A combination of biomarkers, a kit, an assessment system, and an assessment method for prognostic assessment of glioblastoma.

By combining biomarkers from the FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 genes with RT-qPCR detection, the problem of high external validation and testing costs in existing glioblastoma prognostic assessments has been solved, achieving stable and accurate prognostic risk stratification and personalized treatment support.

CN123108149APending Publication Date: 2026-11-17FIRST AFFILIATED HOSPITAL OF XINJIANG MEDICAL UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611362555.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-03
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing prognostic assessment methods for glioblastoma lack external validation, are costly to test, and are difficult to translate into clinical testing. Existing multi-gene models have low universality and are difficult to comprehensively and accurately reflect the prognostic risk of patients.

Method used

A biomarker combination consisting of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 genes was used for prognostic assessment via RT-qPCR detection combined with a risk scoring model. Kits and assessment systems are provided.

Benefits of technology

It achieves stable and accurate prognostic risk stratification for glioblastoma patients, provides reliable quantitative evidence, supports individualized treatment and follow-up management, and has external validation and low-cost testing capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123108149A_ABST
    Figure CN123108149A_ABST
Patent Text Reader

Abstract

This invention discloses a biomarker combination, kit, assessment system, and assessment method for prognostic assessment of glioblastoma. The biomarker combination consists of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 genes, and the expression levels of each gene are detected by RT-qPCR. This invention addresses the problems of insufficient predictive accuracy of existing glioblastoma prognostic assessments relying on single clinical indicators or a small number of molecular markers, poor generalization ability of traditional multi-gene models, reliance on high-cost whole transcriptome sequencing which is difficult to translate clinically, and the lack of a supporting, practical, low-cost RT-qPCR detection system. It achieves stable and accurate prognostic risk stratification for glioblastoma patients, providing reliable quantitative evidence for individualized treatment and postoperative follow-up.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to tumor molecular diagnostic technology, and in particular to a combination of biomarkers, a kit, an assessment system, and an assessment method for prognostic assessment of glioblastoma. Background Technology

[0002] Glioblastoma (GBM) is one of the most malignant primary tumors of the central nervous system and the most common and most malignant primary brain tumor in adults. It is highly aggressive, significantly heterogeneous, and has an extremely poor prognosis, classified as WHO grade IV. Despite advancements in surgical, radiotherapy, and chemotherapy techniques in recent years, the prognosis for GBM patients remains extremely poor, with a median survival of only 12-15 months and a 5-year survival rate of less than 10%. Significant differences exist in treatment response and survival outcomes among different patients. Therefore, establishing a molecular subtyping tool that can accurately assess the prognosis of GBM patients is crucial for clinical risk stratification, individualized treatment decisions, and follow-up management. Current clinical prognostic assessments for GBM primarily rely on clinicopathological parameters such as patient age, tumor resection extent, Karnofsky functional status score, and a few molecular markers such as IDH mutation status and MGMT promoter methylation status. However, due to the high heterogeneity and aggressiveness of GBM, existing single or limited markers are insufficient to comprehensively and accurately reflect the prognostic risk of patients.

[0003] In recent years, based on the development of high-throughput sequencing technology (NGS) and public tumor databases, the construction of multi-gene joint prognostic models based on transcriptome expression profiles has become an important direction for precise tumor stratification. By integrating the expression information of multiple genes that are significantly related to prognosis, the problem of insufficient stability of single-gene markers can be overcome to a certain extent, and the risk prediction ability can be improved. However, the existing multi-gene prognostic models for GBM reported in public still have the following shortcomings: (1) lack of external validation: most models are only constructed based on a single dataset (such as TCGA), lacking validation from independent external cohorts and real clinical samples, resulting in low universality of the models; (2) the models rely on whole transcriptome sequencing (RNA Sequencing) platforms, which have high detection costs and are difficult to translate into clinical applications; (3) few existing solutions further transform the bioinformatics modeling results into reverse transcription quantitative PCR (RT-qPCR) detection systems and matching kits that can be used for clinical testing.

[0004] Therefore, developing a GBM prognostic assessment system and kit that has been validated with external datasets, can be implemented via RT-qPCR, and has good clinical operability and industrialization potential has significant clinical value and market prospects. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a combination of biomarkers, reagent kits, assessment systems and assessment methods for prognostic assessment of glioblastoma, which can achieve stable and accurate prognostic risk stratification for glioblastoma patients and provide reliable quantitative basis for individualized treatment and postoperative follow-up.

[0006] The technical problem to be solved by the present invention is achieved through the following technical solution:

[0007] A biomarker suite for prognostic assessment of glioblastoma, consisting of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 genes.

[0008] Preferably, in the above technical solution, the biomarker combination is used to detect the expression level of each gene by RT-qPCR.

[0009] The application of a reagent for detecting the combination of said biomarkers in the preparation of a glioblastoma prognostic assessment product, wherein said reagent is a specific primer, probe, reverse transcription reagent or PCR amplification reagent, and said product is used for prognostic risk stratification of glioblastoma patients, assisting in the development of individualized treatment plans and postoperative follow-up assessment.

[0010] The application of a combination of biomarkers in the preparation of a glioblastoma prognostic assessment product, wherein the product is an RT-qPCR detection kit, a prognostic assessment calculation system, or a prognostic assessment kit, and the product is used to detect the expression levels of 7 genes and complete the prognostic risk stratification of glioblastoma patients.

[0011] A kit for prognostic assessment of glioblastoma, the kit comprising:

[0012] Specific primers and probes for detecting the expression levels of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 genes, the sequences of which are shown in SEQ ID NO:1-21;

[0013] Primers and probes used to detect the internal reference gene have sequences as shown in SEQ ID NO:22-24;

[0014] Reverse transcription reagent;

[0015] PCR amplification reaction system.

[0016] Preferably, in the above technical solution, the internal reference gene is one or more of GAPDH, ACTB, or 18S rRNA.

[0017] Preferably, the above technical solution further includes RNA extraction reagents, cDNA synthesis reagents, a risk scoring analysis program, an interpretation standard file, and a result reporting module. The risk scoring analysis program is used to automatically calculate risk scores and output grouping results; the interpretation standard file contains threshold determination rules; and the result reporting module is used to generate a prognostic assessment report.

[0018] A prognostic assessment system for glioblastoma includes:

[0019] The sample processing module is used to obtain glioblastoma tissue samples and extract total RNA, and then reverse transcribe it to obtain cDNA.

[0020] The gene expression detection module is used to detect the expression levels of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 by qPCR.

[0021] The data standardization module is used to standardize the target gene expression data using internal reference genes to obtain the relative expression level of the target gene.

[0022] The risk score calculation module is used to calculate the risk score based on the relative expression level of the target gene and its corresponding regression coefficient.

[0023] The prognostic module is used to classify subjects into high-risk or low-risk groups for glioblastoma based on a preset threshold (median risk score of training set samples) and output the corresponding prognostic assessment results.

[0024] A method for prognostic assessment of glioblastoma includes the following steps:

[0025] (1) Obtain tumor tissue samples from patients with glioblastoma and extract total RNA;

[0026] (2) Reverse transcribe total RNA into cDNA;

[0027] (3) Using cDNA as a template, the expression levels of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 were detected by RT-qPCR, and the expression levels of internal reference genes were also detected.

[0028] (4) Calculate the ΔCt value based on the Ct value of the target gene and the Ct value of the internal reference gene, and establish a survival risk prediction model based on the relative expression level of RT-qPCR using the -ΔCt method:

[0029] risk_score=FCN1*(-0.0146)+TPTEP1*0.0773-NPR3*0.1029-CHIT1*0.1445+TSHR*0.0865+IGLV3-21*0.1314-LINC02742*0.0512,

[0030] Among them, FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 represent the relative expression levels of the corresponding target genes; risk_score is the survival risk score for GBM patients;

[0031] (5) Based on the median risk score in the training set, a preset threshold is obtained and the subjects are divided into high-risk group and low-risk group. The high-risk group indicates a poor prognosis, while the low-risk group indicates a relatively good prognosis.

[0032] Preferably, in the above technical solution, the internal reference gene is one or more of GAPDH, ACTB, or 18S rRNA; the ΔCt value = target gene Ct value - internal reference gene Ct value.

[0033] The above-described technical solution of the present invention has the following beneficial effects:

[0034] This invention systematically analyzes transcriptomic and clinical survival data of GBM patients in publicly available databases, screens out key genes significantly associated with patient prognosis, and constructs a multi-gene combined risk model including FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742. Furthermore, this invention establishes an expression quantification method suitable for clinical sample detection based on RT-qPCR, and combines it with model algorithms to achieve stratified assessment of prognostic risk in GBM patients. The advantages are: (1) Multidimensional predictive indicators. The combined model constructed using seven prognostic-related genes has higher stability and accuracy than single biomarkers; (2) Facilitates clinical translation. The seven genes in the model can be rapidly and cost-effectively detected via RT-qPCR, making it suitable for developing GBM prognostic assessment kits; (3) Provides external validation to support predictive efficacy. In the training set, the AUCs for 1-year, 3-year, and 5-year survival predictions reached 0.81, 0.83, and 0.86, respectively; in the independent validation set CGGA, the AUCs for 1-year, 3-year, and 5-year survival predictions were 0.57, 0.61, and 0.66, respectively, demonstrating that the model has good cross-platform and cross-population generalization ability; (4) It has a basis for validation with real samples. In this invention, 39 clinical samples of GBM were further collected to carry out 7-gene RT-qPCR experiments, providing an experimental basis for the biological rationality of the model and its transformation into a clinical testing system. Attached Figure Description

[0035] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments of the invention and, together with their description, serve to explain the principles of the invention.

[0036] Figure 1 This is a gene expression volcano plot of 166 GBM samples from the TCGA training set of this invention;

[0037] Figure 2 The ROC curve of the Cox prognostic risk model in the TCGA training set of this invention;

[0038] Figure 3 The survival differences in the TCGA training set of this invention were compared based on the risk grouping of the Cox model (P<0.0001).

[0039] Figure 4 This is a forest plot of a 7-gene multivariate Cox regression analysis of the present invention;

[0040] Figure 5 The ROC curve of the Cox prognostic risk model on the CGGA validation set of this invention;

[0041] Figure 6 The CGGA validation set of this invention was used to compare survival differences based on risk grouping using the Cox model (P=0.011).

[0042] Figure 7 SPSS graph showing the difference in survival risk values ​​between the long OS group and the short OS group of GBM patients in the Logistic model of 39 RT-qPCR clinical samples of this invention. Detailed Implementation

[0043] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention.

[0044] Unless otherwise specified, the experimental methods used in the following examples are conventional methods, and the materials and reagents used are commercially available. Unless otherwise specified, the equipment used in the experiments is well known to those skilled in the art.

[0045] This invention provides a prognostic assessment system for glioblastoma based on a 7-gene combined model. The 7 genes include FCN1 (Gene ID: 2219), TPTEP1 (Gene ID: 387590), NPR3 (Gene ID: 4883), CHIT1 (Gene ID: 1118), TSHR (Gene ID: 7253), IGLV3-21 (Gene ID: 28796), and LINC02742 (Gene ID: 105376603). By detecting the expression levels of these 7 genes in subject samples and combining them with a pre-constructed risk scoring model, the system performs prognostic assessment and risk stratification for GBM patients, thereby providing auxiliary evidence for clinical treatment decisions and individualized interventions.

[0046] Table 1: Information on 7 target genes

[0047]

[0048] Example 1: Construction of a GBM prognostic prediction model

[0049] This embodiment illustrates the method for constructing the 7-gene combined prognostic model of GBM according to the present invention. Specifically, it includes the following steps:

[0050] First, GBM transcriptome data and corresponding clinical information were downloaded from the TCGA database. Primary tumor samples with survival information were screened, and samples with an overall survival of less than 30 days were excluded. Based on overall survival, the top one-third of the samples were selected as the short survival group (survival <189.6667 days, 83 cases), and the bottom one-third were selected as the long survival group (survival >411.3333 days, 83 cases), for a total of 166 samples included in the analysis.

[0051] Using the short group as the control group and the long group as the experimental group, after filtering out low-expression genes, differential expression analysis was performed using DESeq2, resulting in 91 differentially expressed genes, of which 29 were upregulated and 62 were downregulated. The TPM expression data of these 91 differentially expressed genes were further extracted from 166 samples and standardized using log2(x+1).

[0052] The standardized expression matrix was combined with survival information for LASSO-Cox analysis with parameters set to nfolds=10 and alpha=0.7, resulting in 10 candidate genes: FCN1, TPTEP1, NPR3, CHIT1, CCKAR, CXCL5, TSHR, MIR1-1HG-AS1, IGLV3-21, and LINC02742. Univariate Cox analysis was then performed on these 10 genes, and all results were statistically significant. These genes were then included in a multivariate Cox analysis, and a two-way stepwise regression was used to select the optimal variable set. Finally, a 7-gene multivariate Cox prognostic model consisting of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 was constructed.

[0053] Based on the above model, patient risk scores were calculated, and the model's predictive efficacy was evaluated. The results showed that the model's AUCs for 1-year, 3-year, and 5-year survival predictions in the training set were 0.81, 0.83, and 0.86, respectively, with a C-index of 0.707625, indicating that the 7-gene combined model has good prognostic predictive ability. Furthermore, transcriptome sequencing samples from the CGGA database were used as an independent validation set. Following the same clinical screening principles as the TCGA training set, 624 samples were obtained, and log2(TPM+1) expression data were used for risk score calculation and model validation. The results showed that the model's AUCs for 1-year, 3-year, and 5-year survival predictions in the CGGA validation set were 0.57, 0.61, and 0.66, respectively, indicating that the model has a certain degree of external generalization ability.

[0054] The prognostic assessment system of the present invention includes a sample processing module, a gene expression detection module, a data standardization module, a risk score calculation module, and a prognostic determination module.

[0055] The sample processing module is used to obtain GBM tissue samples from subjects and extract total RNA, and then perform reverse transcription to obtain cDNA; the gene expression detection module uses qPCR to detect the expression levels of the above 7 target genes; the data standardization module is used to standardize the target gene expression data (i.e., target gene ct values) with internal reference genes to obtain the relative expression level of the target gene; the risk score calculation module is used to calculate the risk score based on the expression level of the target gene and its corresponding regression coefficient; and the prognosis determination module is used to divide subjects into GBM high-risk group or low-risk group according to preset thresholds and output the corresponding prognostic assessment results.

[0056] This invention calculates the risk score by multiplying the standardized expression values ​​of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 by their corresponding multivariate Cox regression coefficients and summing the results. The risk score formula can be expressed as: ln(risk_score) = FCN1 * 0.3591658 - TPTEP1 * 0.2453271 + NPR3 * 0.7302472 + CHIT1 * 0.1595180 - TSHR * 0.2433723 + IGLV3-21 * 0.1842869 + LINC02742 * 0.5532045, where the gene name represents the TPM expression level of each gene, and ln(risk_score) is the natural logarithm of the risk score. This model is only used to validate the long-term prognostic predictive value of the seven gene combinations and is not intended for use in clinical RT-qPCR testing systems.

[0057] Furthermore, based on the median risk score in the training set, a preset threshold was obtained, and the subjects were divided into a high-risk GBM group and a low-risk group, with the high-risk group indicating a poor prognosis and the low-risk group indicating a relatively good prognosis.

[0058] Example 2: RT-qPCR gene expression level validation experiment in 39 real GBM tumor samples

[0059] This embodiment is used to illustrate the clinical sample validation of the 7-gene combined prognostic model described in this invention.

[0060] First, tumor tissue samples were collected from 39 GBM patients. Patients were divided into long OS and short OS groups based on the median overall survival (OS / month) after surgery. Patients with OS > 9.3 months were assigned to the long OS group (value = 0), and those with OS ≤ 9.3 months were assigned to the short OS group (value = 1). Total RNA (R4130, Meiji Biotechnology) was extracted using a kit and reverse transcribed into cDNA (RR037, Takara). This cDNA was then used as a template for qPCR to detect the expression levels of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 (RR390, Takara). Simultaneously, the internal reference gene GAPDH was detected, and fluorescence signals were collected. The Ct values ​​of the target genes were obtained based on RT-qPCR detection and standardized using the internal reference gene to calculate the ΔCt value. Furthermore, a risk scoring model based on the RT-qPCR detection system was established using the -log transformed relative expression index, -ΔCt, as the model input variable.

[0061] Table 2: Primer / probe sequences used in RT-qPCR validation experiments

[0062]

[0063] The sequences in Table 2 are numbered from top to bottom as follows: SEQ ID NO.1-24.

[0064] Example 3

[0065] This embodiment illustrates the ability of the 7-gene combined prognostic model of the present invention to discriminate overall survival groups in clinical samples. The present invention further establishes a short-term prognostic risk prediction model based on the relative expression levels of the seven target genes obtained from an RT-qPCR detection platform using a logistic regression method. Using overall patient survival as the clinical outcome, patients are divided into different prognostic groups according to preset time thresholds, and a comprehensive risk score is calculated using RT-qPCR expression data to predict adverse short-term prognoses in GBM patients.

[0066] Based on the relative expression levels of the seven target genes in the above 39 samples, a short-term prognostic risk prediction model was established using the LASSO Logistic regression method. The Logistic coefficients for each gene were -0.0146, 0.0773, -0.1029, -0.1445, 0.0865, 0.1314, and -0.0512, respectively. Therefore, the standardized survival risk score formula is: rs = FCN1*(-0.0146) + TPTEP1*0.0773 - NPR3*0.1029 - CHIT1*0.1445 + TSHR*0.0865 + IGLV3-21*0.1314 - LINC02742*0.0512. The difference in rs between the two groups was compared using an independent samples t-test based on the patients' overall survival (OS). SPSS analysis showed a statistically significant difference in rs between the long OS group and the short OS group (P=0.044), indicating that the RT-qPCR scoring model can significantly distinguish patients with different survival outcomes in the current exploratory cohort and has potential value in predicting the survival risk of GBM patients. The risk score formula established by the logistic regression method uses -ΔCt obtained from RT-qPCR as input and is independent of the Cox model in Example 1.

[0067] The above results further demonstrate that the 7-gene combined prognostic assessment system and kit provided by the present invention have good clinical application value and can be used for prognostic stratification and risk assessment of GBM patients.

[0068] Table 3: Grouping and rs values ​​of the 39 GBM samples included

[0069]

[0070]

[0071] The kit includes: specific primers and probes for detecting the expression levels of 7 genes and an internal reference gene, reverse transcription reagents, and a PCR amplification reaction system. Preferably, the kit may also include RNA extraction reagents, cDNA synthesis reagents, a risk scoring analysis program, interpretation criteria, and a result reporting module. The internal reference gene may be one or more of GAPDH, ACTB, and 18S rRNA. Preferably, the Ct values ​​of each target gene and the internal reference gene are detected using RT-qPCR, and the ΔCt value is calculated based on the Ct values ​​of the target gene and the internal reference gene. The standardized expression level of each target gene is further obtained using the -ΔCt method. The relative expression level can be directly used as the expression input value of each gene in the risk scoring formula, substituted into the risk scoring model based on the RT-qPCR detection system, for risk scoring calculation and prognostic assessment.

[0072] The system or kit described in this invention can be applied to prognostic assessment, risk stratification, treatment planning assistance, and follow-up management of GBM patients. Preferably, by processing the RT-qPCR detection results of seven target genes in clinical samples and substituting them into the risk scoring formula, a quantitative assessment of the prognostic risk of subjects can be achieved. This invention is based on modeling using public databases and combines independent validation sets and clinical sample detection results, demonstrating good clinical application potential and translational value.

[0073] Although the present invention has been disclosed above with reference to embodiments, it is not intended to limit the present invention. Any person skilled in the art can make various different choices and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention is defined by the claims and their equivalents.

Claims

1. A combination of biomarkers for prognostic assessment of glioblastoma, characterized in that, It is composed of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 genes.

2. The biomarker combination according to claim 1, characterized in that, The biomarker combination is used to detect the expression levels of each gene by RT-qPCR.

3. The use of a reagent for detecting the combination of biomarkers described in claim 1 in the preparation of a glioblastoma prognostic assessment product, characterized in that, The reagents are specific primers, probes, reverse transcription reagents, or PCR amplification reagents. The products are used for prognostic risk stratification of glioblastoma patients, assisting in the development of individualized treatment plans, and postoperative follow-up assessment.

4. The application of the biomarker combination according to claim 1 in the preparation of glioblastoma prognostic assessment products, characterized in that, The product is an RT-qPCR detection kit, a prognostic assessment calculation system, or a prognostic assessment kit. The product is used to detect the expression levels of 7 genes and complete the prognostic risk stratification of glioblastoma patients.

5. A kit for prognostic assessment of glioblastoma, characterized in that, The kit includes: Specific primers and probes for detecting the expression levels of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 genes, the sequences of which are shown in SEQ ID NO:1-21; Primers and probes used to detect the internal reference gene have sequences as shown in SEQ ID NO:22-24; Reverse transcription reagent; PCR amplification reaction system.

6. The reagent kit according to claim 5, characterized in that, The internal reference gene is one or more of GAPDH, ACTB, or 18S rRNA.

7. The reagent kit according to claim 5, characterized in that, It also includes RNA extraction reagents, cDNA synthesis reagents, as well as risk scoring analysis procedures, interpretation criteria documents, and result reporting modules.

8. A prognostic assessment system for glioblastoma, characterized in that, include: The sample processing module is used to obtain glioblastoma tissue samples and extract total RNA, and then reverse transcribe it to obtain cDNA. The gene expression detection module is used to detect the expression levels of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 by qPCR. The data standardization module is used to standardize the target gene expression data using internal reference genes to obtain the relative expression level of the target gene. The risk score calculation module is used to calculate the risk score based on the relative expression level of the target gene and its corresponding regression coefficient. The prognosis module is used to classify subjects into high-risk or low-risk groups for glioblastoma based on preset thresholds and output the corresponding prognostic assessment results.

9. A method for prognostic assessment of glioblastoma, characterized in that, Includes the following steps: (1) Obtain tumor tissue samples from patients with glioblastoma and extract total RNA; (2) Reverse transcribe total RNA into cDNA; (3) Using cDNA as a template, the expression levels of FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21 and LINC02742 were detected by RT-qPCR, and the expression levels of internal reference genes were also detected. (4) Calculate the ΔCt value based on the Ct value of the target gene and the Ct value of the internal reference gene. Use the relative expression index -ΔCt after -log transformation as the input variable, and establish a survival risk prediction model based on RT-qPCR expression data using logistic regression: risk_score= FCN1*(-0.0146)+TPTEP1*0.0773-NPR3*0.1029-CHIT1*0.1445+TSHR*0.0865+IGLV3-21*0.1314-LINC02742*0.0512, Where FCN1, TPTEP1, NPR3, CHIT1, TSHR, IGLV3-21, and LINC02742 represent the relative expression levels of the corresponding target genes calculated by RT-qPCR; risk_score is the survival risk score. (5) Based on the patient's total survival time, the patients were divided into a short survival group (OS<9 months) and a long survival group (OS≥9 months), and the rs score was calculated for each group. The high-risk group indicated a poor prognosis, while the low-risk group indicated a relatively good prognosis.

10. The method according to claim 9, characterized in that, The internal reference gene is one or more of GAPDH, ACTB, or 18S rRNA; the ΔCt value = target gene Ct value - internal reference gene Ct value.