Application of diagnostic marker combination in construction of esophageal squamous carcinoma prognosis risk assessment model

Abstract

The invention relates to application of a diagnostic marker combination in construction of a prognosis risk assessment model of esophageal squamous carcinoma. The diagnostic marker combination comprises SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1. The prognosis of the esophageal squamous carcinoma patient is predicted on the basis of detecting the diagnostic marker combination and a corresponding esophageal squamous carcinoma prognosis risk assessment model, effective guidance is provided for a clinician to make a treatment decision on the esophageal squamous carcinoma patient, invalid treatment is reduced, the treatment cost and uncomfortable experience of the patient are reduced, and the clinical application prospect is wide. And accurate prediction and individualized comprehensive treatment of prognosis of the esophageal squamous cell carcinoma patient can be promoted.

Description

Technical Field

[0001] The present invention belongs to the field of biomedicine and relates to the application of a diagnostic marker combination in constructing a prognostic risk assessment model for esophageal squamous cell carcinoma. Background Art

[0002] Esophageal cancer is the sixth leading cause of death worldwide and one of the most common digestive tract malignancies. Esophageal cancer cases in China account for more than 70% of the total global cases, of which esophageal squamous cell carcinoma is the main pathological type. Due to the lack of effective early diagnosis and treatment techniques, most patients are already in the middle and late stages when they are first diagnosed, and the 5-year survival rate of patients with esophageal squamous cell carcinoma in my country is still less than 20%. However, due to the high recurrence, metastasis and drug resistance of esophageal squamous cell carcinoma, patients often have a poor prognosis. Early differentiation of high-risk patients and individualized precision treatment are crucial to improving the clinical outcomes of esophageal squamous cell carcinoma.

[0003] At present, biomarkers such as TNM staging, PD-L1 expression, and tumor mutation load are mainly used in clinical practice to predict the prognosis of patients with esophageal squamous cell carcinoma and the response to immunotherapy. However, these single indicators cannot fully represent the diverse tumor microenvironment, nor do they fully consider the differences in the molecular mechanisms of esophageal squamous cell carcinoma. Therefore, they cannot accurately predict the prognosis of patients with esophageal squamous cell carcinoma and effectively distinguish patients who can benefit from immunotherapy. Therefore, there is an urgent need for a more comprehensive and reliable marker to accurately predict the survival rate and immunotherapy effect of patients with esophageal squamous cell carcinoma.

[0004] Tumor immunotherapy targeting immune checkpoints (ICIs) has significantly improved the clinical efficacy of patients and changed the treatment model of cancer patients. The tumor immune microenvironment has received widespread attention as an important component of immunotherapy. Studies have shown that the tumor immune microenvironment can be used as a major prognostic indicator to improve the efficacy of targeted therapy. Given the important role of T cells in immunity, genes related to T cell-mediated tumor killing have been discovered through clustered regularly interspaced short palindromic repeats (CRISPR) screening and high-throughput experimental methods. Therefore, based on genes related to T cell-mediated tumor killing, developing a prognostic model closely related to the tumor immune microenvironment is both appropriate and feasible for predicting the efficacy of immunotherapy.

[0005] However, biomarkers of genes related to T cell-mediated tumor killing have not yet been developed. Therefore, it is urgent to establish a more accurate and efficient prognostic prediction model based on T cell-mediated tumor killing in esophageal squamous cell carcinoma, effectively distinguish the prognosis of esophageal squamous cell carcinoma patients and accurately identify patients who benefit from tumor immunotherapy. Summary of the invention

[0006] In order to solve the above problems, the main purpose of the present invention is to provide an application of a diagnostic marker combination in constructing a prognostic risk assessment model for esophageal squamous cell carcinoma, thereby establishing an effective indicator for predicting the prognosis and clinical treatment selection of esophageal squamous cell carcinoma.

[0007] To solve the above technical problems, the present invention provides an application of a diagnostic marker combination in constructing a prognostic risk assessment model for esophageal squamous cell carcinoma, wherein the diagnostic marker combination is SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1.

[0008] As described above, the prediction model includes nucleic acid probes for detecting the expression of SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1;

[0009] Among them, the forward probe SLC2A1-F sequence for detecting SLC2A1 is shown in Seq ID No.1, and the reverse probe SLC2A1-R sequence for detecting SLC2A1 is shown in Seq ID No.2; the forward probe KIF11-F sequence for detecting KIF11 is shown in Seq ID No.3, and the reverse probe KIF11-R sequence for detecting KIF11 is shown in Seq ID No.4; the forward probe KCNMA1-F sequence for detecting KCNMA1 is shown in Seq ID No.5, and the reverse probe KCNMA1-R sequence for detecting KCNMA1 is shown in Seq ID No.6; the forward probe BARHL2-F sequence for detecting BARHL2 is shown in Seq ID No.7, and the reverse probe BARHL2-R sequence for detecting BARHL2 is shown in Seq ID No.8; the forward probe CA9-F sequence for detecting CA9 is shown in Seq ID No.9, and the reverse probe CA9-R sequence for detecting CA9 is shown in Seq ID No.10; the sequence of the forward probe TIMP1-F for detecting TIMP1 is shown in Seq ID No.11, and the sequence of the reverse probe TIMP1-R for detecting TIMP1 is shown in Seq ID No.12; the sequence of the forward probe MAGEC3-F for detecting MAGEC3 is shown in Seq ID No.13, and the sequence of the reverse probe MAGEC3-R for detecting MAGEC3 is shown in Seq ID No.14; the sequence of the forward probe PDZK1IP1-F for detecting PDZK1IP1 is shown in Seq ID No.15, and the sequence of the reverse probe PDZK1IP1-R for detecting PDZK1IP1 is shown in Seq ID No.16.

[0010] As described above, the model includes the following content: risk score = (0.0552283605 × KIF11 expression) + (-0.0070408017 × SLC2A1 expression) + (0.0194419232 × KCNMA1 expression) + (-0.0231625843 × BARHL2 expression) + (-0.0214342225 × CA9 expression) + (0.0161917634 × TIMP1 expression) + (-0.0712586182 × MAGEC3 expression) + (-0.0004431282 × PDZK1IP1 expression), when the risk score>0.61 is judged as a high-risk group, and the risk score<0.61 is judged as a low-risk group.

[0011] As described above, the model includes the following: patient risk score, clinical pathological characteristics including TNM stage, tumor grade, alcohol consumption, age, gender, and risk score are integrated into a scoring scale to further calculate the total score of the sample, and use a conversion function to estimate the survival probability of each patient at 1 year, 3 years, and 5 years.

[0012] The present invention also provides a detection kit, comprising nucleic acid probes and instructions for detecting the expression amount of the gene combination SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1;

[0013] Among them, the forward probe SLC2A1-F sequence for detecting SLC2A1 is shown in Seq ID No.1, and the reverse probe SLC2A1-R sequence for detecting SLC2A1 is shown in Seq ID No.2; the forward probe KIF11-F sequence for detecting KIF11 is shown in Seq ID No.3, and the reverse probe KIF11-R sequence for detecting KIF11 is shown in Seq ID No.4; the forward probe KCNMA1-F sequence for detecting KCNMA1 is shown in Seq ID No.5, and the reverse probe KCNMA1-R sequence for detecting KCNMA1 is shown in Seq ID No.6; the forward probe BARHL2-F sequence for detecting BARHL2 is shown in Seq ID No.7, and the reverse probe BARHL2-R sequence for detecting BARHL2 is shown in Seq ID No.8; the forward probe CA9-F sequence for detecting CA9 is shown in Seq ID No.9, and the reverse probe CA9-R sequence for detecting CA9 is shown in Seq ID No.10; the sequence of the forward probe TIMP1-F for detecting TIMP1 is shown in Seq ID No.11, and the sequence of the reverse probe TIMP1-R for detecting TIMP1 is shown in Seq ID No.12; the sequence of the forward probe MAGEC3-F for detecting MAGEC3 is shown in Seq ID No.13, and the sequence of the reverse probe MAGEC3-R for detecting MAGEC3 is shown in Seq ID No.14; the sequence of the forward probe PDZK1IP1-F for detecting PDZK1IP1 is shown in Seq ID No.15, and the sequence of the reverse probe PDZK1IP1-R for detecting PDZK1IP1 is shown in Seq ID No.16;

[0014] The instruction manual records the following content: Risk score = (0.0552283605×KIF11 expression level) + (-0.0070408017×SLC2A1 expression level) + (0.0194419232×KCNMA1 expression level) + (-0.0231625843×BARHL2 expression level) + (-0.0214342225×CA9 expression level) + (0.0161917634×TIMP1 expression level) + (-0.0712586182×MAGEC3 expression level) + (-0.0004431282×PDZK1IP1 expression level). When the risk score>0.61, it is judged to be a high-risk group, and when the risk score<0.61, it is judged to be a low-risk group.

[0015] The present invention also provides the use of the above detection kit in preparing a product for assessing the prognosis risk of patients with esophageal squamous cell carcinoma.

[0016] Preferably, the prognostic risk assessment of patients with esophageal squamous cell carcinoma refers to assessing the overall survival rate of patients with esophageal squamous cell carcinoma.

[0017] The present invention uses a diagnostic marker combination obtained from T cell killing-related genes to construct a prognosis-related model for esophageal squamous cell carcinoma. The marker combination used in this prognostic prediction model not only shows significant differential expression in esophageal squamous cell carcinoma, but is also closely related to the tumor killing function mediated by T cells. At present, the core of tumor immunotherapy is based on the killing function of T cells, and a variety of immunotherapy schemes have been developed around the killing function of T cells, such as PD1 / PD-L1 antibodies, CAR-T cell therapy, etc. Therefore, the prognostic model constructed based on T cell-mediated tumor killing function genes can more accurately predict the prognosis of patients with esophageal squamous cell carcinoma, and may also provide potential guidance for patients with esophageal squamous cell carcinoma receiving immunotherapy, providing new possibilities for the personalization and accuracy of treatment plans.

[0018] Compared with the existing TNM staging system prognosis assessment method, the diagnostic marker combination provided by the present invention combined with the prognosis prediction model is not limited by the pathological characteristics of the tumor and the limited variables related to the tumor, lymph nodes and metastasis; secondly, it is not limited by the correlation between the anatomical progression of the disease and the progression of the stage; most importantly, the present invention fully considers the complex molecular mechanism of tumor diseases and the genetic and molecular characteristics of individuals, can capture the molecular heterogeneity of tumors, and identify differences between patients. This method is more personalized, can consider multiple molecular characteristics, and provide more accurate prognostic information.

[0019] Compared with the existing single gene prognosis assessment method, the present invention adopts a prediction model based on 8 gene combination markers. This model is more adaptable to the complex heterogeneous microenvironment within the tumor and fully considers the diversity of esophageal squamous cell carcinoma, including different subtypes and pathological characteristics, as well as the complexity of the tumor microenvironment. Therefore, the present invention has higher specificity and more accurate diagnostic efficacy.

[0020] Compared with existing multi-gene prognostic assessment methods, the diagnostic marker combination provided by the present invention combined with the prognostic risk assessment model significantly simplifies the process without reducing the accuracy of prediction. Through 8 genetic diagnostic markers that are related to T cell-mediated tumor killing and closely related to the patient's prognosis, the patient's risk score is calculated based on the expression levels of these genes. This means that only these 8 marker genes need to be tested, without the need for complex multi-gene evaluation, which significantly reduces the time cost and cost required for testing. Compared with traditional protein expression level detection methods, there is no need to use expensive antibodies, which greatly reduces the patient's testing costs, and does not require the judgment of pathology experts, reducing the interference of human judgment and operational errors.

[0021] Different from the prior art, the present invention obtains biomarkers related to the prognosis of patients with esophageal squamous cell carcinoma by screening T cell killing-related genes. Not only the predictive efficacy of biomarkers is investigated, but also the credibility of the prognosis prediction results of these biomarkers is improved through multiple verification methods, and the accuracy of the prediction results is also improved. At the same time, the present invention shows excellent accuracy in predicting the 1-, 3-, and 5-year survival rates of patients with esophageal squamous cell carcinoma. Therefore, doctors and patients can understand the treatment prospects more accurately and adopt more precise treatment strategies. This not only helps to improve the treatment effect, but also reduces the medical burden of patients, allowing patients to obtain key prognostic information more quickly, so as to better plan their treatment and rehabilitation paths.

[0022] The beneficial effects of the present invention are:

[0023] The present invention provides an application of a diagnostic marker combination in constructing a prognostic risk assessment model for esophageal squamous cell carcinoma and a detection kit. The prognosis of esophageal squamous cell carcinoma patients is predicted by the diagnostic marker combination and the corresponding prognostic risk assessment model for esophageal squamous cell carcinoma. The risk score 1, 3, and 5-year AUCs are 0.69, 0.67, and 0.70, respectively, proving that the prognostic model constructed based on 8 prognosis-related diagnostic markers has good distinguishing performance for the prognosis of esophageal squamous cell carcinoma patients and can accurately predict the prognosis of esophageal squamous cell carcinoma patients. It provides effective guidance for clinicians to make treatment decisions for esophageal squamous cell carcinoma patients, reduces the occurrence of ineffective treatment, thereby reducing the treatment cost and discomfort experience of patients, and helps to promote accurate prediction of the prognosis of esophageal squamous cell carcinoma patients and individualized comprehensive treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is the intersection of the 2322 differentially expressed genes in esophageal squamous cell carcinoma and the 1109 genes related to T cell killing in Example 1.

[0025] Figure 2 This is a volcano plot of differentially expressed genes related to T cell killing in patients with esophageal squamous cell carcinoma in Example 1.

[0026] Figure 3A This is a parameter diagram of the characteristic genes significantly associated with prognosis determined by LASSO regression analysis in Example 1.

[0027] Figure 3B This is the regression coefficient path diagram of the characteristic genes significantly associated with prognosis determined by LASSO regression analysis in Example 1.

[0028] Figure 4 This is the Kaplan-Meier survival analysis diagram of the patients in Example 1.

[0029] Figure 5This is a multi-factor COX regression analysis diagram of the prognosis prediction model in Example 1.

[0030] Figure 6 1 is the ROC analysis diagram of the 1-, 3- and 5-year survival rates of the patients in Example 1.

[0031] Figure 7 This is the Kaplan-Meier survival analysis diagram of the test set prognosis prediction model in Example 1.

[0032] Figure 8 1, 3 and 5-year survival rate ROC curves of the prognostic prediction model of the test set in Example 1.

[0033] Fig. 9 This is a nomogram for predicting the prognosis of patients with esophageal squamous cell carcinoma in Example 1. The blue dot indicates a representative patient.

[0034] Fig.10 It is a calibration curve between the 1-, 3- and 5-year overall survival predicted by the prognostic prediction model in Example 1 and the actual 1-, 3- and 5-year overall survival of the patient.

[0035] Fig.11 This is the Kaplan-Meier survival analysis of the prognosis validation set prediction model in Example 2.

[0036] Fig.12 1-year and 3-year survival rate ROC curves of the prognosis validation set prediction model in Example 2. DETAILED DESCRIPTION

[0037] The following embodiments of the present invention will be described in detail and in detail so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more specific definition of the protection scope of the present invention. Meanwhile, the following embodiments are only exemplary and explanatory, and do not limit the present disclosure.

[0038] The methods in the following examples are all conventional methods unless otherwise specified. The quantitative tests in the following examples were repeated three times and the results were averaged.

[0039] Example 1: Training set data analysis and prognosis prediction model establishment

[0040] 1. Obtain differentially expressed genes in human esophageal squamous cell carcinoma and genes related to T cell killing function

[0041] The clinical follow-up data and gene expression data of esophageal squamous cell carcinoma samples were collected to form a data set. The clinical follow-up data and gene expression data of esophageal squamous cell carcinoma samples were from a public database, including 179 esophageal squamous cell carcinoma patient cohort samples (GSE53625) (https: / / www.ncbi.nlm.nih.gov / geo / ). The clinical characteristics of the patients are shown in Table 1.

[0042] The samples without clinical follow-up information such as survival status and survival time were eliminated, and the standard was set as p<0.05 after correction and |log2 fold change|>1 (i.e. the change in expression was greater than 2 times). Finally, 2322 significantly differentially expressed genes in esophageal squamous cell carcinoma were obtained. Genes involved in T cell killing function were obtained from the public database (http: / / cis.hku.hk / TISIDB / ), and 1109 T cell killing related genes were obtained.

[0043] Table 1: Clinical characteristics of 179 patients in the GSE53625 dataset

[0044]

[0045] 2. Screening out prognosis-related genes based on differentially expressed genes and constructing a risk assessment model

[0046] In order to screen out genes related to the prognosis of esophageal squamous cell carcinoma from T cell killing-related genes, the intersection of 2322 significantly differentially expressed genes in esophageal squamous cell carcinoma and 1109 T cell killing-related genes was taken, as shown in Figure 2. Figure 1 As shown in the figure, 98 genes related to T cell killing in esophageal squamous cell carcinoma were obtained. Among these 98 genes, 56 genes were up-regulated, i.e., log2FC>1 on the horizontal axis was up-regulated by more than two times, and they were displayed as red dots; 42 genes were down-regulated, i.e., log2FC<-1 on the horizontal axis was down-regulated by more than two times, and they were displayed as blue dots. Figure 2 shown.

[0047] Univariate COX regression analysis was used to screen genes with independent prognostic value for the overall survival of patients in the training set. Univariate COX analysis is a survival analysis method used to evaluate the impact of a single factor on survival time or event rate without considering the impact of other factors. Finally, 10 genes that independently affect prognosis were obtained from 98 genes (p<0.05).

[0048] In order to further screen out the genes with the most prognostic value, the LASSOCOX regression model was used to screen again from the 10 genes. The results are as follows Figure 3A As shown, Figure 3AThis is a parameter diagram of the characteristic genes significantly related to prognosis determined by LASSO regression analysis. The ordinate is the Mean Squared Error, that is, the error size of the model. The dotted line on the left shows the λ model corresponding to the lowest error and fewer gene variables. Figure 3A It can be seen that through the parameter diagram analysis of LASSOCOX regression, when the gene variable is 8, the deviation value displayed on the ordinate is the smallest, and the model fitting effect is the best. The LASSO regression coefficient path diagram is further used for cross-validation, as shown in Figure 3B shown. Figure 3B This is a regression coefficient path diagram of the characteristic genes significantly related to prognosis determined by LASSO regression analysis. Each curve represents the trajectory of a gene changing with the regularization parameter; the vertical axis represents the value of the coefficient, the lower horizontal axis represents logλ, and the upper horizontal axis represents the number of non-zero coefficients in the model; as the λ value increases, the curve shows the time point when the coefficient is compressed to 0, thus reflecting the importance of the gene. Figure 3B It can be seen that the 10 curves of different colors represent the change trajectories of the 10 genes. As the parameter logλ increases, the regression coefficient (i.e., the ordinate value) continues to converge and finally converges to 0. As the λ value changes, the genes whose regression coefficients converge to 0 later are more important, so the key genes that affect the prognosis of patients with esophageal squamous cell carcinoma can be further screened. At this time, 8 key genes are selected to form the model marker combination: SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1.

[0049] Finally, multivariate COX regression analysis was used to obtain the coefficients of each of the eight genes, and the risk score formula was constructed using the coefficients and the expression levels of each gene (the bootstrap method was used for penalized maximum likelihood estimation, repeated 1000 times. The optimal weighting coefficient was determined by the regularization parameter λ using the 1-SE standard). The gene risk score model formula is: risk score = (0.0552283605 × KIF11 gene expression level) + (-0.0070408017 × SLC2A1 gene expression level) The risk score of the patient was calculated and the optimal cutoff point of the risk score was determined to be 0.61 in the survival analysis using the "Surv_Cutpoint" function. Based on this cutoff point, patients with a risk score greater than 0.61 were marked as a high-risk group, and patients with a risk score less than 0.61 were marked as a low-risk group.

[0050] 3. The risk assessment model predicts the prognosis of esophageal squamous cell carcinoma in the training set and the test set.

[0051] According to the above genetic risk scoring model formula, the patients in the training set were divided into high-risk group and low-risk group, and the Kaplan-Meier method was used for survival analysis. Figure 4 The upper part shows the survival curves for high-risk patients (blue) and low-risk patients (yellow). Figure 4 The lower part shows the number of high-risk and low-risk patients at different time points, which comprehensively illustrates that patients with esophageal squamous cell carcinoma in the high-risk group have a worse prognosis.

[0052] Subsequently, clinicopathological characteristics and risk scores were included in univariate and multivariate COX regression analyses to verify whether the risk score of the prognostic model could serve as an independent risk factor for predicting overall survival outcomes. Figure 5 As shown in Figure 2, the risk coefficient showed a significant difference, i.e., p<0.001 (HR=60.51, 95%, confidence interval=8.869-412.907), indicating that the prognostic model can be used as an independent predictor of overall survival. In addition, the ROC curve was used to evaluate the prognostic model, as shown in Figure 2. Figure 6 As shown, the areas under the ROC curve (AUC) of the patients at 1, 3, and 5 years were 0.69, 0.67, and 0.70, respectively, all greater than 0.5 and less than 1, indicating that the prediction model had good accuracy in different survival time periods in the training set.

[0053] The test set is a sample from the esophageal squamous cell carcinoma cohort in the TCGA public database. The clinical follow-up data and gene expression data of the esophageal squamous cell carcinoma samples include 89 esophageal squamous cell carcinoma patients. Based on the above-mentioned predictive prognosis model, the prognosis prediction of the esophageal squamous cell carcinoma case data in the test set was verified. First, the risk scores of the 89 patients were calculated, and the risk threshold was 0.61. The risk score was higher than the threshold and was in the high-risk group, and the risk score was lower than the threshold and was in the low-risk group. The survival rate statistics of the high-risk group and the low-risk group are as follows Figure 7 As shown, patients in the high-risk group had a worse prognosis compared with those in the low-risk group. Figure 8 The data showed that the 1-, 3-, and 5-year AUCs of the risk score were 0.69, 0.72, and 0.75, respectively, all greater than 0.5 and less than 1, indicating that the prediction model had good accuracy in different survival time periods in the validation set.

[0054] 6. Constructing a nomogram for overall prediction

[0055] The prognostic model risk score was further combined with clinical pathological characteristics including TNM stage, tumor grade, alcohol consumption, age, and gender to construct a prediction nomogram, such as Fig. 9 As shown, the blue dot shows a representative patient. It is used as a quantitative analysis tool to predict the survival risk of individual patients and to predict the prognosis of patients with multiple indicators. The integral scale is scored by these covariates, the total score of the sample is calculated, and the conversion function is used to estimate the survival probability of each sample at three time points (1 year, 3 years and 5 years). For example Fig. 9 The blue dot in shows a representative patient.

[0056] Female, 66 years old, no history of alcoholism, risk score 0.69, esophageal squamous cell carcinoma stage III, intermediate. Based on her current disease status, the patient's survival probability in the next 1, 3 and 5 years is predicted to be 57.6%, 17.9% and 10.9% respectively. Fig. 9 Draw a vertical line upwards from the age of 66 in the nomogram, and you can get the corresponding score (Points) of about 70 points. The gender is female, and the corresponding score is about 52 points. And so on, find the corresponding score for each variable state. Finally, add up the scores of all variables to get the patient's total score (Total Points) of about 395 points. Based on the total score, draw a vertical line downwards to know the patient's survival rate for the next 1 year, 3 years, and 5 years, that is, the 1-year (time>365) survival rate is 57.6%; the 3-year (time>1095) survival rate is 17.9%; the 5-year (time>1825) survival rate is 10.9%.

[0057] In addition, the actual survival probability is used to calibrate the predicted survival probability to improve the clinical applicability of the evaluation model. Fig.10 As shown in the figure, the X-axis represents the probability predicted by the nomogram, and the Y-axis represents the actual probability of patients with esophageal squamous cell carcinoma. The perfect prediction corresponds to the gray dotted line. The orange, blue, and red solid lines represent the 1-year, 3-year, and 5-year prediction performances, respectively, indicating that the patient survival rate predicted by this model is close to the actual survival rate.

[0058] Example 2: Clinical patient application

[0059] 1. Get data

[0060] In order to further verify the clinical patient case, tumor tissue samples of 64 patients with esophageal squamous cell carcinoma with clinical follow-up data were obtained. The mRNA of the tumor tissue samples was extracted and reverse transcribed into cDNA, and then hybridized with the nucleic acid probes of SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1 fixed on the gene chip, such as Seq ID No.1 to Seq ID No.16, and the nucleic acid probe of ACTIN shown in Seq ID No.17 to Seq ID No.18 was used as a control to detect the expression levels of SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1 in the patient samples. The specific information of the nucleic acid probes is shown in Table 2.

[0061] Table 2 8 groups of nucleic acid probes

[0062]

[0063]

[0064] 2. Calculate risk score

[0065] By the above method, the expression levels of 8 genes of 64 patients were obtained respectively, and then the gene expression level of each patient was substituted into the risk score model to obtain the risk score of each patient. The formula of the 8-gene risk score model is: risk score = (0.0552283605 × KIF11 gene expression level) + (-0.0070408017 × SLC2A1 gene expression level) + (0.0194419232 × KCNMA1 gene expression level) + (-0.0231625843 × BARHL2 gene expression level) + (-0.0214342225 × CA9 gene expression level) + (0.0161917634 × TIMP1 gene expression level) + (-0.0712586182 × MAGEC3 gene expression level) + (-0.0004431282 × PDZK1IP1 gene expression level). The gene expression levels and risk score data of the patients are shown in Table 3.

[0066] Table 3: Detection results of relative expression and risk score of patients with esophageal squamous cell carcinoma

[0067] Patient ID Risk Scoring Patient ID Risk Scoring Patient ID Risk Scoring Patient ID Risk Scoring Patient 1 0.6899 Patient 17 0.5535 Patient 33 0.7932 Patient 49 0.7609 Patient 2 0.3941 Patient 18 0.4690 Patient 34 0.6664 Patient 50 0.4340 Patient 3 0.5639 Patient 19 0.5004 Patient 35 0.7905 Patient 51 0.4567 Patient 4 0.7793 Patient 20 0.5412 Patient 36 0.6577 Patient 52 0.5679 Patient 5 0.6412 Patient 21 0.5641 Patient 37 0.5562 Patient 53 0.6111 Patient 6 0.5676 Patient 22 0.6199 Patient 38 0.6515 Patient 54 0.3524 Patient 7 0.6369 Patient 23 0.4422 Patient 39 0.6958 Patient 55 0.4026 Patient 8 0.6395 Patient 24 0.5228 Patient 40 0.5113 Patient 56 0.6184 Patient 9 0.6050 Patient 25 0.5635 Patient 41 0.6611 Patient 57 0.5978 Patient 10 0.4426 Patient 26 0.4496 Patient 42 0.6174 Patient 58 0.4109 Patient 11 0.5946 Patient 27 0.6589 Patient 43 0.6336 Patient 59 0.7028 Patient 12 0.6641 Patient 28 0.3873 Patient 44 0.7992 Patient 60 0.6302 Patient 13 0.5358 Patient 29 0.6509 Patient 45 0.8001 Patient 61 0.5265 Patient 14 0.5843 Patient 30 0.5151 Patient 46 0.5289 Patient 62 0.6360 Patient 15 0.4207 Patient 31 0.5105 Patient 47 0.5443 Patient 63 0.5342 Patient 16 0.5615 Patient 32 0.5038 Patient 48 0.7011 Patient 64 0.4907

[0068] 3. Prognostic analysis of patients with esophageal squamous cell carcinoma

[0069] The risk threshold is 0.61. The risk score is higher than the threshold and is in the high-risk group, while the risk score is lower than the threshold and is in the low-risk group. The survival rate statistics of the high-risk group and the low-risk group are as follows Fig.11 As shown. Fig.11 It can be seen that the survival rate of the high-risk group shown by the blue curve is significantly lower than that of the low-risk group shown by the yellow curve. The patients in the high-risk group are predicted to die, and the patients in the low-risk group are judged to survive. The predicted results are compared with the actual survival results of 1 year and 3 years to create an ROC curve, as shown in Figure 2. Fig.12 As shown, the AUC values ​​are 0.66 and 0.68, respectively, both greater than 0.5 and less than 1, indicating that the prediction model has good accuracy.

[0070] As can be seen from the above examples, the diagnostic marker combination provided by the present invention is combined with LASSO-Cox regression analysis to construct a prognosis prediction model for esophageal squamous cell carcinoma in the GSE53625 training set. Its accuracy was verified by the TCGA external test set, and the prognosis prediction ability of the model and the correlation between the model and other clinical characteristics were analyzed. At the same time, the model was verified by clinical patient examples, and the results showed that the prediction model has good distinguishing performance for the prognosis of patients with esophageal squamous cell carcinoma and can accurately predict the prognosis of patients. Therefore, the model can be used as a prognosis guide for patients with esophageal squamous cell carcinoma and promotes personalized treatment of patients with esophageal squamous cell carcinoma.

[0071] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. Application of diagnostic marker combination in constructing a prognostic risk assessment model for esophageal squamous cell carcinoma, characterized in that: The diagnostic marker combination is SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1.

2. The use according to claim 1, characterized in that: The prediction model includes nucleic acid probes for detecting the expression of the gene combination SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1; Among them, the sequence of the forward probe SLC2A1-F for detecting SLC2A1 is shown in Seq ID No. 1, and the sequence of the reverse probe SLC2A1-R for detecting SLC2A1 is shown in Seq ID No. 2; The sequence of the forward probe KIF11-F for detecting KIF11 is shown in Seq ID No. 3, and the sequence of the reverse probe KIF11-R for detecting KIF11 is shown in Seq ID No. 4; The sequence of the forward probe KCNMA1-F for detecting KCNMA1 is shown in Seq ID No. 5, and the sequence of the reverse probe KCNMA1-R for detecting KCNMA1 is shown in Seq ID No. 6; The sequence of the forward probe BARHL2-F for detecting BARHL2 is shown in Seq ID No. 7, and the sequence of the reverse probe BARHL2-R for detecting BARHL2 is shown in Seq ID No. 8; The sequence of the forward probe CA9-F for detecting CA9 is shown in Seq ID No. 9, and the sequence of the reverse probe CA9-R for detecting CA9 is shown in Seq ID No. 10; The sequence of the forward probe TIMP1-F for detecting TIMP1 is shown in Seq ID No. 11, and the sequence of the reverse probe TIMP1-R for detecting TIMP1 is shown in Seq ID No. 12; The sequence of the forward probe MAGEC3-F for detecting MAGEC3 is shown in Seq ID No. 13, and the sequence of the reverse probe MAGEC3-R for detecting MAGEC3 is shown in Seq ID No. 14; The sequence of the forward probe PDZK1IP1-F for detecting PDZK1IP1 is shown in Seq ID No.15, and the sequence of the reverse probe PDZK1IP1-R for detecting PDZK1IP1 is shown in Seq ID No.

16.

3. The use according to claim 1, characterized in that The model includes the following content: risk score = (0.0552283605 × KIF11 expression) + (-0.0070408017 × SLC2A1 expression) + (0.0194419232 × KCNMA1 expression) + (-0.0231625843 × BARHL2 expression) + (-0.0214342225 × CA9 expression) + (0.0161917634 × TIMP1 expression) + (-0.0712586182 × MAGEC3 expression) + (-0.0004431282 × PDZK1IP1 expression). When the risk score>0.61, it is judged as a high-risk group, and the risk score<0.61 is judged as a low-risk group.

4. The use according to claim 1, characterized in that The model also includes the following: patient risk score, clinical pathological characteristics including TNM stage, tumor grade, alcohol consumption, age, gender, and risk score are integrated into an integral scale to further calculate the total score of the sample, and use a conversion function to estimate the survival probability of each patient at 1 year, 3 years, and 5 years.

5. A detection kit, characterized in that: Contains nucleic acid probes and instructions for detecting the expression of the gene combination SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1; Among them, the sequence of the forward probe SLC2A1-F for detecting SLC2A1 is shown in Seq ID No. 1, and the sequence of the reverse probe SLC2A1-R for detecting SLC2A1 is shown in Seq ID No. 2; The sequence of the forward probe KIF11-F for detecting KIF11 is shown in Seq ID No. 3, and the sequence of the reverse probe KIF11-R for detecting KIF11 is shown in Seq ID No. 4; The sequence of the forward probe KCNMA1-F for detecting KCNMA1 is shown in Seq ID No. 5, and the sequence of the reverse probe KCNMA1-R for detecting KCNMA1 is shown in Seq ID No. 6; The sequence of the forward probe BARHL2-F for detecting BARHL2 is shown in Seq ID No. 7, and the sequence of the reverse probe BARHL2-R for detecting BARHL2 is shown in Seq ID No. 8; The sequence of the forward probe CA9-F for detecting CA9 is shown in Seq ID No. 9, and the sequence of the reverse probe CA9-R for detecting CA9 is shown in Seq ID No. 10; The sequence of the forward probe TIMP1-F for detecting TIMP1 is shown in Seq ID No. 11, and the sequence of the reverse probe TIMP1-R for detecting TIMP1 is shown in Seq ID No. 12; The sequence of the forward probe MAGEC3-F for detecting MAGEC3 is shown in Seq ID No. 13, and the sequence of the reverse probe MAGEC3-R for detecting MAGEC3 is shown in Seq ID No. 14; The sequence of the forward probe PDZK1IP1-F for detecting PDZK1IP1 is shown in Seq ID No. 15, and the sequence of the reverse probe PDZK1IP1-R for detecting PDZK1IP1 is shown in Seq ID No. 16; The instruction manual records the following content: Risk score = (0.0552283605×KIF11 expression level) + (-0.0070408017×SLC2A1 expression level) + (0.0194419232×KCNMA1 expression level) + (-0.0231625843×BARHL2 expression level) + (-0.0214342225×CA9 expression level) + (0.0161917634×TIMP1 expression level) + (-0.0712586182×MAGEC3 expression level) + (-0.0004431282×PDZK1IP1 expression level). When the risk score>0.61, it is judged to be a high-risk group, and when the risk score<0.61, it is judged to be a low-risk group.

6. Use of the detection kit as claimed in claim 5 in the preparation of a product for prognostic risk assessment of patients with esophageal squamous cell carcinoma.

7. The use according to claim 6, characterized in that The prognostic risk assessment of patients with esophageal squamous cell carcinoma refers to assessing the overall survival rate of patients with esophageal squamous cell carcinoma.

Images