DNA methylation markers for predicting liver cancer recurrence and uses thereof
By using specific DNA methylation markers and Infinium HumanMethylation 850K microbead array technology, combined with machine learning, the problem of difficult to accurately predict liver cancer prognosis is solved, and accurate prediction and personalized treatment of liver cancer recurrence are achieved.
Patent Information
- Application Number
- CN202080017304.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2019-02-28
- Filing Date
- 2020-02-26
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2040-02-26
AI Technical Summary
The prior art is difficult to accurately predict the prognosis of patients with liver cancer, which leads to risks in treatment choices and is unable to effectively distinguish between high-risk and low-risk patients.
Compositions that predict liver cancer recurrence were isolated and kits and methods for predicting liver cancer recurrence were developed using DNA methylation markers cg21325760, cg10544510, cg06702718 and cg15997204 combined with Infinium HumanMethylation 850K microbead array and machine learning technology. .
Accurate prediction of liver cancer recurrence is achieved, personalized and accurate treatment is improved, unnecessary overtreatment is reduced, and prediction performance is significantly improved.
Smart Images

Figure CN113544288B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a DNA methylation marker for predicting liver cancer recurrence according to liver cancer prognosis by utilizing differences in genomic DNA methylation levels and its use, and specifically relates to a DNA methylation marker for predicting liver cancer recurrence in normal liver tissue of liver cancer patients after treatment. Background Art
[0002] Liver cancer can be roughly divided into primary liver cancer (hepatocellular carcinoma, liver cancer) arising from liver cells themselves and metastatic liver cancer in which cancer of other tissues has metastasized to the liver. Approximately more than 90% of liver cancers are primary liver cancers, and the commonly understood term liver cancer refers to primary liver cancer. According to the "2017 Cause of Death Statistics" released by Statistics Korea in September 2018, the number of cancer deaths in 2017 was 153.9 per 100,000 population, an increase of 1.0 (0.6%) from the previous year. The cancer mortality rate is ranked from high to low as lung cancer (35.1), liver cancer (20.9), colorectal cancer (17.1), stomach cancer (15.7) and pancreatic cancer (11.3).
[0003] Like other cancers, the treatment of liver cancer is mainly divided into radical treatment and conservative treatment. Here, radical treatment refers to treatment that completely eliminates the cancer itself, including liver resection surgery, liver transplantation, radiofrequency ablation, etc. In addition, if these treatments are not possible, carotid artery chemoembolization, radiotherapy, and anti-cancer treatment are performed. However, this treatment method is almost difficult to cure completely, so the purpose of medical examinations is to provide radical treatment through early detection. In the case of liver resection surgery, about 50% of patients will experience cancer recurrence because the resected liver is also exposed to hepatitis instead of a normal liver.
[0004] The prognosis of liver cancer refers to the various conditions of the patient predicted according to liver cancer after liver cancer is diagnosed, such as the possibility of complete liver cancer treatment, the possibility of recurrence after treatment, and the possibility of survival of the patient, which depends on various circumstances, such as the severity of the disease, the time of diagnosis, and the course of treatment. Liver cancer can only be effectively treated by appropriately applying various treatment methods according to the prognosis. For example, for patients with an expected good prognosis, risky treatment methods that may cause serious side effects to the patient, such as aggressive chemotherapy or surgery and radiotherapy, should be avoided, and relatively mild, conservative, and safe treatment methods should be selected. Conversely, for patients with an expected poor prognosis, efforts should be made to increase the survival period or survival rate through active mobilization of treatment methods such as chemotherapy or surgery and radiotherapy.
[0005] However, it is difficult to accurately predict the prognosis of liver cancer patients based on existing technologies. In order to accurately predict the prognosis, an analysis method is needed to classify patients by risk group, but until now, there has been a situation where the prognosis is determined only by the clinical pathological stage of liver cancer at the time of diagnosis and initial surgical treatment, and there is no means to accurately predict the prognosis of liver cancer.
[0006] As a result of efforts to solve this problem, the present inventors prepared a method capable of predicting the prognosis of liver cancer patients based on DNA methylation data of liver cancer tissues. Summary of the invention
[0007] Technical issues
[0008] The purpose of the present invention is to provide a composition comprising DNA methylation markers for predicting liver cancer recurrence and a method for predicting the possibility of liver cancer recurrence using the composition.
[0009] Technical Solution
[0010] To achieve the above object, one aspect of the present invention provides a composition for predicting liver cancer recurrence, comprising one or more DNA methylation markers selected from the group consisting of DNA methylation markers represented by SEQ ID NOs: 1 to 4.
[0011] In the present invention, the DNA methylation markers represented by the above SEQ ID NOs: 1 to 4 refer to cg21325760, cg10544510, cg06702718 and cg15997204, respectively, which are DNA methylation markers represented by CG locus symbols ( Infinium HumanMethylation 850K Bead Array CpG ID).
[0012] As used herein, the term "liver cancer" is a general term for cancers occurring in liver tissue, and liver cancer can be divided into primary liver cancer (hepatocellular carcinoma, liver cancer) produced by liver cells themselves and metastatic liver cancer in which cancers of other tissues have metastasized to the liver. Various treatment methods such as liver resection, liver transplantation, radiofrequency ablation, and carotid artery chemoembolization can be used for the treatment of liver cancer, but the 5-year survival rate is only 33.6% (average value from 2011-2015; statistics from the Ministry of Health and Welfare), and the recurrence rate also reaches 70%. Compared with other cancers, the 5-year survival rate of liver cancer is lower and the recurrence rate is higher, so a method capable of predicting recurrence is needed.
[0013] As used herein, the term "DNA methylation" refers to the covalent binding of a methyl group to the C5 position of a cytosine base in genomic DNA. The methylation level refers to, for example, the amount of methylation present in the DNA base sequence of all genomic regions and some non-genomic regions, and in the present invention, refers to the degree of methylation of the DNA methylation markers shown in SEQ ID NOs: 1 to 4. In the DNA methylation markers of SEQ ID NOs: 1 to 4, methylation may occur throughout the sequence.
[0014] In the present invention, the composition for predicting liver cancer recurrence can be a composition comprising one or more, two or more, three or more or four DNA methylation markers. Specifically, the composition for predicting liver cancer recurrence can respectively comprise one of the following DNA methylation markers: cg21325760, cg10544510, cg06702718 and cg15997204, and can include cg21325760 and cg10544510; cg21325760 and cg06702718; cg21325760 and cg15997204; cg10544510 and cg06702718; cg10544510 and cg15997204; or a combination of cg06702718 and cg15997204, which is a combination of two markers. In addition, the composition for predicting the recurrence of liver cancer may include three of the DNA methylation markers, such as cg21325760, cg10544510 and cg06702718; cg21325760, cg10544510 and cg15997204; cg21325760cg06702718 and cg15997204; or a combination of cg10544510, cg06702718 and cg15997204. In addition, the composition for predicting the recurrence of liver cancer may include all of the DNA methylation markers cg21325760, cg10544510, cg06702718 and cg15997204.
[0015] In the present invention, the DNA methylation marker cg15997204 is a marker present in the MYT1L gene, and the MYT1L (Myelin Transcription Factor 1Like) gene encodes a transcription factor belonging to the CCHC zinc finger protein. According to data published by TCGA (The Cancer Genome Atlas), MYT1L is expressed in a variety of cancers, but the correlation between the methylation level of this gene and the recurrence of liver cancer is still unclear.
[0016] In the present invention, the DNA methylation marker cg21325760 is present in the MAGEL2 (MAGE (melanoma-associated antigen) family member L2) gene and is highly expressed in testicular cancer according to TCGA data, but the correlation between the methylation level of this gene and liver cancer recurrence is unclear.
[0017] In one embodiment of the present invention, the inventors isolated genomic DNA from liver cancer tissues and used The Infinium HumanMethylation 850K microbead array confirmed the methylation level. The results confirmed that the liver cancer tissues were divided into two groups according to the methylation level, and it was known that there was a significant difference in the recurrence rate of liver cancer between the two groups. Thereafter, DNA methylation markers with significantly different methylation levels were identified in the two groups by machine learning techniques such as random forests, and DNA methylation markers were found (see Table 3). In addition, the present invention completes the present invention by confirming that when four DNA methylation markers are used together instead of using one DNA methylation marker, the accuracy of predicting liver cancer recurrence converges to 1, thereby completing the present invention ( Figure 5 and Table 2 ).
[0018] In the present invention, liver cancer may be hepatocellular carcinoma generated by liver cells themselves, or metastatic liver cancer in which cancer of other tissues metastasizes to the liver, but preferably hepatocellular carcinoma. Hepatocellular carcinoma accounts for about 90% of malignant tumors of the liver and is common in Korea, Japan, Southeast Asia, China, etc. Most occur in cirrhosis, but some are caused by chronic hepatitis B or C.
[0019] As used herein, the term "recurrence" refers to abnormal cell proliferation in a liver cancer patient after liver cancer treatment, i.e., liver cancer reoccurs in the liver. As treatment methods for liver cancer patients, hepatectomy, radiofrequency ablation, carotid artery chemoembolization, percutaneous ethanol injection, etc. can be used, but are not limited thereto.
[0020] In the present invention, the composition for predicting the prognosis of liver cancer can be implemented in the form of a kit for predicting the recurrence of liver cancer, which also includes reagents necessary for identifying the methylation level of the DNA methylation marker.
[0021] One aspect of the present invention provides a method for predicting liver cancer recurrence, comprising the following steps:
[0022] (A) Steps for isolating genomic DNA from liver tissue to be analyzed;
[0023] (b) a step of detecting the methylation level of one or more DNA methylation markers selected from the group consisting of the DNA methylation markers represented by SEQ ID NOs: 1 to 4; and
[0024] (c) A step of comparing the methylation level of the detected DNA methylation marker with the methylation level of the DNA methylation marker detected in normal liver tissue.
[0025] In the present invention, the DNA methylation markers represented by SEQ ID NOs: 1 to 4 refer to cg21325760, cg10544510, cg06702718 and cg15997204, respectively, which are DNA methylation markers represented by CG locus symbols ( Infinium HumanMethylation 850K Bead Array CpG ID).
[0026] In one embodiment of the present invention, the liver tissue to be analyzed can be liver cancer tissue, hepatocellular carcinoma tissue, or metastatic liver cancer or metastatic cancer tissue at a site where liver cancer has metastasized, preferably hepatocellular carcinoma tissue. As a method for separating the liver tissue to be analyzed from the subject to be analyzed, methods such as percutaneous biopsy and jugular vein biopsy known in the technical field to which the present invention belongs can be used.
[0027] In one embodiment of the present invention, the method for predicting the prognosis of liver cancer can be a method for identifying the methylation of at least two DNA methylation markers in step (b). The inventors found four DNA methylation markers (probes) that can predict liver cancer recurrence (see Table 2), and evaluated the performance of predicting liver cancer recurrence according to the number of markers. As a result of the evaluation, it was confirmed that when there are two or more markers, the performance such as accuracy and sensitivity converge to 0.9, and when the number of markers is 3 or more, the performance of liver cancer recurrence prediction converges to 1 ( Figure 5 ).
[0028] From this result, it can be seen that the DNA methylation marker of the present invention has excellent liver cancer prognosis prediction performance even when used alone, but its performance is significantly excellent when used in multiples. Therefore, the DNA methylation marker of the present invention can be used alone or in combination of two or more markers to predict liver cancer recurrence.
[0029] Specifically, for the DNA methylation markers of the present invention, cg21325760, cg10544510, cg06702718 and cg15997204 can be used as a single marker, or a combination of two markers (cg21325760 and cg10544510; cg21325760 and cg06702718; cg21325760 and cg15997204; cg067027510 and cg06702718; cg067027 510 and cg15997204; or cg06702718 and cg15997204) or in a combination of three markers (cg21325760, cg10544510 and cg06702718; cg21325760, cg10544510 and cg15997204; cg21325760, cg06702718 and cg15997204; or cg10544510, cg06702718 and cg15997204). In addition, all four markers can be combined and used.
[0030] In one embodiment of the present invention, the method for predicting the prognosis of liver cancer may also include, after step (c), a step of determining that the possibility of liver cancer recurrence is higher when the methylation level of the DNA methylation marker identified in the sample to be analyzed is lower than the methylation level of normal liver tissue.
[0031] In the present invention, the method for identifying the methylation level of DNA methylation markers can be performed by PCR, methylation-specific PCR, real-time methylation-specific PCR, and PCR using methylated DNA-specific binding proteins, quantitative PCR, DNA chips, pyrophosphate sequencing, commercially available chips, etc., which are well known in the art, but are not limited thereto.
[0032] The present inventors have confirmed that in patients with recurrent liver cancer (MG2), the methylation level of DNA methylation markers is significantly lower than that in normal liver tissue (N)( Figure 3 , 7 and 8), therefore, DNA methylation markers can be effectively used to predict the recurrence of liver cancer.
[0033] Beneficial Effects
[0034] When the DNA methylation marker for predicting liver cancer recurrence of the present invention is used, the recurrence of liver cancer can be easily predicted, and more personalized treatment can be provided for patients at high risk of liver cancer recurrence, and clinical information can be provided to avoid unnecessary overtreatment. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1Shown are the results of identifying the methylation levels of DNA isolated from hepatocellular carcinoma samples (A; n=140)) and normal liver samples (B; n=96).
[0036] Figure 2 Shown are the results of classifying hepatocellular carcinoma samples (n=140) into methylation group 1 (MG1; n=81) and methylation group 2 (MG2; n=59) according to prognosis, and then classifying each sample by a consensus clustering method according to the DNA methylation pattern.
[0037] Figure 3 Results identifying recurrence-free survival (RFS) of methylation group 1 (MG1) and methylation group 2 (MG2) are shown.
[0038] Figure 4 Shown is the process of confirming the variable importance plot (varImpPlot) in methylation group 1 (MG1) and methylation group 2 (MG2) (A), and then selecting the top 50 probes involved in methylation group classification (B).
[0039] Figure 5 The results of confirming the efficiency of liver cancer prognosis prediction according to the number of probes involved in methylome classification are shown as area under the curve (AUC) values.
[0040] Figure 6 Results identifying the methylation levels of probes involved in methylome classification in normal liver samples (N), methylome 1 (MG1), and methylome 2 (MG2) are shown.
[0041] Figure 7 Results of classification of the autonomously generated liver cancer validation cohort data using probes involved in methylome classification are shown.
[0042] Figure 8 The results of classifying liver cancer data published in TCGA (The Cancer Genome Atlas) using probes involved in methylome classification are shown. DETAILED DESCRIPTION
[0043] Hereinafter, one or more embodiments will be described in more detail by way of examples. However, these examples are for illustrating one or more embodiments, and the scope of the present invention is not limited to these examples.
[0044] Experimental methods
[0045] 1. Recruitment of patients with hepatocellular carcinoma
[0046] The studies related to the present invention were approved by the Review Board of Seoul National University Hospital, and written consent was obtained from all patients. 184 patients who underwent hepatocellular carcinoma (hereinafter referred to as HCC) resection in the Department of Surgery of Seoul National University Hospital from 2011 to 2016 were recruited, and the microscopic resection margin was clean and the tumor was completely removed by postoperative CT scan. Patients visited the hospital regularly and were monitored for recurrence every 3 months according to standard protocols, including serum tumor markers, alpha-fetoprotein (AFP) levels, contrast CT scans or magnetic resonance imaging (MRI). Recurrence was classified into early recurrence (≤1 year) and late recurrence (>1 year), excluding patients who relapsed within 1 year after HCC resection. Demographic data and clinicopathological characteristics were determined on the hospital medical record, and patients were followed up until death after resection. The reason for excluding patients with recurrence within 1 year is that in the case of early recurrence, HCC resection was incorrect or recurred due to reasons other than liver cancer. It is common, therefore, that it does not meet the purpose of the present invention, that is, to find markers related to the recurrence of new cancers in a completely normal liver after cancer treatment.
[0047] 2. Generation of methylation data
[0048] Among the 182 HCC patients, 42 patients who had HCC recurrence within 1 year after surgery were excluded from the analysis. TM DNA was extracted from frozen HCC samples of the remaining 140 patients using the 5M genomic DNA extraction kit (Bioneer). Each sample was analyzed using the Illumina Infinium HumanMethylation EPIC 850K Beadchip (Illumina) using 500 ng of DNA and the Illumina iScan System (Illumina, CA, USA) according to the manufacturer's standard protocol. The Illumina Infinium HumanMethylation EPIC Beadchip contains DNA methylation values for 850,000 or more CpG sites for each single nucleotide.
[0049] 3. DNA methylation data processing
[0050] For the data obtained using Illumina Infinium HumanMethylation EPIC BeadChip in 2 above, the methylation data were preprocessed using the minfi R package in the R platform (https: / / cran.r-project.org / bin / windows / base / ; version 3.3.2), which is an open software. Specifically, the original IDAT files of HCC samples were processed, and background correction and dye bias were corrected. Thereafter, the signal intensity of methylated and unmethylated probes was measured, thereby quantifying the DNA methylation value of each probe as a b value ranging from 0 (unmethylated) to 1 (fully methylated). Next, data quality control was performed on all samples, and all 140 samples passed the quality control standards. Finally, in order to reduce the batch effect problem, the methylation data were normalized using the functional normalization method.
[0051] 4. DNA methylation value
[0052] The methylation b values were converted to M values according to the following formula 1 for statistical analysis.
[0053] [Formula 1]
[0054] M value = log((b-value) / (1-(b-value)))
[0055] The M value was used for consensus clustering, and the b value was used for heatmap visualization and boxplot generation.
[0056] 5. Consistency Clustering
[0057] ConsensusClusterPlus (https: / / bioconductor.org / packages / release / bioc / vignettes / ConsensusClusterPlus / inst / doc / ConsensusClusterPlus.pdf) was used for consensus clustering to identify HCC subgroups. The top 3000 most variable probes (median absolute deviation (MAD)) (MAD 3K) were used as input.
[0058] 6. Selection of prognostic markers
[0059] The random forest method is based on a machine learning algorithm developed for modeling nonlinear effects. The inventors used the random forest R package of the R platform to find probes that can distinguish HCC groups with good and poor prognosis, and randomly divided HCC tumor samples into 4 training cohorts and 1 test cohort for five-fold cross validation. As a result of the execution, the inventors established a random forest model in which the expected ratio of each group was predicted by the confusion matrix (predicted samples vs observed samples), and the performance of each model was determined based on the area under the curve (AUC), sensitivity, and specificity values.
[0060] The present inventors measured the importance value of each probe in consideration of the Gini index, the overall reduction of probe impurities, and the dot plot of variable importance measured by random forest. As a result, the present inventors found the top 50 probes ( Figure 4 ). Hereinafter, when comparing the performance efficiency of 1 probe and the top 50 probes, the minimum number of probes with the highest efficiency was confirmed by considering the AUC value.
[0061] 7. Gene Ontology and Pathway Analysis
[0062] The inventors used innateDB (http: / / www.innatedb.com / ) to perform various pathway searches and gene ontology analyses. In innateDB, gene ontology biological processes, KEGG (Kyoto Encyclopedia of Genes and Genomes), and Reactome resources were used. The significant cutoff criterion was p value < 0.05.
[0063] 9. Statistical Analysis
[0064] Using Pearson's chi(χ 2 ) test or Fisher's exact test were used for statistical analysis of categorical variables. The median recurrence-free survival (RFS) was calculated using the Kaplan-Meyer method, and the log-rank test was used for comparison between different groups.
[0065] Experimental Results
[0066] 1. Patient selection and basic characteristics
[0067] Of the 182 HCC patients, 42 patients who had HCC recurrence within 1 year after surgery were excluded from the analysis. From the methylation identification results of 140 HCC samples, it can be seen that HCC samples showed large methylation changes between samples, while normal tissues showed almost constant methylation ( Figure 1 ).
[0068] 2. Comparison of MG1 and MG2 methylation profiles
[0069] The results of classifying 140 HCC samples according to their DNA methylation patterns confirmed that they were divided into methylation group 1 (n=81) and group 2 (n=59) (hereinafter described as MG1 and MG2, respectively) ( Figure 2 In addition, the methylation profile of MG2 was most different from that of non-tumor liver tissue, while the methylation profile of MG1 was similar to that of non-tumor liver tissue ( Figure 2 B in Figure 1). It can be seen that the recurrence-free survival (RFS) of MG1 is longer, while that of MG2 is shorter, leading to a poor prognosis ( Figure 3 ).
[0070] This epigenetic change is related to signal transduction, MAPK (mitogen activated protein kinase) and G protein coupled receptor (GPCR) signaling pathways (Table 1). When the inventors interpreted a series of genes based on the GO (Gene Ontology) system of this group, the top 50 probes were related to catabolism, neurotransmitter secretion, molecular function regulation, Wnt signaling pathway, cell migration regulation, mRNA metabolism, small GTPase-mediated signal transduction, gene expression regulation and cell apoptosis process regulation.
[0071] [Table 1]
[0072]
[0073]
[0074] 3. Selection of biomarkers for predicting HCC recurrence
[0075] Using the variable importance plot (varImpPlot), the top 50 probes that can classify HCC samples into MG1 and MG2 were found ( Figure 4 ). Below, while comparing the performance efficiency according to the number of probes, the minimum number of probes with the maximum efficiency is selected considering the AUC (area under the curve) value. From the results, it can be seen that when the number of probes is three or more, the accuracy of liver cancer recurrence prediction converges to 1 ( Figure 5 and Table 2). The letters described in Table 2 below represent accuracy (Accu.), sensitivity (Sen.), and specificity (Spec.), respectively.
[0076] [Table 2]
[0077]
[0078]
[0079] Based on the results of Table 2 above, the probes described in Table 3 below were finally selected, where the probes are recorded in descending order of importance value.
[0080] [Table 3]
[0081]
[0082]
[0083] The sequences of the probes described in Table 3 are reported in Table 4 below.
[0084] [Table 4]
[0085]
[0086] *In genomic sequences, [CG] indicates the methylation site targeted by the relevant CG marker
[0087] 3. Validation of HCC recurrence prediction markers
[0088] From the results of identifying the methylation levels of the probes described in Table 3 using the random forest method, it can be confirmed that the methylation level in normal liver tissue (green) is close to 100% and is at a constant level, and MG1 (red) shows a methylation level similar to that of normal liver tissue, while the methylation level of MG2 (gray) is about 50% or lower compared with that of normal liver tissue ( Figure 6 ; The X-axis is the selected marker, and the Y-axis is the b value indicating the methylation level).
[0089] Furthermore, 140 HCC samples were randomly divided into training data and validation data, and according to the test results of testing again with the probes in Table 3, it was confirmed that the existing MG1 was reclassified as MG1, and the existing MG2 was also reclassified as MG2 (Table 5).
[0090] [Table 5]
[0091]
[0092] The liver cancer cohort data generated by the inventors were verified by the probes in Table 3. The results showed that the liver cancer cohort data were divided into two groups (Pre_MG1 and Pre_MG2) according to the methylation level, and the recurrence rate of Pre_MG2 was higher ( Figure 7 In addition, the results of confirming RFS confirmed that the RFS of Pre_MG1 was significantly longer than that of Pre_MG2 (p=0.0136; Figure 7 B). This result means that the probe of the present invention can successfully predict the recurrence of liver cancer.
[0093] In addition, based on the validation results of TCGA (Cancer Genome Atlas) liver cancer data (an open database) in the same manner using the probes in Table 3, it can be confirmed that liver cancer patients are divided into two groups (Pre_MG1 and Pre_MG2) ( Figure 8 A in ), and these groups had different prognoses ( Figure 8 B). <110> ROPEDI CO., LTD. <120> DNA methylation markers for predicting liver cancer recurrence and uses thereof <130> G21E10C0523P / CN <150> KR 10-2019-0024379 <151> 2019-02-28 <160> 4 <170> KoPatentIn 3.0 <210> 1 <211> 122 <212> DNA <213> Artificial sequence <220> <223> Synthetic polynucleotides <400> 1 ggcctgccag agcggtggct gggtggccag gacctgtggg gcaggtcgga tgggcggcgg 60 cgcctggcgg atcagcggcg gggcctggcg gatcacaggt ggagcctggc ggatcacagg 120 tg 122 <210> 2 <211> 122 <212> DNA <213> Artificial sequence <220> <223> Synthetic polynucleotides <400> 2 ccttcagaga ggaagaatct taaggcgcgt ccaagacaag cggagtcgag gactccgttt 60 cggggaggtt tacctgaaaa gctgtgtgga aaaacccgac tgcgactttc aagcttgtgc 120 tt 122 <210> 3 <211> 122 <212> DNA <213> Artificial sequence <220> <223> Synthetic polynucleotides <400> 3 ctgcaacatc accgacaaag agcatgaaaa tgtgaatgac aggactctaa aaagaacgca 60 cgtttactcc gagagcggaa acaagaaggc gggtgcggcc tcactccttc ccggccaggc 120 ac 122 <210> 4 <211> 122 <212> DNA <213> Artificial sequence <220> <223> Synthetic polynucleotides <400> 4 cttctgtgag ggggacgcag ccgggcctcc cagacgcccg cgttccttac gtgaggggga 60 cgcagccggg ccttccgggc cttcccaagt cccgcgttcc ttctgtgagg cggacgcagc 120 cg 122
Claims
1. A composition for predicting recurrence of hepatocellular carcinoma, comprising a reagent for detecting the methylation level of the DNA methylation marker represented by SEQ ID NO:
2.
2. The composition according to claim 1, wherein the composition further comprises a reagent for detecting the methylation level of one or more DNA methylation markers represented by a sequence selected from the group consisting of SEQ ID NO: 1, SEQ ID NO: 3 and SEQ ID NO:
4.
3. The composition according to claim 1, wherein the composition predicts the prognosis of hepatocellular carcinoma recurrence in the normal liver of a hepatocellular carcinoma patient after treatment.
4. Use of a reagent for detecting the methylation level of a DNA methylation marker represented by SEQ ID NO: 2 in the preparation of a kit for implementing a method for predicting the recurrence of hepatocellular carcinoma, the method comprising the following steps: (a) isolating genomic DNA from a liver tissue sample to be analyzed; (b) detecting the methylation level of the DNA methylation marker represented by SEQ ID NO: 2; and (c) Comparison of the methylation levels of the detected DNA methylation markers with those detected in normal liver tissues.
5. The use according to claim 4, wherein the method for predicting recurrence of hepatocellular carcinoma further identifies the methylation level of one or more DNA methylation markers represented by a sequence selected from the group consisting of SEQ ID NO: 1, SEQ ID NO: 3 and SEQ ID NO: 4 in step (b).
6. The method according to claim 4, wherein the method for predicting the recurrence of hepatocellular carcinoma further comprises, after step (c), a step of determining that the possibility of hepatocellular carcinoma recurrence is higher when the methylation level of the DNA methylation marker identified in the liver tissue to be analyzed is lower than the methylation level of normal liver tissue.
7. The use according to claim 4, wherein the method predicts the prognosis of hepatocellular carcinoma recurrence in the normal liver of a hepatocellular carcinoma patient after treatment.
