differently modifies methylated DNA and unmethylated DNA; (b'') a step of confirming the methylation status of one or more gene CpG island regions selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 in the genomic DNA or a fragment thereof treated with the reagent; and (b''') a step of determining thyroid cancer if the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 increases compared to a normal control group.

[0097] In any one of the 15th to 19th embodiments, the 20th embodiment provides a method for providing information necessary for predicting the prognosis of cancer, wherein the method further comprises: (c) a step of normalizing the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 measured in step (b) to calculate a score; (d) a step of summing the methylation scores calculated in step (c) to integrate them into a comprehensive methylation score; and (e) a step of classifying the comprehensive methylation score integrated in step (d) into a high-risk group if it is 0.4 or higher and classifying it into a low-risk group if it is less than 0.4;

[0098] In any one of the 15th to 20th embodiments, the 21st embodiment provides a method for providing information necessary for predicting the prognosis of cancer, wherein in step (e), the high-risk group is evaluated as having a lower survival rate and a higher likelihood of undergoing total thyroidectomy compared to the low-risk group.

[0099] In one embodiment of the present invention, the 22nd aspect provides a cancer diagnostic device comprising: (a) a detection unit for detecting the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 in DNA isolated from a biological sample; (b) a calculation unit for normalizing the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 detected by the detection unit to calculate a score and integrating the results; and (c) an evaluation unit for classifying the high-risk group if the integrated score from the calculation unit is 0.4 or higher, and the low-risk group if it is less than 0.4, and evaluating that the high-risk group has a lower survival rate and a higher likelihood of undergoing total thyroidectomy compared to the low-risk group.

[0100] In the 22nd embodiment, the 23rd embodiment provides a cancer diagnostic device comprising the CpG sites of AGAP2 chr12:58132478-58132734, EHBP1L1 chr11:65352231-65353134, PRDM8 chr4:81109887-81110460, CD37 chr19:49841187-49841628, RIN3 chr14:93153278-93154759, CARMIL2 chr16:67686860-67687674, and GPR84 chr12:54764065-54764510.

[0101] In the 22nd or 23rd embodiment, a cancer diagnostic device is provided in which the cancer is papillary thyroid cancer.

[0102] In one embodiment of the present invention, the 25th aspect provides a method for diagnosing and treating cancer, comprising the step of analyzing the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

[0103] In the 25th embodiment, the 26th embodiment provides a method for diagnosing and treating cancer, wherein the cancer is any one selected from the group comprising thyroid cancer, breast cancer, pancreatic cancer, stomach cancer, blood cancer, lung cancer, liver cancer, colorectal cancer, rectal cancer, esophageal cancer, bile duct cancer, kidney cancer, bladder cancer, prostate cancer, ovarian cancer, uterine cancer, cervical cancer, head and neck cancer, skin cancer, and brain tumor.

[0104] In the 25th or 26th embodiment, the 27th embodiment provides a method for diagnosing and treating cancer, wherein the thyroid cancer is papillary thyroid cancer.

[0105] In any one of the 25th to 27th embodiments, the 28th embodiment provides a method for diagnosing and treating cancer, wherein the AGAP2 comprises chr12:58132478-58132734, EHBP1L1 comprises chr11:65352231-65353134, PRDM8 comprises chr4:81109887-81110460, CD37 comprises chr19:49841187-49841628, RIN3 comprises chr14:93153278-93154759, CARMIL2 comprises chr16:67686860-67687674, and GPR84 comprises chr12:54764065-54764510.

[0106] In any one of the 25th to 28th embodiments, the 29th embodiment provides a method for diagnosing and treating cancer, wherein the step of analyzing the methylation level is performed through a preparation for analyzing the methylation level, and the preparation for analyzing the methylation level is one or more selected from the group consisting of a primer pair capable of amplifying a fragment containing a methylated CpG island site, a probe capable of hybridizing with a methylated CpG island site, a methylation-specific binding protein capable of binding with a methylated CpG island site, a methylation-specific binding antibody or aptamer, a methylation-sensitive restriction endonuclease, a sequencing primer, a sequencing biosynthesis primer, and a sequencing bioligation primer.

[0107] In any one of the 25th to 29th embodiments, the 30th embodiment provides a method for diagnosing and treating cancer, wherein the step of analyzing the methylation level comprises the step of contacting genomic DNA treated with a reagent that differently modifies methylated DNA and non-methylated DNA with a substance capable of detecting whether the CpG island region of any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 is methylated.

[0108]

[0109] According to the present invention, the accuracy of predicting papillary thyroid cancer can be improved based on specific biomarkers. In addition, by distinguishing thyroid cancer tumors into two subgroups, it can contribute to diagnosis, prognosis prediction, and treatment strategies.

[0110] Furthermore, the effects of the present invention are not limited to the effects described above, and should be understood to include all effects that can be inferred from the configuration of the invention described in the detailed description or claims of the present invention.

[0111]

[0112] Figure 1 is data confirming candidate CpG island regions.

[0113] Figure 2 shows the results of the target bisulfite sequencing evaluation.

[0114] Figure 3 is a workflow for selecting differential methylation regions (DMRs) using target non-base sequencing data.

[0115] Figures 4 and 5 are graphs distinguishing subgroups (PTC1, PTC2) within a cohort of patients with papillary thyroid cancer (PTC).

[0116] Figure 6 is a graph showing the correlation coefficients of methylation levels in subgroups (PTC1, PTC2) of patients with papillary thyroid cancer (PTC).

[0117] Figure 7 is a Kaplan-Meier curve showing the overall survival rate of subgroups (PTC1, PTC2) of patients with papillary thyroid cancer (PTC).

[0118] Figure 8 is a graph showing the differential methylation region (DMR).

[0119] Figure 9 is a table showing the gene distribution of differential methylation regions (DMR).

[0120] Figure 10 is a graph showing the relationship between binding motifs and differential methylation regions (DMRs) for the components of a subgroup complex.

[0121] Figure 11 is a graph confirming the hypermethylation concentration area.

[0122] Figure 12 is a graph confirming the separation between subgroups using Principal Component Analysis (PCA).

[0123] Figures 13 and 14 are the results of generating volcano plots visualizing differentially expressed genes (DEGs) between each subgroup.

[0124] Figures 15 and 16 show the results of performing K-means clustering on DEGs and classifying them into three clusters according to expression patterns.

[0125] Figures 17 and 18 are graphs showing that tumor-related genes are upregulated in PTC1 as a result of gene set enrichment analysis (GSEA).

[0126] Figure 19 shows the selection of 77 genes that differ differentially in methylation expression levels between PTC1 and PTC2.

[0127] Figure 20 shows the results of classifying 5 patients into PTC1 and 23 patients into PTC2 through hierarchical clustering7.

[0128] Figure 21 shows the results of separating the gene expression patterns of ATAC-seq data into subgroups using principal component analysis (PCA).

[0129] Figure 22 defines differential access regions (DARs) for each subgroup based on chromatin accessibility differences using Diffbind36.

[0130] Figure 23 shows the results of annotating the locations of differentially accessible regions (DARs) within the genome.

[0131] Figure 24 shows that the results of the ontology analysis of genes near the differentially accessible region of PTC1 are consistent with the gene ontology results of the TCGA RNA-seq data, indicating that these regions are associated with immune cell activation.

[0132] Figure 25 shows the results of an ontology analysis of genes near differentially accessible regions, indicating that regions with reduced accessibility are associated with development, differentiation, and proliferation of various cell types.

[0133] Figure 26 shows the results of identifying genes with significant changes in expression after demethylation agent treatment.

[0134] Figure 27 shows the results of selecting six genes from the DNA methylation subgroup using VSURF.

[0135] Figure 28 shows the accuracy results of applying the Random Forest algorithm to the test set for six genes of DNA methylation subgroups selected using VSURF.

[0136] Figure 29 shows that AGAP2 is a representative candidate biomarker in all Random Forest algorithm test datasets.

[0137] Figures 30 to 34 show the results confirming the association between CD37 overexpression and tumors.

[0138] Figure 35 shows the design of methylation-specific PCR (MSP) primers.

[0139] Figure 36 shows the methylation-specific PCR (MSP) results.

[0140] Figure 37 shows the results of applying methylation-specific PCR (MSP) primers to the genomic DNA of 5-Aza-treated BCPAP cells.

[0141] Figure 38 shows the result of confirming the difference in DNA methylation levels between PTC1 and PTC2 using the sum of the methylation signals of the biomarkers.

[0142] Figure 39 shows the results of verifying the biomarker gene expression of two subgroups through quantitative RT-PCR.

[0143] Figure 40 shows the results of successfully separating subgroups based on the DNA methylation scores of seven candidate biomarkers using a small amount of DNA extracted from fine needle aspiration (FNAB) samples.

[0144] Figure 41 shows the results confirming that the candidate gene expression level of fine needle aspiration (FNAB) samples classified as PTC1 via qRT-PCR is higher than that of PTC2.

[0145] Figure 42 clearly shows that patients are divided into two distinct subgroups from the principal component analysis (PCA) plot of gene expression patterns combining qMSP and qRT-PCR data.

[0146] Figure 43 shows the results of confirming the difference in survival rates between PTC1 and PTC2 in the same TNM weapon system.

[0147]

[0148] The present invention will be described in more detail below through examples. These examples are intended solely to explain the present invention more specifically, and it will be obvious to those skilled in the art that the scope of the present invention is not limited by these examples according to the gist of the invention.

[0149]

[0150] Experimental method

[0151] Biological sample collection

[0152] Thyroid tumor tissue and adjacent normal tissue were collected in pairs from 55 patients with papillary thyroid cancer (PTC). Comprehensive demographic information, including age, sex, and TNM stage, is listed in Table 1.

[0153] VariablesPTC1n = 34PTC2n = 21p-valueSurgeryTotal22160.550Partial115SexMale650.731Female2816SubtypeConventional32180.405Diffuse sclerosing10Encapsulated01Follicular11Tall cell01MultiplicityBilateral790.260Unilateral63No209CapsuleYes720.456No2619ThyroiditisYes1340.117No2017T StageI23120.438II56III33IV00N stageIa1040.392Ib1813Age, years40.9733.10.007Tumor size, cm1.782.240.2622Number of positive central neck LN4.064.550.6799Number of positive lateral neck LN3.733.20.9283Maximal size of lymph nodes1.151.210.898Calcitonin (before operation)2.0042.060.318

[0154] * p-values ​​were calculated using the chi-square test or Fisher's exact test based on the minimum sample size of a single box.

[0155] * Abbreviations: LN, lymph node; capsule, capsule infiltration

[0156]

[0157] Surgically removed tissues were immediately transported and stored in Dulbecco's modified Eagle's medium (Cytiva, Marlborough, MA, USA) supplemented with 10% FBS (Cytiva, Marlborough, MA, USA). Fine Needle Aspiration Biopsy (FNAB) samples were collected from 17 patients. FNAB was performed using a 20 mL plastic syringe and syringe holder (Cameco syringe pistol, Precision Dynamics, San Fernando, CA), and 21–23 gauge needles were used depending on the ultrasound pattern of the nodules. Upon arrival, the tissues were washed twice with PBS (Cytiva, Marlborough, MA, USA) and then rapidly frozen using liquid nitrogen to maintain nucleic acid integrity.

[0158]

[0159] CpG Island Selection for Probe Pool Design for Targeted Bisulfite Sequencing

[0160] To select candidate CpG islands for bisulfite sequencing, RRBS data obtained from TCGA thyroid cancer (THCA) data and GSE107738 thyroid cancer data were utilized, along with Infinium HumanMethylation450 BeadChip data. The beta values ​​of each CpG location were averaged according to the human reference genome (NCBI GRCh37; hg19) to represent the methylation values ​​of the corresponding CpG islands. Subsequently, the methylation values ​​of normal tissue samples were averaged to calculate representative normal tissue methylation values, and the difference between the average values ​​of tumor samples and normal tissues was calculated and summarized in a table.

[0161]

[0162] VariablesPTC1n = 95PTC2n = 304p-valueTNM StageI541720.570II425III2569IV (IV, IVA, IVC)1237SexMale24850.607Female71219HistologyClassical842750.391Tall cell1126Capsular invasionModerate / advanced (T4)6120.587Minimal (T3)2990None56189MultiplicityBilateral16570.8881Unilateral73228Isthmus416BRAF-RASclassBRAF-like671860.004RAS-like448TERTpromoter( C228T)Mutated350.4WT67220BRAFmutation(V600E)Mutated541760.865WT39122RASmutationYes0220.007No93276RETfusionYes12160.014No81282Age, years48.946.50.364Sample weight199.65197.080.6Differentiation score-0.54-0.20.016ERK score16.78.70.0014

[0163] * p-values ​​were calculated using the chi-square test or Fisher's exact test based on the minimum sample size of a single box.

[0164] CpG islands with a methylation difference of 10% or greater in more than 10% of patients were selected. Based on this criterion, 5,812 were selected by comparing TCGA normal and tumor samples, 2,612 by comparing GSE107738 normal and benign tumor samples, and 2,484 by comparing normal and malignant samples. Arbor Bioscience's probe design tool (MyBaits) was used to remove non-specific target regions within the genome, and ultimately, 7,217 CpG islands were selected as targets.

[0165]

[0166] Targeted bisulfite library preparation and sequencing

[0167] Genomic DNA was extracted from tissue samples using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) and performed according to the manufacturer's recommended guidelines. Subsequently, DNA concentration and purity were measured using a UV spectrophotometer (Nanodrop 2000; Thermo Fisher, Carlsbad, CA, USA). A total of 500 ng of genomic DNA was disrupted in low-EDTA TE buffer using a Covaris M220 Focused Ultrasonicator (Woburn, MA, USA), and the quality, quantity, and fragment size (major peaks in the 250–300 bp range) of the disrupted DNA were verified using an Agilent Technologies 2100 Bioanalyzer system (Santa Clara, CA, USA).

[0168] The DNA libraries were subjected to bisulfite conversion using Zymo Research’s EZ DNA Methylation-Gold Kit (Irvine, CA, USA), and subsequent library preparation was carried out according to the manufacturer’s protocol using Swift Biosciences’ Accel-NGS® Methyl-Seq DNA Library Kit (Ann Arbor, MI, USA) and other specified enzymes, buffers, and reagents. Finally, eight libraries were assembled and incubated with a probe pool tailored to the target region, and sequenced using a HiSeq2500 sequencer (Illumina, San Diego, CA, USA) in a 2x 100 bp paired format, generating 2 Gb of sequencing data from each sample.

[0169]

[0170] ATAC library preparation and sequencing

[0171] The preparation procedure for the ATAC library of BCPAP cells is as follows. Cells were lysed using fresh ATAC-seq lysis and wash buffer according to the preparation guidelines. The cells were centrifuged to form a pellet, the supernatant was removed, and the cells were resuspended in lysis buffer. After incubation, the samples were diluted and washed to form a pellet again. Subsequently, the transposition mixture was added and incubated; after incubation, the DNA was purified and concentrated. Barcoding was performed by attaching a unique adapter and performing PCR amplification and purification, after which the library concentration was quantified for sequencing. The prepared libraries were sequenced using the 2X 100 bp paired-end method on an Illumina HiSeq2500 sequencer, generating 2 Gb of sequencing data from each sample.

[0172]

[0173] Next-generation sequencing data preprocessing

[0174] Targeted bisulfite sequencing data was quality checked using FastQC (version 0.11.9), and adapter sequences and low-quality sequences were removed using Trim Galore (version 0.6.7) and Cutadapt 16 (version 2.8).

[0175] Sequencing reads were aligned with Bowtie2 (version 2.4.4) using Bismark 17 (version 0.22.3) based on the human CpG islet reference genome hg19, the alignment quality of each dataset was evaluated using Picard's CollectHsMetrics, and all quality validation data were summarized using multiQC 18.

[0176] Methylated and unmethylated cytosines at each CpG site were identified in the post-processed data using Bismark's methylation extractor, and only sites methylated tenfold or more were selected for subsequent analysis. Finally, the methylation values ​​of CpG sites within the same CpG island were calculated by averaging based on hg19. nRNA-seq data were aligned using HISAT2, and SAM files were aligned using Samtools. Read counts of BAM files were quantified using HTSeq (version 0.11.1) and normalized using DESeq2 (version 3.12).

[0177] Principal component analysis (PCA) of gene expression patterns was performed using ggplot2 (version 3.3.3), and correlations between samples were analyzed.

[0178] Differentially expressed genes (DEGs) between samples were identified using DESeq2, and genes with a minimum 2-fold difference in expression and an adjusted p-value < 0.05 were selected. ATAC sequencing (ATAC-seq) data from the nBCPAP cell line and public data (GSE162515) were processed using the PEPATAC pipeline, and PEPATAC ensures reproducibility using RefGenie Asset Manager for consistent reference genome annotation.

[0179] Adapters and mitochondrial DNA were removed using Trimmomatic, and read alignment and duplicate removal were performed using Bowtie2 and Picard, respectively.

[0180]

[0181] Differentially methylated region (DMR) selection

[0182] The analysis was performed under the assumption that each CpG island represents the mean value. Target bisulfite sequencing data were reviewed to identify targets in which DNA methylation changed by more than 20% compared to normal tissue paired with the tumor in more than 20% of 55 thyroid cancer patients. Additionally, differential methylation regions (DMRs) were identified separately in patients with and without thyroiditis, enabling a comparative analysis of epigenetic changes associated with this condition. A total of 333 DMRs were identified, and these regions were further classified into hypermethylated and hypomethylated groups based on the average methylation level across the entire DNA methylation subgroup.

[0183]

[0184] Next-generation sequencing data analysis

[0185] Before analysis, 399 patients with papillary thyroid cancer (PTC) were selected from the THCA (TCGA, Firehose Legacy) dataset provided by cBioPortal (19,20). The DNA methylation levels of CpG islands in TCGA samples were determined by subtracting the average methylation level of normal thyroid samples from the methylation level of each tumor sample. K-means clustering was applied to divide patients within each cohort into two subgroups; to ensure the stability and reliability of the clusters, 1,000 iterations and distinct random centroids were used, and up to 5,000 iterations were allowed for convergence. Genomic regions of interest were annotated using the ChIPseeker (21) package in R and the TxDB.Hsapiens.UCSC.hg19.knownGene and org.HS.eg.db packages. In addition, i-cisTarget was utilized to explore predictable consensus motifs within 329 candidate differential methylation regions (DMRs). Gene ontology analysis and gene set enrichment analysis (GSEA) were performed using the gprofiler2 and fgsea packages, utilizing custom gene matrix transpose (GMT) files from MSigDB. Data visualization for DNA methylation microarray, bisulfite sequencing, RNA-seq, and ATAC-seq was performed using the ComplexHeatmap (version 4.2.2) package in R. Differential access regions between conditions were analyzed using Diffbind with a threshold of |Fold Change| > 2, while the criterion |Fold Change| > 0.25 was applied for candidate biomarker selection.To refine the initial 28 candidate genes into a core gene set and remove redundancy, the random forest algorithm implemented in the VSURF package in R was utilized.

[0186]

[0187] Quantitative Methylation Specific PCR (qMSP)

[0188] Before quantifying the DNA methylation levels of the target region, 500 ng of genomic DNA and DNA methylation regulator DNA (Takara Bio Inc., Japan) extracted from patients with papillary thyroid cancer (PTC) were each treated with sodium bisulfite (EZ DNA Methylation-Lightning Kits, Zymo Research, Irvine, CA, USA). The concentration of bisulfite-treated genomic DNA was quantified using a UV spectrophotometer (Nanodrop 2000; Thermo Fisher Scientific, Carlsbad, CA, USA). For nqMSP analysis, a master mix for GC-rich PCR (KAPA SYBR FAST qPCR Master Mix (2X), Kapa Biosystems) and a PCR cycler (LightCycler 480 II; Roche Diagnostics) were used, and 2 ng of bisulfite-treated genomic DNA was used for each qMSP experiment. After 45 iterations of quantitative PCR (qPCR), the crossover point (Cp) was determined by directly adjusting the signal threshold. The DNA methylation level of each CpG island was calculated using the formula in Equation 1:

[0189] [Mathematical Formula 1]

[0190] 2(Cp of Unmet primer) - (Cp of Met primer)

[0191] To distinguish Methylation Specific PCR (MSP) results between samples, the following scoring system was established: First, the individual DNA methylation score of each biomarker was calculated using the formula in Equation 2:

[0192] [Mathematical Formula 2]

[0193]

[0194] To normalize the background signal inherent in the normal thyroid tissue of each patient, a normalized score was calculated by subtracting the tumor score from the normal score, and the cumulative DNA methylation scores of seven biomarkers for each patient were summed and integrated into a comprehensive DNA methylation score.

[0195]

[0196] qPCR (Quantitative PCR)

[0197] To confirm the expression of each candidate gene in the patients, cDNA was synthesized from 1 μg of total RNA using reverse transcriptase (Invitrogen, Carlsbad, CA, USA). For qPCR, KAPA SYBR FAST qPCR Master Mix (2X, Kapa Biosystems) and a PCR cycler (LightCycler 480 II; Roche Diagnostics, Basel, Switzerland) were used, and 5 ng of cDNA was used for each qPCR experiment. After 45 qPCR cycles, Cp values ​​were calculated by manually adjusting the signal threshold, and the expression level of each target gene was determined using the comparative Cp method (2-ΔΔCp) with GAPDH expression as a control. To classify the papillary thyroid carcinoma (PTC) subgroup, the sum of the expression levels of seven biomarkers (AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84) for each patient was calculated and integrated into a composite score.

[0198]

[0199] Cell Culture and Transformation

[0200] BCPAP (Elabscience, Houston, Texas, USA) cells were cultured in RPMI-1640 medium (Gibco, Houston, Texas, USA) supplemented with 10% fetal serum (Hyclone, Logan, UT, USA) and 1% penicillin-streptomycin (Gibco, Houston, Texas, USA), while 293T (ATCC, Manassas, VA, USA) cells were cultured in Dulbecco's modified Eagle medium (Hyclone, Logan, UT, USA) supplemented with 10% FBS (Hyclone, Logan, UT, USA) and 1% penicillin (Gibco, Houston, Texas, USA). Cell lines were cultured under 5% CO₂ conditions in a humid environment at 37°C, and cells were subcultured every 3 days when they reached 80-90%.

[0201] To overexpress the CD37 gene, the CD37 sequence was cloned into the pcDNA3.1 V5 / His A vector (Invitrogen, Waltham, MA, USA), and the pcDNA3.1 CD37 V5 / His A plasmid (12 μg, 100 mm scale) was transformed into cells using Lipofectamine 2000 (Invitrogen, Waltham, MA, USA) according to the manufacturer's instructions.

[0202]

[0203] proliferation analysis

[0204] For IncuCyte® cell proliferation analysis, 1×10⁴ 293T cells were seeded into a 96-well plate, and cell growth over time was monitored and images were captured using the IncuCyte® Live-Cell Analysis System. For automated cell count analysis, the IncuCyte® S3 Live-Cell Analysis System (Sartorius, Gottingen, Germany) was used.

[0205]

[0206] Analysis of migration and infiltration

[0207] Infiltration analysis was performed using an 8-μm pore size dual-chamber (Transwell) system (Costa; Corning Incorporated, Corning, NY, USA). In this experiment, 1×10 5 Canine 293T cells were inoculated into an upper chamber coated with Matrigel; the upper chamber was filled with serum-free RPMI medium, while the lower chamber was filled with serum-containing RPMI medium to act as a chemoattractant. After 48 hours of incubation, non-infiltrating or non-migrating cells remaining on the membrane surface were removed with a cotton swab, and infiltrating cells that had migrated through the membrane were stained and counted. Wound healing analysis was performed in 24-well plates, with 2 × 10⁶ cells in each well. 5 Dog cells were inoculated. Wounds were created using a scratcher (SPL; Pocheon, Gyeonggi-do, South Korea), and wound healing images were taken at 0 and 24 hours. Wound width was measured using ImageJ software.

[0208]

[0209] Colony formation analysis

[0210] 2.5 × 10³ cells were seeded into a 6-well plate and cultured for 12 days. Afterward, the cells were fixed with 4% paraformaldehyde and stained with 1% crystal violet to make it easier to count the number of colonies.

[0211]

[0212] Statistical analysis

[0213] To investigate the correlations between DNA methylation subgroups within each group, the mean relative methylation levels of tumor samples in each subgroup were calculated, and Pearson correlation coefficients were derived using the 'ggpairs' function. The chi-square test was primarily used to verify the statistical significance of the clinical analysis, while Fisher's exact test was applied to cells with an expected frequency of less than 5 to ensure statistical validity. Survival analysis was performed using the Kaplan-Meier method and the log-rank test.

[0214] Experimental results

[0215] Identification of Differentially Methylated Regions (DMRs) in Papillary Thyroid Carcinoma (PTC)

[0216] Accurately measuring methylation changes at all CpG sites within selected CpG islands plays a crucial role in the development of highly specific methylation-specific PCR (MSP) primers. To identify candidate CpG island regions, TCGA and public data (RRBS, GSE107738) were analyzed, and ultimately, 7,217 CpG island regions of interest were identified (Fig. 1).

[0217] Figure 1 shows the construction of a panel of targeted CpG islands through preliminary analysis. Figure 1A is the preliminary analysis procedure for selecting candidate CpG islands. The methylation difference between tumor tissue and normal tissue was calculated by averaging the beta values ​​of CpG sites. CpG islands with a methylation change of 10% or more in more than 10% of the total patients were selected as candidates. Figure 1B summarizes the selected CpG islands from each dataset. 5,812 CpG islands were selected from the TCGA-THCA data, 2,612 from benign tumors of GSE107738, and 2,484 from malignant tumors, respectively. Figure 1C is the targeted bisulfite sequencing data preprocessing pipeline. The methylation status of each CpG site was determined based on a coverage of at least 10x. The average methylation level within each CpG island was calculated.

[0218] To determine the target base sequencing (TBS) from the above regions, thyroid tumor tissue and adjacent normal tissue were isolated from 55 Korean papillary thyroid cancer (PTC) patients, and the data quality was evaluated after sequencing. The results showed significantly higher read depth compared to other sequencing data in terms of base sequence quality, reference genome alignment, and read depth (Fig. 2).

[0219] Figure 2 relates to quality control of target bisulfite sequencing. In Figure 2, the corresponding indicators A–C were evaluated using Picard. The bar graphs represent the fold enrichment across the entire genome relative to the baited region (A, left), the proportion of on-target aligned to the baited region (B, left), and the average coverage of the baited region (C, left), respectively. The box plot shows the average values ​​of each indicator (A–C) (right).

[0220] 333 differentially methylated regions (DMRs) were identified using a selection workflow (Fig. 3), of which 329 DMRs were ultimately used for subsequent analysis.

[0221] Figure 3 is a workflow for selecting differential methylation regions (DMRs) using target bisulfite sequencing data. In Figure 3, the blue panel at the top represents specific criteria for selecting target CpG islands in the probe panel, as described in A of Figure 1. The green panel at the bottom of Figure 3 represents the procedure for selecting differentially methylated CpG islands from target bisulfite sequencing data of the study cohort.

[0222] K-means clustering identified two distinct patient subgroups within our cohort (the 55 papillary thyroid cancer patients mentioned above) and the TCGA cohort (the 399 papillary thyroid cancer patients mentioned above) (Fig. 4), and a clear distinction between the two subgroups within our cohort was derived through principal component analysis (PCA) of gene expression patterns using DNA methylation levels of 329 differential methylation regions (DMRs) (Fig. 5). To explore the correlation between the two subgroups, the correlation coefficients of DNA methylation levels for each subgroup were also evaluated. The results showed high correlation coefficients of 0.863 between the PTC1 subgroups and 0.442 between the PTC2 subgroups, suggesting the existence of corresponding subgroups within each cohort (Fig. 6). In particular, the difference in DNA methylation levels of the Differential Methylation Region (DMR) between PTC1 and PTC2 was more pronounced in our cohort, which may be due to the fact that targeted base sequencing reads the methylation levels of all CpG sites within the region of interest. In summary, heterogeneous PTC subgroups can be detected through targeted bisulfite sequencing. Furthermore, the clinical significance of subgroups defined by DNA methylation levels of the Differential Methylation Region (DMR) was evaluated, and a survival analysis between the two groups was conducted using clinical data from TCGA.

[0223] Kaplan-Meier curves showed that PTC1 had a lower overall survival rate than PTC2 (Fig. 7), and in our cohort, there were no significant differences in clinicopathological characteristics except for age at diagnosis (Table 1). In the TCGA cohort, the RAS-like papillary thyroid carcinoma (PTC) classification and RAS mutations were frequently found in PTC2 patients (Table 2). The thyroid differentiation score, a gene score related to thyroid function and metabolism based on the expression levels of 16 genes, was lower in PTC1, while the ERK score, calculated using 52 gene signatures evaluating ERK (MAPK) activity, was higher in PTC1 (Table 2). In normal thyroid tissue, ERK activity is strictly regulated; otherwise, it can induce abnormal cell proliferation, differentiation, and apoptosis. This suggests the possibility that PTC1 may exhibit a more aggressive phenotype than PTC2. In summary, DNA methylation levels identified in 329 DMRs demonstrate that they can serve as classification criteria to divide the PTC cohort into two groups, indicating prognostic and pathological differences.

[0224]

[0225] Characterization and Functional Significance of DNA Methylation Patterns

[0226] Before exploring the functional relevance of the 329 selected differential methylation regions (DMRs), we classified them according to the average DNA methylation levels of the subgroups in our cohort. By calculating the average DNA methylation levels of the PTC1 and PTC2 subgroups, we redefined the differential methylation regions (DMRs) in PTC1 as hypermethylated regions (245) and hypomethylated regions (84) compared to PTC2 (Fig. 8A). Analysis of the annotations of the differential methylation region locations within the genome revealed that DNA methylation occurred at a higher frequency in the promoter and remote intergenic regions of PTC1 (Fig. 8B). To investigate the functional roles of genes associated with these differential methylation regions (DMRs), we performed gene ontology analysis and discovered that genes related to developmental processes were significantly distributed in the hypermethylated differential methylation regions of PTC1 (Fig. 9). In addition, motif analysis was performed to identify transcription factors associated with differential methylation regions (DMRs), and this analysis confirmed that binding motifs for components of the PRC1 and PRC2 complex were abundant in the differential methylation regions (DMRs) (Fig. 10). Coverage plot analysis, which visualizes genome-wide locations, showed that hypermethylation was concentrated in the HOXA and HOXB clusters (Fig. 11), which is consistent with results observed in various cancers, including thyroid cancer, in other studies. Taken together, our findings demonstrate the existence of characteristic methylation patterns among papillary thyroid carcinoma (PTC) subgroups, characterized by significant hypermethylation of genes important for development. This suggests that epigenetic changes in the PTC1 subgroup of papillary thyroid carcinoma are associated with the developmental process and can contribute to understanding the heterogeneity of papillary thyroid carcinoma (PTC).

[0227]

[0228] Transcriptome Profiling and Differential Gene Expression Analysis of Papillary Thyroid Carcinoma (PTC) Subgroups

[0229] RNA-seq data from TCGA were analyzed to compare the transcriptome profiles of the two subgroups and normal tissue. As expected, principal component analysis (PCA) of gene expression patterns showed a clear separation between the subgroups based on principal components 1 and 2 (Fig. 12). DESeq2 was used to investigate differentially expressed genes (DEGs) between each subgroup, and volcano plots were generated to visualize the DEGs (Figs. 13, 14). In Fig. 14, A relates to PTC1 and normal tissue, and B relates to PTC2 and normal tissue; Fig. 14 is a volcano plot showing differential expression between PTC1 and PTC2.

[0230] K-means clustering was performed on all DEGs, and they were classified into three clusters based on their expression patterns (Fig. 15). Cluster 1 consisted of genes that were commonly low in PTC1 and PTC2 compared to normal tissue, Cluster 2 consisted of genes that were specifically increased in PTC1, and Cluster 3 consisted of genes that were commonly increased in both PTC1 and PTC2 compared to normal tissue. Consistent with previous research results, ontology analysis showed that differentially expressed gene (DEG) cluster 1 included genes known to be low in thyroid cancer, whereas cluster 3 included genes that are high in thyroid cancer.

[0231] Differentially expressed gene (DEG) cluster 2 is rich in immune-related terms (Fig. 16). Gene set enrichment analysis (GSEA) shows that genes related to tumor formation and tumor evasion are upregulated in PTC1 (Figs. 17, 18). In Fig. 18, C is a comparison of the top 10 enriched terms in PTC1 with PTC2 based on normalized enrichment scores, and D is a comparison of the bottom 10 enriched terms in PTC1 with PTC2 based on normalized enrichment scores.

[0232]

[0233] Integrated Analysis of DNA Methylation, Gene Expression, and Chromatin Accessibility in Papillary Thyroid Carcinoma (PTC) Subgroups

[0234] To identify genes directly affected by methylation changes, a comprehensive analysis was performed by integrating our methylation data with publicly available RNA-seq and ATAC-seq data. First, 77 genes that were differentially methylated and showed different expression levels between PTC1 and PTC2 were selected (Fig. 19). To investigate the chromatin accessibility of each subgroup, publicly available ATAC-seq data (GSE162515) was utilized, and DNA methylation subgroups were inferred by comparing the expression patterns of these 77 genes with the RNA-seq data from TCGA. Through hierarchical clustering, 5 patients could be classified as PTC1 and 23 as PTC2 (Fig. 20). Principal component analysis (PCA) of the gene expression patterns in the ATAC-seq data shows separation by subgroup (Fig. 21). Figure 21 shows the analysis of chromatin accessibility among methylation subgroups in papillary thyroid carcinoma (PTC), and presents the results of principal component analysis (PCA) using ATAC sequencing data of PTC. Subgroups are distinguished by color.

[0235] Differences in chromatin accessibility were observed using Diffbind36, and differential accessable regions (DARs) were defined for each subgroup (Fig. 22). Fig. 22 shows volcano plots of differential chromatin accessibility between PTC1 and normal tissue (B), PTC2 and normal tissue (C), and PTC1 and PTC2 tissues (D).

[0236] In particular, as a result of annotating the locations of differentially accessible regions (DARs) within the genome, it was found that in PTC1, the proportion of promoters was significantly higher in regions with increased chromatin accessibility compared to PTC2, while there were fewer promoters in regions with reduced accessibility and they were more distributed in remote intergenic regions (Fig. 23). Fig. 23 shows the genomic annotation of differentially accessible regions (DARs). The DARs were annotated by referencing TxDB and org.HS.eg.db.

[0237] Genetic ontology analysis of genes near differentially accessible regions in PTC1 revealed that these regions are associated with immune cell activation, which is consistent with the genetic ontology analysis results of the TCGA RNA-seq data (Fig. 24). Conversely, regions with reduced accessibility indicated association with development, differentiation, and proliferation of various cell types (Fig. 25). Our integrated analysis identified a potential set of biomarkers for PTC1 and highlights the importance of DNA methylation, gene expression, and chromatin accessibility. Furthermore, the significant abundance of immune-related regions in PTC1 supports the potential role of the immune system in the progression of thyroid cancer.

[0238]

[0239] Discovery and Prioritization of Candidate Genes for PTC Subgroup Methylation Detection System

[0240] As a result of the integrated analysis of the bisulfite sequencing, publicly available RNA-seq, and ATAC-seq data of the present invention, 35 genomic regions showing distinct changes were observed in each dataset, and it was predicted that these changes would be interrelated. These regions include a total of 28 genes (Table 3). Table 3 is an integrated catalog of candidate genes combined from the targeted bisulfite sequencing, RNA-seq, and ATAC-seq analyses.

[0241] Genomic locationGeneTargeted bisulfite sequencingRNA-seq(TCGA)RNA-seq(GSE162515)ATAC-seq(GSE162515)Met diff (T1 - T2)LFCPadjLFCPadjLFChr11:65352231-65353134EHBP1L134.280.704.4E-201.285.5E-030.44chr11:65359292-65360328EHBP1L129.750.704.4E-201.285.5E-030.27chr12:58132478-58132734AGAP229.191.431.4E-362.741.6E-061.60chr4:81109887-811104 60PRDM828.441.811.8E-472.292.1E-040.48chr5:10649367-10650352ANKRD33B27.730.913.1E-122.201.2E-040.39chr14:93153 278-93154759RIN326.370.822.6E-281.893.4E-050.39chr12:58130870-58132047AGAP224.941.431.4E-362.741.6E-060.58chr12 :58119909-58121551AGAP224.361.431.4E-362.741.6E-060.36chr16:67686860-67687674CARMIL224.282.921.7E-774.303.4E-060.25chr15:40583093-40583526PLCB223.261.612.1E-472.623.4E-050.28chr19:49841187-49841628CD3722.722.195.2E-573.85 4.0E-060.98chr19:17877468-17877777FCHO122.700.772.0E-062.473.1E-030.28chr5:43037259-43037520ANXA2R22.321.241.1E-272.067.4E-051.99chr2:11774310-11774521GREB122.24-0.874.7E-05-2.152.0E-020.72chr19:15563869-15564223RASAL321.031.765.7E-573.013.2E-060.32chr19:15568027-15569227RASAL320.1 01.765.7E-573.013.2E-060.87chr21:45789090-45789373TRPM220.001. 382.0E-361.953.1E-040.63chr4:81128229-81128691PRDM819.741.811.8E-472.292.1E-040.29chr16:67681975-67683924CARMIL218.682.921. 7E-774.303.4E-060.76chr2:198029068-198029438ANKRD4417.221.088.1E-352.755.1E-070.34chr12:54764065-54764510GPR8416.721.488.2E-331.901.4E-020.30chr17:3847999-3848570ATP2A316.401.276.8E-322.843.4E-061.47chr17:14201726-14202052HS3ST3B115.681.298.7E-283 .003.0E-071.88chr19:49842654-49843628CD3715.672.195.2E-573.85 4.0E-060.64chr19:1070985-1071812ARHGAP4515.061.095.4E-332.613. 4E-050.52chr7:5336513-5336894SLC29A4-16.26-0.905.2E-06-2.229. 6E-032.48chr22:50483350-50483579IL17REL-16.262.551.1E-295.481. 9E-030.32chr16:29675845-29676120SPN-17.671.772.5E-442.646.5E-042.61chr11:65408344-65408631SIPA1-18.320.623.8E-141.725.4E-052.42chr19:13207375-13207621LYL1-19.670.941.0E-162.401.1E-051.84chr11:67176945-67177169TBC1D10C-21.630.799.2E-151.911.5E-042.71chr17:72347924-72348322BTBD17-21.72-1.761.1E-05-3.493.5E-0 20.70chr19:3178741-3179986S1PR4-24.251.712.0E-453.324.7E-072. 61chr12:6664425-6665336IFFO1-27.251.073.0E-371.464.1E-042.66chr11:63974829-63975048FERMT3-37.591.832.4E-542.681.0E-052.47.

[0242] * Abbreviations: Met diff, mean DNA methylation difference; LFC, Log2 (variation scale); Padj, adjusted p-value

[0243] Before developing a system to sensitively detect methylation changes in these regions, target genes were selected. To this end, ATAC-seq analysis was performed after treating the thyroid cancer cell line BCPAP with the demethylating agent 5-azacytidine (5-Aza) to identify the region showing the greatest change in chromatin accessibility. As a result, changes in the expression of AGAP2, EHBP1L1, GPR84, and PRDM8 were observed (Fig. 26). In addition, target genes were selected using a random forest algorithm, and six genes (PRDM8, EHBP1L1, CARMIL2, CD37, RIN3, SIPA1) were selected by using VSURF to identify suitable and non-overlapping classifiers of DNA methylation subgroups (Fig. 27). When this algorithm was applied to the test set, the overall accuracy reached 87.4% (Fig. 28), and consistent changes were observed in two genes (EHBP1L1 and PRDM8) in both the BCPAP cell line ATAC-seq results and the random forest analysis. In addition, AGAP2 was identified as a representative candidate biomarker in all datasets (Fig. 29). Thus, a system for detecting methylation changes was constructed based on eight genes that showed differences in the BCPAP cell line ATAC-seq results and the random forest analysis.

[0244] CD37, one of the candidate biomarkers, was selected, and its role was verified in the laboratory. After overexpressing CD37 in 293T and BCPAP cell lines, functional experiments such as proliferation, invasion, wound healing, and colony formation were performed to evaluate the effect of CD37 on tumorigenic characteristics (Figs. 30 to 34). As a result, it was found that CD37 significantly improved the tumorigenic characteristics of both cell lines.

[0245] Figure 30 confirms the successful overexpression of CD37 in 293T cells after transfection through western blot analysis. Figure 31 shows that cell proliferation increased in 293T cells overexpressing CD37 compared to the control group, based on proliferation analysis results using IncuCyte. Figure 32 shows the results of an invasion analysis demonstrating increased invasiveness of 293T cells after CD37 overexpression, including representative images of invaded cells and quantitative analysis. Figure 33 shows that wound healing analysis results indicate accelerated wound closure and increased motility in 293T cells overexpressing CD37. Figure 34 shows that colony formation analysis results in BCPAP cells overexpressing CD37 indicate increased tumorigenicity due to increased colony formation compared to the control group.

[0246] These results suggest that CD37 may play a significant role in promoting tumor progression. From this, the possibility was confirmed that candidate biomarkers discovered through the analysis of the present invention, particularly CD37, are involved in influencing the aggressiveness of PTC.

[0247]

[0248] Validation and Clinical Application of a Methylation Subgroup Classification System Based on Methylation Specific PCR (MSP)

[0249] Methylation-specific PCR (MSP) is a widely used and simple method for evaluating DNA methylation levels within specific genomic regions. To construct a molecular system for classifying PTC patients into PTC1 or PTC2, we utilized the Methylation-specific PCR (MSP) assay. In the validation experiments, hypermethylated regions, where methylation signals appear more clearly, were prioritized for validation because they are easier to detect than hypomethylated regions. Therefore, CpG islands in the hypomethylated SIPA1 region of PTC1 were excluded from further validation to ensure methodological consistency and clarity in result interpretation. CpG sites within candidate differential methylation regions (DMRs) were ranked based on the variance (variability) of DNA methylation level differences between PTC1 and PTC2, and between PTC1 and corresponding normal tissues, and Methylation-specific PCR (MSP) primers targeting these top CpG sites were designed (Fig. 35). Figure 35 shows the binding regions of qMSP primers according to differences in DNA methylation levels. The primer binding regions are represented in the BED format indicated by green boxes. MethPrimer software was used for primer design. To evaluate the efficiency of methylation-specific PCR (MSP) primers, methylation-specific PCR (MSP) was performed on bisulfite-treated unmethylated and methylation-controlled DNA samples. The signal ratio between the detected methylated and unmethylated DNA primers quantitatively reflected the DNA methylation levels and corresponded to the ratio of methylated DNA to unmethylated DNA (Figure 36). Additionally, to verify the detection effect of the methylation-specific PCR (MSP) primers, they were applied to the genomic DNA of 5-Aza-treated BCPAP cells. As expected, a significant decrease in methylation signals was observed in candidate regions after 5-Aza treatment (Figure 37).Subsequently, the ability of methylation-specific PCR (MSP) primers to distinguish differences in DNA methylation levels between PTC1 and PTC2 was evaluated, and a scoring system was introduced to classify patient-specific DNA methylation subgroups. This score was calculated as the sum of methylation signals from seven biomarkers normalized relative to normal tissue, and the results indicated that the MSP system very effectively distinguished between PTC1 and PTC2 (p < 0.0001) (Fig. 38). Finally, the biomarker gene expression of the two subgroups was verified via quantitative RT-PCR, and it was confirmed that gene expression increased in PTC1 patients as expected (Fig. 39).

[0250] For methylation and expression verification, 2 ng of nucleic acid is required for a single qMSP, and 5 ng for a single qRT-PCR. These amounts are readily available from fine-needle aspiration biopsy (FNAB) samples. Previous studies have shown that these amounts are sufficient even when a small number of cancer cells are present in FNAB samples. The present invention aims to evaluate whether a sensitive approach can detect DNA methylation changes in candidate differential methylation regions (DMRs) in minimally invasive biopsy samples. qMSP was performed on 17 FNAB samples from a newly selected group of papillary thyroid carcinoma (PTC) patients (Table 4).

[0251] VariablesPTC1PTC2p-valuen = 11n = 6SexMale500.102Female66Age, years38.939.50.9

[0252] * The p-value was calculated using Fisher's exact test.

[0253] Even with a small amount of DNA extracted from fine needle aspiration (FNAB) samples, subgroups were successfully classified according to the DNA methylation scores of seven candidate biomarkers (Fig. 40). In addition, qRT-PCR confirmed that the expression levels of candidate genes in FNAB samples classified as PTC1 were higher than those of PTC2 (Fig. 41), which verified the correlation between DNA methylation and expression levels in FNAB samples. Furthermore, a PCA plot combining qMSP and qRT-PCR data clearly showed that patients were divided into two distinct subgroups (Fig. 42). In particular, since the qMSP method enables more distinct and accurate separation of patient groups than qRT-PCR, it is emphasized that DNA methylation is a highly specific marker for patient classification.

[0254] The TNM staging system (Tumor, Node, Metastasis system) serves as a fundamental framework for classifying malignant tumors and is widely used for cancer diagnosis and predicting patient prognosis. The goal is to evaluate the potential for improving survival prediction by integrating the existing TNM staging system with the classification system of the present invention. In particular, the survival analysis of the present invention was performed by combining TNM staging data from TCGA with DNA methylation subgroup status (PTC1 and PTC2), and the results showed a significant difference in survival rates. Patients in TNM stages III-IV and the PTC1 subgroup exhibited significantly lower survival rates than patients in PTC2, despite having the same TNM stage (Fig. 43). This result suggests that DNA methylation subgroups can act as independent prognostic factors and can complement traditional TNM classification. Therefore, distinguishing between PTC1 and PTC2 is crucial for determining surveillance and treatment strategies for thyroid cancer.

[0255]

[0256] In this invention, 7,217 CpG islands of interest were identified through DNA methylation data, and subsequently, 333 differential methylation regions (DMRs) were isolated from our cohort using bisulfite sequencing. Two distinct patient subgroups were identified using K-means clustering and principal component analysis, and these results were verified through correlation analysis with TCGA data. In particular, the subgroup designated as PTC1 had a lower overall survival rate compared to PTC2, and this group contained 245 hypermethylated regions and 84 hypomethylated regions located mainly in promoter and remote interzegnic regions. Functional analysis revealed that the hypermethylated differential methylation regions (DMRs) of PTC1 were rich in developmental genes and genes associated with Polycomb Repressive Complex 1 and 2 (PRC1 / PRC2) binding motifs, suggesting an association with tumor progression and worsening prognosis. By integrating bisulfite sequencing, RNA sequencing, and ATAC-seq data, 35 genomic regions and 28 genes were identified as potential biomarkers, of which 7 genes were established as true biomarkers confirmed through qMSP and qPCR analysis. For clinical application, methylation changes in fine-needle aspiration (FNAB) samples could be detected using a highly sensitive quantitative system, which is a minimally invasive method routinely used for preoperative evaluation.

[0257] Among these seven biomarkers, RIN3 and AGAP2 were found to play important roles in the Ras pathway and are critically involved in the progression of thyroid cancer. RIN3 appears to be involved in the Ras signaling pathway due to the inclusion of a Ras-associated domain, while AGAP2 exerts anti-apoptotic effects by activating nuclear phosphate inositide 3-kinase (PI3K), a process frequently initiated by Ras activation. Furthermore, AGAP2, EHBP1L1, and CARMIL2 were shown to play significant roles in cancer metastasis by contributing to cell migration and invasion. Conversely, CD37 and GPR84 are associated with the immune system; this aligns with the immune-related transcriptomic characteristics of PTC1, suggesting a potential role in regulating immune responses within thyroid cancer. These seven biomarkers represent an area of ​​thyroid cancer research that has not yet been fully explored. Their findings deepen the understanding of the molecular complexity of thyroid cancer (THCA) and demonstrate their potential as therapeutic targets or diagnostic markers when combined with novel technologies. In particular, their diverse functions, ranging from signaling pathways to interactions with the immune system, will provide new insights into thyroid cancer research.

[0258] Binding sites for MAX and E2F6, components of the PRC1 complex, and SUZ12 and EZH2, components of the PRC2 complex, were observed to be abundant in hypermethylated differential methylation regions (DMRs). Previous studies have emphasized that the interaction between the PRC1 complex and DNA methylation plays a crucial role in regulating the expression of genomic defense genes during mammalian development. These genes play a key role in maintaining cellular genomic stability by detecting and repairing DNA damage, minimizing mutations, and inhibiting the activity of selfish DNA elements such as transposons. Furthermore, the PRC2 complex, known for mediating H3K27 methylation and genomic silencing, recruits DNA(cytosine-5)-methyltransferase 3 beta (DNMT). Abnormal activity of the PRC2 complex is widely recognized as contributing to cancer development by altering transcriptional regulation. However, further bioinformatics analysis and experimental verification are required to confirm whether these hypothetical mechanisms are characteristic of the PTC1 subgroup.

[0259] Fine-needle aspiration biopsy (FNAB) plays a pivotal role in the preoperative evaluation of thyroid nodules and is primarily used to determine malignancy and decide the necessity of surgery. This study demonstrates that specific cancer subgroups can be identified using qMSP and qRT-PCR, advanced detection techniques that analyze DNA methylation patterns and biomarker expression levels in FNAB samples. By evaluating methylation levels and gene expression together, we can effectively classify various cancer subgroups, providing crucial information for establishing personalized treatment strategies based on tumor aggressiveness. In particular, qMSP demonstrated higher specificity and clearer subgroup separation than qRT-PCR due to the inherent instability of RNA and variability in expression levels, as qMSP detects more stable methylation patterns within DNA. This reliability is critical as it enables the definitive classification of patients with minimal DNA samples during surgical planning. The primers of the present invention successfully classified subgroups even under these conditions, reaffirming that qMSP is a useful tool for distinguishing tumor subtypes of thyroid cancer (see Table 5).

[0260] Primers used in the present invention (SEQ ID NOs 1 to 28) Gene Methylated Forward (5'→3') Methylated Reverse (5'→3') Unmethylated Forward (5'→3') Unmethylated Reverse (5'→3') AGAP2GGATTTAAGGGATAGGGGTCGC (SEQ ID NO. 1) ACCAAACCGAACCCAAACGA (SEQ ID NO. 2) GGGATTTAAGGGATAGGGGTTGTG (SEQ ID NO. 3) AACCAAACCAAACCCAAACAA (SEQ ID NO. 4) EHBP1L1GTTGCGACGGTTTTCGGTTAAC (SEQ ID NO. 5) CATCGAAAACGAACTACTATTCCGC (SEQ ID NO. 6) TGAGGTTGTGATGGTTTTTGGTTAAT (SEQ ID NO. 7) CCATCAAAAACAAACTACTATTCCAC (SEQ ID NO. 8)PRDM8TTAAGTTTCGGAGTTTGGTAGGACG(Sequence No. 9)TAAAAACAACCACGAAAAAAACTCG(Sequence No. 10)TTAAGTTTTGGAGTTTGGTAGGATG(Sequence No. 11)AAAAACAACCACAAAAAAAACTCAC(Sequence No. 12)CD37TGTGTTTTTTTTAATTTTGCGTTCG(Sequence No. 13)CCCATAACCCTAAAACTCTAACCGC(Sequence No. 14)TGTGTTTTTTTTAATTTTGTGTTTG(Sequence No. 15)CCCATAACCCTAAAACTCTAACCAC(Sequence No. 16)RIN3TAGATAGTTTGGCGTTCGTTATTCG(Sequence No. 17)GACGAAACCCTATCTCCTCAAACG(Sequence No. 18)TAGATAGTTTGGTGTTTGTTATTTG(Sequence No. 19)AACAAAACCCTATCTCCTCAAACAC(Sequence No. 20)CARMIL2TTTCGGGGTTGTTGAGAGATGTC(Sequence No. 21)ACTCCACCCGACCTACTACCGA(Sequence No. 22)TTTTTTTGGGGTTGTTGAGAGATGTT(Sequence No. 23)CCACTCCACCCAACCTACTACCA(Sequence No. 24)GPR84TCGTTTTTCGTTCGTTTTAGGGTAC(Sequence No.25)AACAACGACGAACCTACTACCGC(Sequence No. 26)TGTTTTTTGTTTGTTTTAGGGTATG(Sequence No. 27)ACAACAACAAACCTACTACCAC(Sequence No. 28)

[0261] For surgeons, this means that beyond the diagnosis of malignant tumors, fine-needle aspiration biopsy (FNAB) provides preoperative prognostic information, aiding in determining the scope of surgery and assessing the need for additional treatments, such as radioactive iodine (RAI) therapy. The ability to accurately classify tumor subtypes preoperatively can help surgeons decide between total thyroidectomy and lobectomy, determine the necessity of lymph node dissection, or formulate more aggressive treatment plans for patients identified with the high-risk PTC1 subtype.

[0262] Although the cohort of 55 PTC patients in this invention is significantly large compared to similar studies, the long-term prognosis of these patients has not yet been fully determined due to the prospective nature of the cohort. To compensate for this, the clinical significance was indirectly inferred through association with the TCGA (Cancer Genome Archive) cohort, which helps to overcome some of the limitations.

[0263] This invention has significant implications for establishing personalized treatment plans, improving prognostic accuracy, and enhancing patient counseling regarding surgical options and expected outcomes by providing a molecular tool for predicting tumor behavior. This approach also aligns with the principles of precision medicine, which aims for personalized treatment based on individual tumor biology.

[0264] In conclusion, the present invention presents a novel minimally invasive method for preoperative risk classification of papillary thyroid cancer using epigenetic biomarkers detectable in fine needle aspiration (FNAB) samples. By detecting specific DNA methylation patterns associated with tumor aggressiveness, qMSP (quantitative methylation-specific PCR) can accurately classify patients with minimal DNA input. These insights provide surgeons with a useful tool for surgical decisions, enabling more precise surgical interventions to improve patient outcomes and optimize resource utilization in the treatment of thyroid cancer.

[0265]

[0266] Foregoing, specific parts of the present invention have been described in detail. It is evident to those skilled in the art that such specific descriptions are merely preferred embodiments and do not limit the scope of the invention. Accordingly, the actual scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A composition for cancer diagnosis or prognosis prediction comprising an active ingredient that analyzes the methylation level of any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

2. In Paragraph 1, A composition for diagnosing cancer or predicting prognosis, wherein the above cancer is any one selected from the group comprising thyroid cancer, breast cancer, pancreatic cancer, stomach cancer, blood cancer, lung cancer, liver cancer, colorectal cancer, rectal cancer, esophageal cancer, bile duct cancer, kidney cancer, bladder cancer, prostate cancer, ovarian cancer, uterine cancer, cervical cancer, head and neck cancer, skin cancer, and brain tumor.

3. In Paragraph 2, A composition for cancer diagnosis or prognosis prediction, wherein the above-mentioned thyroid cancer is papillary thyroid cancer.

4. In Paragraph 1, A composition for cancer diagnosis or prognosis prediction, wherein the above AGAP2 comprises a CpG site of chr12:58132478-58132734, EHBP1L1 comprises a CpG site of chr11:65352231-65353134, PRDM8 comprises a CpG site of chr4:81109887-81110460, CD37 comprises a CpG site of chr19:49841187-49841628, RIN3 comprises a CpG site of chr14:93153278-93154759, CARMIL2 comprises a CpG site of chr16:67686860-67687674, and GPR84 comprises a CpG site of chr12:54764065-54764510.

5. In Paragraph 1, A composition for cancer diagnosis or prognosis prediction, wherein the agent for analyzing the methylation level is one or more selected from the group consisting of a primer pair capable of amplifying a fragment containing a methylated CpG island site, a probe capable of hybridizing with a methylated CpG island site, a methylation-specific binding protein capable of binding with a methylated CpG island site, a methylation-specific binding antibody or aptamer, a methylation-sensitive restriction endonuclease, a sequencing primer, a sequencing biosynthesis primer, and a sequencing bioligation primer.

6. In Paragraph 1, A composition for cancer diagnosis or prognosis prediction, wherein the above methylation is confirmed by contacting genomic DNA treated with a reagent that modifies methylated DNA and non-methylated DNA differently with a substance capable of detecting whether the CpG island region of any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 is methylated.

7. In Paragraph 6, A composition for cancer diagnosis or prognosis prediction, wherein the genomic DNA is isolated from a biological sample selected from the group consisting of cells, tissues, biopsies, paraffin tissues, blood, serum, plasma, fine needle aspiration specimens, urine, and combinations thereof derived from a target individual.

8. In Paragraph 6, A composition for cancer diagnosis or prognosis prediction, wherein the above-mentioned reagent is one or more selected from the group consisting of bisulfite, hydrogen sulfite, disulfite, or combinations thereof.

9. In Paragraph 1, A composition for cancer diagnosis or prognosis prediction, wherein the above methylation is detected by a method selected from the group consisting of PCR, methylation specific PCR, real-time methylation specific PCR, PCR using a methylation DNA specific binding protein, PCR using a methylation DNA specific binding antibody or aptamer, quantitative PCR, nucleic acid chip, sequencing, sequencing biosynthesis, and sequencing bioligation.

10. A cancer diagnostic kit comprising the composition of any one of claims 1 to 9.

11. In Paragraph 10, The above cancer is a cancer diagnostic kit for papillary thyroid cancer.

12. A nucleic acid chip for cancer diagnosis comprising a probe capable of hybridizing with a fragment containing a CpG island methylated for any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

13. In Paragraph 12, The above cancer is papillary thyroid cancer, a nucleic acid chip for cancer diagnosis.

14. In Paragraph 12, A nucleic acid chip for cancer diagnosis, wherein the above-mentioned AGAP2 includes the CpG site chr12:58132478-58132734, EHBP1L1 includes the CpG site chr11:65352231-65353134, PRDM8 includes the CpG site chr4:81109887-81110460, CD37 includes the CpG site chr19:49841187-49841628, RIN3 includes the CpG site chr14:93153278-93154759, CARMIL2 includes the CpG site chr16:67686860-67687674, and GPR84 includes the CpG site chr12:54764065-54764510. 15.(a) a step of isolating DNA from a biological sample derived from a target individual; and (b) detecting the methylation level for one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 in the DNA separated in step (a); comprising a method for providing information necessary for predicting the prognosis of cancer.

16. In Paragraph 15, A method for providing information necessary for predicting the prognosis of cancer, wherein the cancer is any one selected from the group comprising thyroid cancer, breast cancer, pancreatic cancer, stomach cancer, blood cancer, lung cancer, liver cancer, colorectal cancer, rectal cancer, esophageal cancer, bile duct cancer, kidney cancer, bladder cancer, prostate cancer, ovarian cancer, uterine cancer, cervical cancer, head and neck cancer, skin cancer, and brain tumor.

17. In Paragraph 16, A method for providing information necessary for predicting the prognosis of cancer, wherein the above-mentioned thyroid cancer is papillary thyroid cancer.

18. In Paragraph 15, A method for providing information necessary for predicting the prognosis of cancer, wherein the above AGAP2 includes the CpG sites chr12:58132478-58132734, EHBP1L1 includes chr11:65352231-65353134, PRDM8 includes chr4:81109887-81110460, CD37 includes chr19:49841187-49841628, RIN3 includes chr14:93153278-93154759, CARMIL2 includes chr16:67686860-67687674, and GPR84 includes chr12:54764065-54764510.

19. In Paragraph 15, The above step (b) is (b') A step of treating the genomic DNA separated in step (a) with a reagent that modifies methylated DNA and unmethylated DNA differently; (b'') a step of confirming the methylation status of one or more gene CpG island sites selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 in genomic DNA or a fragment thereof treated with the above reagent; and (b''') a step of determining thyroid cancer if the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2 and GPR84 increases compared to a normal control group; a method for providing information necessary for predicting the prognosis of cancer, further comprising 20. In Paragraph 15, A method for providing information necessary for predicting the prognosis of the above cancer (c) a step of normalizing the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 measured in step (b) above to calculate a score; (d) a step of summing the methylation scores calculated in step (c) above and integrating them into a comprehensive methylation score; and (e) a step of classifying the integrated comprehensive methylation score from step (d) as high-risk if it is 0.4 or higher and as low-risk if it is less than 0.4; further comprising a method for providing information necessary for predicting the prognosis of cancer.

21. In Paragraph 20, A method for providing information necessary for predicting the prognosis of cancer, wherein in step (e) above, the high-risk group is evaluated as having a lower survival rate and a higher likelihood of undergoing total thyroidectomy compared to the low-risk group. 22.(a) A detection unit for detecting the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 in DNA isolated from a biological sample derived from a target individual; (b) a calculation unit that calculates a score by normalizing the methylation level of one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 detected by the detection unit and integrates the same; and (c) an evaluation unit that classifies the score integrated in the above calculation unit as high-risk group if it is 0.4 or higher, and as low-risk group if it is less than 0.4, and evaluates that the high-risk group has a lower survival rate and a higher likelihood of undergoing total thyroidectomy compared to the low-risk group; a cancer diagnostic device comprising 23. In Paragraph 22, A cancer diagnostic device comprising the CpG sites of the above AGAP2 chr12:58132478-58132734, EHBP1L1 chr11:65352231-65353134, PRDM8 chr4:81109887-81110460, CD37 chr19:49841187-49841628, RIN3 chr14:93153278-93154759, CARMIL2 chr16:67686860-67687674, and GPR84 chr12:54764065-54764510.

24. In Paragraph 22, The above cancer is papillary thyroid cancer, a cancer diagnostic device.

25. A method for diagnosing and treating cancer, comprising the step of analyzing the methylation level of any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84.

26. In Paragraph 25, A method for diagnosing and treating cancer, wherein the above cancer is any one selected from the group comprising thyroid cancer, breast cancer, pancreatic cancer, stomach cancer, blood cancer, lung cancer, liver cancer, colorectal cancer, rectal cancer, esophageal cancer, bile duct cancer, kidney cancer, bladder cancer, prostate cancer, ovarian cancer, uterine cancer, cervical cancer, head and neck cancer, skin cancer, and brain tumor.

27. In Paragraph 26, The above-mentioned thyroid cancer is papillary thyroid cancer, a method for diagnosing and treating cancer.

28. In Paragraph 25, A method for diagnosing and treating cancer, wherein the above AGAP2 includes a CpG site of chr12:58132478-58132734, EHBP1L1 includes a CpG site of chr11:65352231-65353134, PRDM8 includes a CpG site of chr4:81109887-81110460, CD37 includes a CpG site of chr19:49841187-49841628, RIN3 includes a CpG site of chr14:93153278-93154759, CARMIL2 includes a CpG site of chr16:67686860-67687674, and GPR84 includes a CpG site of chr12:54764065-54764510.

29. In Paragraph 25, A method for diagnosing and treating cancer, wherein the step of analyzing the methylation level is performed using a preparation for analyzing the methylation level, and the preparation for analyzing the methylation level is one or more selected from the group consisting of a primer pair capable of amplifying a fragment containing a methylated CpG island site, a probe capable of hybridizing with a methylated CpG island site, a methylation-specific binding protein capable of binding with a methylated CpG island site, a methylation-specific binding antibody or aptamer, a methylation-sensitive restriction endonuclease, a sequencing primer, a sequencing biosynthesis primer, and a sequencing bioligation primer.

30. In Paragraph 25, A method for diagnosing and treating cancer, comprising the step of analyzing the methylation level, wherein genomic DNA treated with a reagent that differently modifies methylated DNA and non-methylated DNA is contacted with a substance capable of detecting whether the CpG island region of any one or more genes selected from the group consisting of AGAP2, EHBP1L1, PRDM8, CD37, RIN3, CARMIL2, and GPR84 is methylated.

Images