阅读说明：本技术 基于基因检测判断肝癌药物敏感性和远期预后的预测工具及其应用 (Prediction tool for judging liver cancer drug sensitivity and long-term prognosis based on gene detection and application thereof ) 是由 徐俊杰 蔡秀军 潘宇 梁霄 夏顺杰 于 2021-01-21 设计创作，主要内容包括：本发明公开了一种基于基因检测判断肝癌药物敏感性和远期预后的预测工具,本发明通过统计分析TCGA数据中的与肝癌预后相关有氧糖酵解通路基因,在此基础上采用LASSO回归分析简化预后相关基因,建立基于有氧糖酵解通路基因的预测工具,简称有氧糖酵解指数。将该指数于多个公共数据库及邵逸夫医院临床样本中验证,发现该指数可准确预测肝癌患者对索拉菲尼疗法的敏感和远期预后。本发明可以有效的筛选对索拉菲尼疗法敏感的肝细胞肝癌患者,为肝癌患者精准、综合治疗提供新思路。(The invention discloses a prediction tool for judging liver cancer drug sensitivity and prognosis in a long term based on gene detection, which is used for simplifying genes related to prognosis by adopting LASSO regression analysis on the basis of statistically analyzing aerobic glycolysis pathway genes related to liver cancer prognosis in TCGA data and establishing the prediction tool based on the aerobic glycolysis pathway genes, namely aerobic glycolysis indexes. The index is verified in a plurality of public databases and clinical samples of the Shore-Yi-Fu hospital, and the index can accurately predict the sensitivity and the long-term prognosis of the liver cancer patient to the Sorafenib therapy. The screening method can effectively screen the liver cell liver cancer patients sensitive to the Sorafenib therapy, and provides a new idea for accurate and comprehensive treatment of the liver cancer patients.)

1. A predictive tool for determining drug sensitivity and prognosis of liver cancer based on gene detection, wherein the predictive tool is aerobic glycolysis index, wherein the aerobic glycolysis index is LDHA gene expression level 0.163+ STC2 gene expression level 0.004+ GPC1 gene expression level 0.034+ TKTL1 gene expression level 0.0001+ SLC2A1 gene expression level 0.083+ SRD5A3 gene expression level 0.032+ PLOD2 gene expression level 0.070+ G6PD gene expression level 0.083+ HMMR gene expression level 0.040+ HOME 1 gene expression level 0.001+ RARS 2 gene expression level 0.132-GOT2 gene expression level 0.146+ CENPA 050.001-053.

2. The prediction tool of claim 1, wherein the method/technique for detecting the expression level of a gene comprises: second generation RNA sequencing or third generation RNA sequencing or gene chip technology.

3. Use of the prediction tool of claim 1 in determining the susceptibility of a patient to a liver cancer drug.

4. The application of claim 3, wherein the surfminer data packet is utilized to obtain an optimal threshold value of aerobic glycolysis index corresponding to a corresponding detection method of the data set according to the survival data of the data set, if the aerobic glycolysis index of the patient is higher than the threshold value, the sensitivity of the patient to the Sorafenib therapy is indicated to be poor, and otherwise, the sensitivity of the patient to the Sorafenib therapy is indicated to be good.

5. Use of the prediction tool of claim 1 for prognosis prediction of a patient with liver cancer.

6. The use of claim 5, wherein the survivor data of the data set is used to obtain the optimal threshold value of aerobic glycolysis index corresponding to the detection method corresponding to the data set, and if the aerobic glycolysis index of the patient is higher than the threshold value, the prognosis of the patient with liver cancer is poor, otherwise, the prognosis of the patient with liver cancer is good.

7. The use of any one of claims 5 to 6, wherein the liver cancer is hepatocellular carcinoma.

8. A kit for judging drug sensitivity and prognosis of liver cancer based on gene detection is characterized by comprising reagents for measuring the expression levels of LDHA gene, STC2 gene, GPC1 gene, TKTL1 gene, SLC2A1 gene, SRD5A3 gene, PLOD2 gene, G6PD gene, HMMR gene, HOMER1 gene, RARS1 gene, GOT2 gene, CENPA gene and SLC2A2 gene.

9. The kit of claim 8, wherein the reagent is a primer or probe that specifically binds to the gene.

Technical Field

The invention belongs to the fields of biotechnology and medicine, and particularly relates to gene detection related to antitumor drug resistance and application thereof.

Background

Liver cancer is the sixth most common malignant tumor in China and also worldwide, and ranks the fourth highest among tumor-related causes of death. Despite the great progress in therapeutic approaches, the five-year survival rate of liver cancer is still between 25% and 55%. Distant metastasis, intrahepatic recurrence, and poor sensitivity to various therapies are the main causes of poor prognosis of liver cancer. The gene mutation, chromosome abnormality and cell signal pathway abnormality are closely related to the occurrence and development of liver cancer. The method has the advantages that the liver cancer is classified through molecular biological characteristics, accurate treatment is realized, and the prognosis of a liver cancer patient is improved.

Aerobic glycolysis is a large hallmark of tumor malignancy, which mainly means that tumor cells obtain energy largely through glycolysis even at physiological oxygen concentrations. Through the change of the sugar metabolism mode, the tumor cells rapidly obtain energy and simultaneously produce a large amount of metabolites required by physiological synthesis. Furthermore, aerobic glycolysis is closely related to multiple oncogene signaling pathways. Thus, the molecular typing of liver cancer by aerobic glycolysis levels may reveal a new molecular classification of liver cancer.

Sorafenib is currently the first-line therapeutic for advanced liver cancer. However, the sorafenib drug resistance phenomenon is very common in clinic. How to screen out the patients sensitive to the sorafenib therapy, accurate medication is of great importance to improving the prognosis of the patients with liver cancer. Tumor metabolism, changes in tumor microenvironment, epigenetics, etc. are also considered to be likely associated with liver cancer sorafenib resistance. However, the leading mechanism or the key gene still is the main problem which puzzles the research of the liver cancer on the sorafenib drug resistance.

Therefore, there is an urgent need in the art to find a new method capable of predicting the sensitivity of liver cancer to sorafenib and the prognosis of liver cancer at a distant stage, by which precise treatment is achieved and the prognosis of patients is improved.

Disclosure of Invention

The invention aims to find a novel prediction tool for predicting the sensitivity and the long-term prognosis of liver cancer to sorafenib, aiming at the defects of the prior art.

The invention is realized by the following technical scheme:

1. statistically screening aerobic glycolysis pathway genes related to liver cancer prognosis in TCGA data by using a single-factor Cox regression model;

2. on the basis, LASSO regression analysis is adopted to simplify the relevant genes for prognosis, and a prediction tool based on the aerobic glycolysis pathway gene is established, which is called aerobic glycolysis index for short; aerobic glycolysis index (LDHA gene expression amount) 0.163+ STC2 gene expression amount 0.004+ GPC1 gene expression amount 0.034+ TKTL1 gene expression amount 0.0001+ SLC2a1 gene expression amount 0.014+ SRD5A3 gene expression amount 0.032+ PLOD2 gene expression amount 0.070+ G6PD gene expression amount 0.083+ HMMR gene expression amount 0.040+ gor 1 gene expression amount 0.001+ RARS1 gene expression amount 0.132-GOT2 gene expression amount 0.146+ CENPA gene expression amount 0.053-SLC2a2 gene expression amount.

The detection method/technology of the gene expression level comprises the following steps: second generation RNA sequencing or third generation RNA sequencing or gene chip technology.

3. The index is verified in a plurality of public databases and clinical samples of the Shore-Yi-Fu hospital, and the index can accurately predict the long-term prognosis of the liver cancer patient; and acquiring an optimal aerobic glycolysis index threshold corresponding to the corresponding detection method of the data set by using a surfminer data packet according to the survival data of the data set, wherein if the aerobic glycolysis index of the patient is higher than the threshold, the prognosis of the liver cancer patient is poor, and otherwise, the prognosis of the liver cancer patient is good.

4. The index is verified in GDSC and CCLE databases and clinical samples of STORM tests, the index is found to be negatively correlated with the sorafenib sensitivity, and the sensitivity of a patient to sorafenib therapy can be accurately predicted. And acquiring an aerobic glycolysis index optimal threshold corresponding to a corresponding detection method of the data set by using a survminer data packet according to the survival data of the data set, wherein if the aerobic glycolysis index of the patient is higher than the threshold, the patient is indicated to have poor sensitivity to the sorafenib therapy, and otherwise, the patient is indicated to have good sensitivity to the sorafenib therapy.

The invention also provides a kit for judging drug sensitivity and prognosis of liver cancer based on gene detection, which comprises reagents for measuring the expression quantity of LDHA gene, STC2 gene, GPC1 gene, TKTL1 gene, SLC2A1 gene, SRD5A3 gene, PLOD2 gene, G6PD gene, HMMR gene, HOMER1 gene, RARS1 gene, GOT2 gene, CENPA gene and SLC2A2 gene.

Preferably, the reagent is a primer or probe that specifically binds to the gene.

The invention has the beneficial effects that: the index of the invention is only based on 14 gene expression levels, and the method is simple, has high prediction accuracy, is easy to popularize and has very good clinical transformation value.

Drawings

The invention is further explained below with reference to the figures and examples;

FIG. 1 is a graph showing that single-factor Cox analysis suggests that 80 aerobic glycolysis-related genes are associated with liver cancer prognosis;

FIG. 2 is a simplified LASSO regression analysis of prognostic-related genes, establishing aerobic glycolysis indices based on 14 gene expression levels;

FIG. 3 is a graph of overall survival (a) and tumor-free survival (b) for liver cancer patients in a TCGA database for which the aerobic glycolytic index is predictive; in the figure, 2 represents the survival curve of low AGI, and 1 and 3 are error curves of the survival curve of low AGI, respectively; the high AGI survival curve is shown at 5, and the error bars at 4 and 6, respectively, are high AGI survival curves.

FIG. 4 is a ROC plot of TCGA-LIHC data;

FIG. 5 is a graph of the overall survival rate of a GSE14520(a) and LIRI-JP database (b) predicted by aerobic glycolysis index and a liver cancer patient (c) in the Shore-fugaf hospital; in the figure, 2 represents the survival curve of low AGI, and 1 and 3 are error curves of the survival curve of low AGI, respectively; the high AGI survival curve is shown at 5, and the error bars at 4 and 6, respectively, are high AGI survival curves.

FIG. 6 is a graph showing the negative correlation between sorafenib sensitivity and aerobic glycolysis index for hepatoma cell lines in GDSC (a) and CCLE database (b);

figure 7 is "STORM" clinical data showing that aerobic glycolysis index can predict response to sorafenib therapy.

FIG. 8 is a graph of AUC of "STORM" clinical data.

Detailed Description

The present invention is further illustrated by the following experiments in conjunction with examples, it being understood that these examples are for illustrative purposes only and in no way limit the scope of the present invention.

Sequencing and clinical data and reagent sources involved in this example:

TCGA-LIHC data are downloaded in UCSC database (https:// xenambrowser. net/datapages), LIRI-JP data are downloaded in HCCDB database (http://lifeome.net/database/hccdb/ download.html) GSE14520 and GSE109211 data are listed in GEO database (https:// www.ncbi.nlm.nih.gov/GEO /). hepatoma cell line sensitivity to Sorafenib data are listed in GDSC database (https:// www.cancerrxgene.org) and CCLE databasehttps:// portals.broadinstitute.org/ccle/data) The data of the Shao Yi Fu hospital are derived from 102 cases of clinic visits from 1 month 2008 to 1 month 2018 of the Shao Yi Fu hospital affiliated to the Zhejiang university medical school, liver cancer is confirmed in all the 102 cases of clinic visits, the TNM stage is I-IV, the T stage is T1-T4, the age is 32-88 years, and the clinic follow-up time is more than 2 years.

Example (b):

selecting the sequencing data and clinical follow-up information of 371 liver cancer patients in TCGA-LIHC data, and analyzing the influence of aerobic glycolysis genes on the total survival rate of the 371 patients through single-factor COX regression. The results show that a total of 80 genes significantly affected the overall survival rate of liver cancer patients, as shown in fig. 1.

Through LASSO regression analysis, the relevant genes were simplified and aerobic glycolysis indexes based on the expression levels of 14 genes were established, and as shown in fig. 2, the values were assigned specifically to Aerobic Glycolysis Index (AGI) ═ LDHA gene expression level 0.163+ STC2 gene expression level 0.004+ GPC1 gene expression level 0.034+ TKTL1 gene expression level 0.0001+ SLC2a1 gene expression level 0.014+ SRD5A3 gene expression level 0.032+ PLOD2 gene expression level 0.070+ G6PD gene expression level 0.083+ HMMR gene expression level 0.040+ HOMER1 gene expression level 0.001+ RARS1 gene expression level 0.132-GOT 2.001 + cea 2.

Cases can be grouped based on Aerobic Glycolysis Index (AGI), where the threshold for grouping is the point where the prognosis of the two separate groups of patients is most different, e.g., based on patient survival data, the optimal threshold is obtained using the R language software "surfmer" data packet, it is noted that the threshold may be different for different sequencing methods. The following is described in detail with reference to specific validation sets:

the influence of aerobic glycolysis indexes on the long-term prognosis of a liver cancer patient is verified in TCGA-LIHC data, namely, the expression level of each gene of a liver cancer tissue of the patient is detected through an Illumina HiSeq 2000RNA sequencing platform, the aerobic glycolysis indexes of the liver cancer patients are calculated after standardization processing, according to survival data of the patients, an R language software 'survminer' data packet is used, an optimal threshold value of 4.05 is selected, the aerobic glycolysis indexes are lower than 4.05, the aerobic glycolysis indexes are low aerobic glycolysis index groups (low AGI groups), and the aerobic glycolysis indexes are higher than 4.05, and the aerobic glycolysis indexes are high aerobic glycolysis index groups (high AGI groups). Among them, the aerobic glycolysis index was found by Kaplan-Meier survival curves and log-rank survival analysis to suggest a worse long-term prognosis for liver cancer patients in the high AGI group, including overall survival and tumor-free survival, as shown in FIG. 3.

Meanwhile, the ROC curve graph is used for evaluating the clinical accuracy of the model in the embodiment, the ROC curve graph is shown in figure 4, the abscissa is 1-specificity, the ordinate is sensitivity, and under the condition that the five-year survival rate is a node, when 4.05 is taken, the specificity is 0.65, the sensitivity is 0.69, the AUC value of the calculated model is 0.714, and the accuracy of the model prediction result is higher. The area under the ROC curve is between 1.0 and 0.5, with AUC greater than 0.5, the closer the AUC is to 1, indicating better diagnostic results.

Furthermore, COX regression analysis is adopted to verify relevant risk factors of aerobic glycolysis indexes on the long-term prognosis of the liver cancer patient in TCGA, and the multi-factor regression analysis finds that clinical indexes such as age (greater than or equal to 60 years old and control less than 60 years old), gender (male and control female), tumor differentiation degree (G3 grade, G2 grade and control G1 grade), tumor stage (IV stage, III stage and II stage and control I stage), blood vessel invasion (large blood vessel infiltration, micro infiltration and control non-infiltration) and the like are not independent risk factors of the long-term prognosis of the liver cancer patient, and the aerobic glycolysis indexes are independent risk factors of the long-term prognosis of the liver cancer patient, as shown in fig. 5. The results show that the aerobic glycolytic index of the invention can be used for independently predicting the long-term prognosis of liver cancer patients, and is not influenced by clinical indexes such as age, sex, tumor differentiation degree, tumor stage, vascular invasion and the like.

The influence of aerobic glycolytic indexes on the long-term prognosis of liver cancer patients is further verified in 243 liver cancer patients in a GSE14520 database, 200 liver cancer patients in a LIRI-JP database and 102 liver cancer patients in an Shaofu hospital, and similarly, the aerobic glycolytic indexes of the liver cancer patients are respectively detected by Affymetrix Human Genome U133A 2.0.0 Array (GSE14520), Illumina RNA-Seq (LIRI-JP) and Illumina (Shaofu hospital) sequencing platforms, after standardization treatment, the aerobic glycolytic indexes of the liver cancer patients are calculated and the optimal threshold values (3.245(GSE14520), 1.785(LIRI-JP) and 1.64 (Shaofu hospital)) are taken, wherein the aerobic glycolytic indexes are lower than the optimal threshold value and are respectively in a low AGI group, and the AGI is higher than the optimal threshold value and is respectively in a high AGI group. The aerobic glycolytic index suggested a worse overall survival rate for liver cancer patients in the high AGI group, as shown in figure 6.

The liver cancer cell line was suggested to be positively correlated with the aerobic glycolysis index for sorafenib IC50 concentration in the GDSC database. In the CCLE database, it was suggested that the concentration of sorafenib EC50 in the liver cancer cell line was positively correlated with the aerobic glycolysis index, as shown in fig. 7a, 7 b.

Among 67 patients with liver cancer who received sorafenib treatment in the "STORM" database, the aerobic glycolysis index was effective in predicting the sensitivity of the patients with liver cancer to sorafenib, with an area under the curve of 0.879, as shown in FIG. 8. The threshold 3.488 corresponds to a sensitivity of 0.905 and a specificity of 0.848.

The embodiment also provides a method for predicting the sensitivity of a patient to sorafenib therapy by using the method, which specifically comprises the following steps:

1. collecting tissue sample (such as operation specimen and puncture specimen) of liver cancer patient, and extracting total RNA in tissue.

2. And selecting a proper sequencing platform to detect the aerobic glycolysis index related gene, and calculating the aerobic glycolysis index.

3. And judging the aerobic glycolysis index level of the sample by referring to the optimal threshold value of the aerobic glycolysis index in the database according to the established aerobic glycolysis index database.

4. If the aerobic glycolysis index level of the detected sample is lower than the threshold value, the patient has better prognosis and is sensitive to the sorafenib therapy, and sorafenib adjuvant therapy can be performed to improve the prognosis. On the contrary, if the aerobic glycolysis index level of the detected sample is higher than the threshold value, the patient has poorer prognosis, is not sensitive to the sorafenib therapy and is not suitable for sorafenib adjuvant therapy.