Biomarker, model and system for prognosis evaluation of acute myelogenous leukemia and application of biomarker, model and system

Abstract

The invention provides a biomarker, a model and a system for prognosis evaluation of acute myelogenous leukemia and application of the biomarker, the model and the system. According to the invention, a prognosis model for evaluating the risk of acute myelogenous leukemia is constructed by using a biomarker consisting of GZMB, OPTN, DNM1, MPO and GZMH genes. The method comprises the following steps: extracting marrow fluid nucleated cell RNA of a patient with acute myelogenous leukemia, detecting the relative expression quantity of the five genes in a quantitative detection manner, substituting into a constructed risk prediction model, and performing risk prognosis evaluation according to scores. The method is low in detection cost, low in technical threshold and better in prediction performance, and has independent prognosis evaluation value.

Description

Technical Field

[0001] The present invention belongs to the field of gene technology and biomedical technology, and in particular relates to a biomarker, model, system and application thereof for prognosis assessment of acute myeloid leukemia. Background Art

[0002] Acute myeloid leukemia (AML) is a highly invasive and heterogeneous hematological malignancy caused by the clonal malignant proliferation of bone marrow progenitor cells. Due to its rapid progression and complex molecular pathological mechanisms, the treatment and management of AML have always faced great challenges. Most AML patients lack obvious symptoms in the early stage, usually with acute complications such as bone marrow failure as the main manifestation, and the prognosis is poor. At present, chemotherapy and stem cell transplantation remain the main treatments for AML. Although the treatment of AML has made significant progress in recent years with the development of targeted therapy and immunotherapy (including kinase inhibitors, monoclonal antibodies, chimeric antigen receptor T cells (CAR-T)), the high heterogeneity of tumors and the complex molecular pathological background still lead to relapse, drug resistance and treatment-related toxicity, which seriously affect the long-term survival of patients. Therefore, improving the prognostic assessment and risk stratification of AML patients is crucial to optimize treatment decisions.

[0003] Traditional AML prognosis assessment mostly relies on factors such as the patient's immune phenotype, cytogenetics, and molecular characteristics. Among them, abnormalities in cytogenetics and molecular markers are currently the most important means of prognostic assessment. With the popularization of high-throughput sequencing (such as next-generation sequencing, NGS), the molecular characteristics of AML have been studied more deeply, and common gene mutations include FLT3, NPM1, DNMT3A, TET2, IDH1 / 2, CEBPA, RUNX1, and TP53. Some molecular abnormalities have shown clear prognostic value, such as CEBPA biallelic mutations, NPM1 mutations, and FLT3-ITD negative or low levels indicate a good prognosis, while FLT3-ITD positive, TP53 mutations, etc. indicate a poor prognosis. However, these prognostic models usually rely on complex molecular detection technologies, such as NGS, but the widespread application of these technologies is limited by cost and technical popularity, especially in areas with limited economic conditions, and it is difficult to become a routine detection method.

[0004] In clinical practice, some AML patients still have a poor prognosis even though no significant mutations or abnormalities are found in conventional molecular genetic testing. This suggests that in addition to known molecular genetic changes, there may be other unidentified prognostic factors. Most of the current molecular prognostic models are based on clear gene mutations or chromosomal aberrations, and there is a lack of effective risk stratification methods for patients who do not show conventional molecular genetic changes. This model focuses on the relationship between the expression pattern of 5 PF4-related genes and prognosis in patients without obvious molecular genetic changes. Therefore, this model can not only provide prognostic assessment for patients with known gene mutations, but also provide more instructive prognostic predictions for patients without conventional molecular genetic changes. Through this analysis of gene expression levels, the risk stratification of AML patients can be further refined to make up for the limitations of traditional molecular genetic testing. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a biomarker, model, system and application thereof for the prognosis assessment of acute myeloid leukemia. The predictive model constructed by the biomarker can accurately predict the prognosis of patients and guide clinical treatment.

[0006] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0007] The present invention provides a biomarker for risk prognosis assessment of acute myeloid leukemia, wherein the biomarker consists of GZMB, OPTN, DNM1, MPO and GZMH genes.

[0008] The present invention also provides the use of the biomarker in preparing a product for assessing the risk and prognosis of acute myeloid leukemia.

[0009] The present invention also provides an acute myeloid leukemia prognosis assessment model, which uses the expression level of the biomarker as an input variable and calculates the risk score according to the following formula: risk score = e^((-0.08062789)*log2(MPO relative expression)+(0.05275965)*log2(OPT N relative expression)+(0.14933799)*log2(GZMB relative expression)+(0.04348332)*log2(GZMH relative expression)+(-0.13506054)*log2(DNM1 relative expression)+0.6812835).

[0010] The present invention also provides a system for prognosis risk assessment of acute myeloid leukemia, the system comprising a calculation unit, the calculation unit calculating the risk score using the acute myeloid leukemia prognosis assessment model; the system further comprising a detection unit, an information acquisition unit and an evaluation unit; the detection unit is used to detect the expression level of the biomarker; the information acquisition unit is used to execute the operation of acquiring the test information of the subject, the test information including the expression level of the biomarker; the evaluation unit is used to judge the risk probability of the prostate cancer prognosis of the subject according to the calculation result of the calculation unit, and give rational prevention and treatment suggestions.

[0011] Preferably, the expression level of the biomarker is determined by fluorescent quantitative PCR.

[0012] Preferably, the primers involved in the fluorescence quantitative PCR assay include forward and reverse primers of the GZMB gene, forward and reverse primers of the OPTN gene, forward and reverse primers of the DNM1 gene, forward and reverse primers of the MPO gene, forward and reverse primers of the GZMH gene, and forward and reverse primers of the GAPDH gene.

[0013] Preferably, the fluorescent quantitative PCR assay involves a PCR amplification system comprising: 10 μl of 2×QuantiNova SYBR Green PCR masterMix, 0.7 pmol of forward primer, 0.7 pmol of reverse primer, 10 ng of template cDNA, and 2 μl of RNAase-Free water.

[0014] Preferably, the fluorescent quantitative PCR assay involves a PCR amplification procedure of: 95°C×2min; 95°C×5s, 60°C×10s, 40 cycles; entering a 4°C state after 40 cycles.

[0015] The present invention also provides the use of the acute myeloid leukemia prognosis assessment model or the acute myeloid leukemia prognosis assessment system in the preparation of an acute myeloid leukemia risk prognosis assessment product.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] The risk prognosis assessment model constructed by the present invention has three advantages: (1) Low detection cost and low technical threshold: only five genes, GZMB, OPTN, DNM1, MPO and GZMH, need to be quantitatively detected, which significantly reduces the detection cost. For primary hospitals that do not have a cytogenetic detection system, the model can be used for risk assessment with only quantitative PCR technology, which has wide clinical applicability and promotion value. (2) Better prediction performance: The risk score calculated by the prediction model was compared with the CALGB cytogenetic risk level using decision curve analysis (DCA). The results showed that the use of this model for prognostic assessment has better prediction performance than the use of cytogenetic risk level. (3) It has independent prognostic assessment value: Based on the TCGA-AML data set, through multivariate Cox regression analysis, it was found that the patient risk score is an independent risk factor independent of the patient's cytogenetic risk level, age, and whether chemotherapy is performed. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Analysis of PF4 expression levels and survival status in different data sets.

[0019] Figure 2 The figure shows the Venn diagram of differentially expressed genes in the GSE13159 and GSE9476 datasets.

[0020] Figure 3 This is the Venn diagram of PF4-related genes and AML-related genes.

[0021] Figure 4 Forest plot of univariate Cox regression analysis of 6 prognosis-related genes.

[0022] Figure 5 LASSO regression screening graph for 6 prognosis-related genes. The horizontal axis of the left graph is log(Lambda), and the vertical axis represents the cross-validation error; the horizontal axis deviance of the right graph represents the proportion of residuals explained by the model, showing the relationship between the number of genes and the proportion of residuals explained, and the vertical axis is the coefficient of the gene.

[0023] Figure 6 is the ROC curve (TCGA-AML dataset).

[0024] Figure 7 is the KM curve of the risk model (TACG-AML dataset).

[0025] Figure 8 is the risk curve of the risk model (TACG-AML dataset).

[0026] Fig. 9 is the ROC curve (TARGET-AML dataset).

[0027] Fig.10 is the KM curve of the risk model (TARGET-AML dataset).

[0028] Fig.11 is the risk curve of the risk model (TARGET-AML dataset).

[0029] Fig.12 This is the result of the decision curve analysis to evaluate the prognostic ability of the risk score.

[0030] Fig.13 The Schoenfeld residual method was used to determine the PH assumption results of Cox proportional hazards regression analysis.

[0031] Fig.14 Forest plot for multivariate Cox analysis.

[0032] Fig.15 Nomogram for independent prognostic factors.

[0033] Fig.16 This is the correction curve of the nomogram.

[0034] Fig.17 is the ROC curve (GSE71014 dataset).

[0035] Fig.18 This is a univariate Cox analysis graph (GSE71014 dataset).

[0036] Fig.19 Forest plot for multivariate Cox analysis (GSE71014 dataset).

[0037] Fig. 20 This is a nomogram of independent prognostic factors (GSE71014 dataset).

[0038] Fig.21 This is the nomogram correction curve (GSE71014 dataset).

[0039] Fig. 22 This is a graph showing the prognostic ability assessment of risk scores using the decision curve analysis method.

[0040] Fig.23 This is the clinical ROC curve of the risk prediction model.

[0041] Fig.24 Clinical survival analysis for risk prediction models. DETAILED DESCRIPTION

[0042] The present invention provides a biomarker for the prognosis assessment of acute myeloid leukemia, wherein the biomarker is composed of GZMB, OPTN, DNM1, MPO and GZMH genes. The present invention uses univariate Cox regression analysis and LASSO regression to screen all PF4-related differentially expressed genes in TCGA-AML and TARGET-AML data sets, and obtains differentially expressed genes GZMB, OPTN, DNM1, MPO and GZMH genes for subsequent modeling.

[0043] The present invention also provides the use of the biomarker in preparing a product for evaluating the prognosis of acute myeloid leukemia. The product of the present invention includes a data model / system, a tissue chip, a kit, a reagent, and a nuclear membrane strip.

[0044] The present invention also provides an acute myeloid leukemia prognosis assessment model, which uses the expression level of the biomarker as an input variable and calculates the risk score according to the following formula: risk score = e^((-0.08062789)*log2(MPO relative expression)+(0.05275965)*log2(OPT N relative expression)+(0.14933799)*log2(GZMB relative expression)+(0.04348332)*log2(GZMH relative expression)+(-0.13506054)*log2(DNM1 relative expression)+0.6812835).

[0045] The present invention also provides a system for evaluating the prognosis of acute myeloid leukemia, the system comprising a calculation unit, the calculation unit using the expression level of the biomarker as an input variable, and calculating the risk score according to the following formula: Risk score = e^((-0.08062789)*log2(relative expression of MPO)+(0.05275965)*log g2(relative expression of OPTN)+(0.14933799)*log2(relative expression of GZMB)+(0.04348332)*log2(relative expression of GZMH)+(-0.13506054)*log2(relative expression of DNM1)+0.6812835); the system also includes a detection unit, an information acquisition unit and an evaluation unit; the detection unit is used to detect the expression level of the biomarker; the information acquisition unit is used to execute the operation of acquiring the subject's detection information, and the detection information includes the expression level of the biomarker; the evaluation unit is used to judge the risk probability of the subject's prostate cancer prognosis according to the calculation result of the calculation unit, and give reasonable prevention and treatment suggestions.

[0046] In the present invention, the expression level of the biomarker is determined by fluorescent quantitative PCR. The primers involved in the fluorescent quantitative PCR determination of the present invention include forward and reverse primers of the GZMB gene, forward and reverse primers of the OPTN gene, forward and reverse primers of the DNM1 gene, forward and reverse primers of the MPO gene, forward and reverse primers of the GZMH gene and forward and reverse primers of the GAPDH gene, and the nucleotide sequences of the above primers are shown in SEQ ID No.1-12. The GAPDH gene of the present invention is an internal reference gene. The primer sequences used in the PCR amplification system of the present invention are not limited to the above sequences, and the corresponding primers or probe sequences of the fluorescent quantitative PCR detection system suitable for determining the genes described in this application can also be selected.

[0047] In the present invention, the fluorescent quantitative PCR assay involves a PCR amplification system of: 2×QuantiNova SYBRGreen PCR masterMix 10μl, forward primer 0.7pmol, reverse primer 0.7pmol, template cDNA 10ng, RNAase-Free water 2μl. The reagents used in the PCR amplification system of the present invention are not limited to the above reagents, and reagents suitable for fluorescent quantitative PCR detection of the genes described in the present application can also be selected.

[0048] Preferably, the fluorescent quantitative PCR assay involves a PCR amplification procedure of: 95°C for 2 min (this stage is a hot start step and does not participate in the cycle); 95°C for 5 s, 60°C for 10 s, 40 cycles (there is a melting curve analysis process after each cycle, and the fastest temperature rise mode is selected); after 40 cycles, enter a low temperature state of 4°C. The PCR amplification procedure of the present invention is not limited to the above-mentioned procedure, and an amplification procedure suitable for fluorescent quantitative PCR detection of the gene described in the present application can also be selected.

[0049] The present invention also provides the use of the acute myeloid leukemia prognosis assessment model or the acute myeloid leukemia prognosis assessment system in the preparation of an acute myeloid leukemia risk prognosis assessment product.

[0050] The technical solutions provided by the present invention are described in detail below in conjunction with the embodiments, but they should not be construed as limiting the protection scope of the present invention.

[0051] Example 1

[0052] Based on the median expression level of platelet factor 4-related gene (PF4), AML patients in TCGA-AML and TARGET-AML datasets were divided into PF4 high expression group and PF4 low expression group, respectively. Kaplan-Meier survival analysis was performed on the two groups of patients to evaluate the relationship between PF4 expression and overall survival rate of AML patients. The results are as follows Figure 1As shown, it was found that whether in the TCGA dataset or the TARGET dataset, the expression of PF4 showed the same trend as the patient's survival prognosis, that is, high PF4 expression often means a worse survival prognosis.

[0053] To understand the functional and pathway differences between the high and low expression groups of PF4, the transcriptomes of patients with high and low PF4 expression in the TCGA-AML dataset were analyzed using the stat_compare_means function of the R language ggpubr package to perform differential analysis of standardized gene expression using the Wilcoxo test. With P value less than 0.05 and log2(FC)>1 as the threshold, among all 15,230 highly expressed genes, 672 genes showed significant upregulation in the PF4 high expression group, and 156 genes showed significant downregulation in the PF4 high expression group.

[0054] In addition, GSE13159 and GSE9476 datasets were selected, and transcriptome differential analysis was performed on AML patients and their corresponding normal tissue samples. For the GSE13159 dataset, 214 of the 20,188 genes were significantly upregulated in AML patients, and 645 were significantly downregulated. For the GSE9476 dataset, 207 of the 12,403 genes were significantly upregulated in AML patients, and 540 were significantly downregulated. For the genes differentially expressed in the GSE13159 dataset and the GSE9476 dataset, the intersection was taken, and 359 genes showed differential expression in both the GSE13159 and GSE9476 gene sets ( Figure 2 ).

[0055] For these 359 genes differentially expressed in AML, we intersected them with the PF4-related genes differentially expressed in the PF4 high-expression and PF4 low-expression patient subgroups obtained previously based on the TCGA dataset. Figure 3 The results showed that 58 PF4-related genes were obtained, and these genes were differentially expressed in AML patient tissues compared with normal control tissues.

[0056] A prognostic prediction model was constructed based on PF4-related differentially expressed genes: Univariate Cox regression analysis was performed on the 58 PF4-related genes obtained from the above analysis. According to the survival prognosis information of TCGA-AML patients, 6 genes among the 58 genes showed a significant correlation with the survival prognosis of AML patients (P<0.02) ( Figure 4 ).

[0057] Further LASSO regression screening was performed on these 6 genes, such as Figure 5As shown in the figure, the left dotted line position on the left is the position with the minimum cross-validation error. The corresponding horizontal axis log(Lambda) is determined according to this position (lambda.min). The upper horizontal axis shows the number of model genes. Find the optimal log(Lambda) value, find the corresponding gene and its coefficient in the right figure, and the proportion of the residual explained by the model. Select the Lasso model at the Lambda.min position when the error is the smallest as the Lasso analysis result. Five genes were obtained through Lasso regression analysis. These five genes have the best model construction effect, namely: GZMB, OPTN, DNM1, MPO, GZMH. These five genes are used as model genes.

[0058] Based on the relevant data of 151 AML patients in the TCGA-AML dataset, a multivariate Cox regression model was constructed for the genes finally retained, and the patient risk score was calculated according to the model. Prediction model risk score (risk Score) = e^((-0.08062789)*log2(MPO relative expression)+(0.05275965)*log2(OPTN relative expression)+(0.14933799)*log2(GZMB relative expression)+(0.04348332)*log2(GZMH relative expression)+(-0.13506054)*log2(DNM1 relative expression)+0.6812835).

[0059] Example 2

[0060] Using the TCGA-AML dataset, the survival ROC package of R language was used to draw the receiver operating characteristic curve (ROC). According to the obtained prognostic risk model, the risk score of each patient was calculated. The area under the ROC curve (AUC) was used to evaluate the effectiveness of the model with 1, 3, and 5 years as the survival time nodes. The results are shown in the figure. The AUC values ​​of the ROC curve are 0.748, 0.713, and 0.717 at 1, 3, and 5 years, respectively. Figure 6 ), all greater than 0.6. Therefore, these five model genes can effectively predict the survival status of AML patients, and the prognostic model has good performance.

[0061] Based on the above data set and the constructed evaluation model, the risk score of each patient was calculated, and the median value was calculated by the function in the R code. The samples were divided into high-risk group and low-risk group with the median value of the risk score as the boundary. The KM curve of the survival of high-risk and low-risk groups was drawn using survminer in R language, and the survival difference between high-risk and low-risk groups was compared by log-rank test. Figure 7As shown, there was a significant difference in the survival of AML patients between the high-risk and low-risk groups (P < 0.01), and the survival status of patients in the high-risk group was worse than that of patients in the low-risk group.

[0062] After sorting the risk scores from low to high, draw the risk curve. Figure 8 As shown, there are more death samples in the high-risk group than in the low-risk group. Therefore, it can be inferred that the mortality rate of AML patients in the low-risk group is lower than that in the high-risk group, and the survival status of the high-risk group samples is worse.

[0063] Example 3

[0064] Using the TARGET-AML dataset, the risk score was calculated based on the prognostic risk model, and ROC and survival curves were drawn. Fig. 9 As shown in the figure, the AUC values ​​of the ROC curve of the TARGET-AML dataset were greater than 0.6 at 1, 3, and 5 years.

[0065] The samples were divided into high-risk group and low-risk group based on the risk score cutoff value, and the survival difference between the high-risk group and the low-risk group was compared by long-rank test. Fig.10 As shown in the figure, the survival rate of patients in the high-risk group was lower than that in the low-risk group, and P < 0.05, which was significantly different.

[0066] Based on the TARGET-AML dataset, the risk scores are sorted from low to high, and the risk curve is drawn ( Fig.11 ). In the figure, there are more death samples in the high-risk group than in the low-risk group, and the survival time of the high-risk group samples is lower than that of the low-risk group.

[0067] Example 4

[0068] (1) Evaluation of Prediction Performance

[0069] The risk score calculated by this prediction model was compared with the CALGB cytogenetic risk level of AML patients in the TCGA database using decision curve analysis (DCA). The results suggest that the use of this model for prognostic assessment has a better predictive performance than using cytogenetic risk level ( Fig.12 ).

[0070] (2) Independent prognostic assessment

[0071] Based on the TCGA-AML dataset, multivariate Cox regression analysis found that the patient risk score is an independent risk factor independent of the patient's cytogenetic risk level, age, and whether chemotherapy is performed. The specific evaluation process is as follows:

[0072] Before conducting a multivariate Cox regression analysis of cytogenetic risk level, age, neoadjuvant therapy, and patient risk score, the Schoenfeld residual method was used to determine the PH assumption of the Cox proportional hazard regression analysis. Fig.13 As shown, it can be seen that the effects of the four prognostic factors on survival rate do not change with time, and the P values ​​of the four factors are all greater than 0.05, indicating that each variable meets the PH test, and the overall test P value of the model is 0.119, and the model as a whole meets the PH test.

[0073] Then, the above factors were subjected to multivariate Cox prognostic analysis. Fig.14 As shown in the results, it was found that the patient's cytogenetic risk level, age, chemotherapy and risk score were still significantly correlated with the patient's prognosis (P<0.05), suggesting that the above factors can be used as independent prognostic factors.

[0074] Using the R language rms package, we scored according to the genetic risk level, age, chemotherapy, and risk score. Each factor corresponds to a score, and the sum of the scores of each factor corresponds to the total score (TotalPoint). Then, we predicted the 1-, 3-, and 5-year survival rates based on the total score. The higher the score, the lower the survival rate. Finally, we plotted the prediction results into a nomogram ( Fig.15 ). Based on the nomogram prediction model, draw the calibration curve corresponding to the nomogram ( Fig.16 ), the horizontal axis of the calibration curve is the predicted probability, and the vertical axis is the actual probability. The closer the slope of the calibration curve is to 1, the more accurate the model prediction is. As can be seen from the figure, the slope of the curve is close to 1 in the prediction of 1, 3, and 5-year survival rates, indicating that the model has good predictive ability.

[0075] Example 5

[0076] In order to further evaluate the prognostic risk model, the external validation set GSE71014 dataset was used to validate the prognostic risk model. The risk score was calculated based on the prognostic risk model, and ROC and survival curves were drawn at the same time. Fig.17 As shown in the figure, the AUC values ​​of the ROC curve of this data set were greater than 0.6 at the 3rd and 5th years.

[0077] The four clinical characteristics of the patients in the data set, including cytogenetic risk level, age, gender, and whether neoadjuvant therapy was performed, were included in the study together with the patient risk score calculated previously, and a univariate Cox analysis was performed. Fig.18 As shown, the patient's cytogenetic risk grade, age, whether or not neoadjuvant treatment was performed, and risk score were significantly correlated with the patient's prognosis.

[0078] Then, a multivariate Cox prognostic analysis was performed on the above factors. The results are as follows: Fig.19 As shown, the patient's cytogenetic risk grade, age, whether neoadjuvant treatment was performed, and risk score were significantly correlated with the patient's prognosis, suggesting that several factors can serve as independent prognostic factors.

[0079] Using the R language rms package, we scored according to the genetic risk level, age, whether neoadjuvant therapy was performed, and risk score. Each factor corresponds to a score, and the sum of the scores of each factor corresponds to the total score (TotalPoint). Then, we predicted the 1-, 3-, and 5-year survival rates based on the total score. The higher the score, the lower the survival rate. Finally, we plotted the prediction results into a nomogram ( Fig. 20 ). Based on the nomogram prediction model, draw the calibration curve corresponding to the nomogram ( Fig.21 ), the slope modulus of the calibration curve was close to 1, indicating that the calibration curve had good predictive ability for 1-, 3-, and 5-year survival rates. To further verify the prognostic significance of the prediction model, the risk score was compared with the CALGB cytogenetic risk level of AML patients, and the prognostic ability of the risk score was evaluated using decision curve analysis (DCA). Fig. 22 As shown, the risk score curve is farther away from the None and All lines than the cytogenetic risk level curve, indicating that using the risk score to predict patient prognosis has a better predictive performance than using the cytogenetic level.

[0080] Example 6

[0081] Using AML specimens from the Department of Hematology of the Affiliated Hospital of Jining Medical College (approved by the Ethics Committee), we extracted nucleated cell RNA from bone marrow fluid, detected the relative expression of the above five genes by real-time fluorescence quantitative PCR, and substituted them into the constructed risk prediction model for risk prediction. According to the latest version of the NCCNAML guidelines, the cytogenetic risk level of clinical patients was determined, and the cytogenetic risk level was divided into low-risk and high-risk patients, and divided into two groups according to their true labels (high risk vs low risk); the corresponding risk prediction scores were calculated based on the risk prediction model, and ROC curve analysis was performed ( Fig.23 The area under the ROC curve was 0.8750 (P=0.0004, statistically significant).

[0082] Real-time fluorescence quantitative PCR detection method: First, extract the bone marrow AML cells of the newly diagnosed AML patients (approved by the ethics committee) and use red blood cell lysis buffer (BD FACS TMLysing Solution, Catalog No. 349202) lysed red blood cells, and RNA was extracted from the remaining cells (Takara, Trizol, Catalog No. T9108). The RNA was extracted using the TransGene kit ( First-Strand cDNA Synthesis SuperMix, Catalog No. AE301-03) was used to reverse transcribe the sample RNA into cDNA. The QIAGEN kit (QuantiNovaSYBR Green PCR Kit, Catalog No. 208054) was used for quantitative analysis (qPCR) of each target gene. The qPCR reaction conditions are as follows: 95℃×2min (this step is a hot start step and does not participate in the cycle)→95℃×5s→60℃×10s (40 cycles), there is a fluorescence level detection process after each cycle, and there is a melting curve analysis process after the amplification is completed. The fastest mode is selected for the heating mode, and the 4℃ low temperature state is entered after 40 cycles. The primer sequences of the genes involved are shown in Table 1.

[0083] Table 1 Primer sequences of each gene

[0084]

[0085] According to the risk score of each patient, the data were divided into high, medium and low risk groups using the quantile function and cut function of R language with 0.33 and 0.66 as the classification cutoff values, respectively. Risk scores below the 33% percentile were low risk, between the 33% and 66% percentiles were medium risk, and above the 66% percentiles were high risk. Survival analysis was performed on patients with low, medium and high risk risk scores. The results showed that the survival time of patients in the high-risk group was significantly shorter than that of patients in the low-risk group, and P = 0.0308, which was significantly different ( Fig.24 ).

[0086] 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 principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A biomarker for prognosis assessment of acute myeloid leukemia, characterized in that: The biomarkers consist of GZMB, OPTN, DNM1, MPO and GZMH genes.

2. Use of the biomarker as claimed in claim 1 in the preparation of a product for risk and prognosis assessment of acute myeloid leukemia.

3. A prognosis assessment model for acute myeloid leukemia, characterized in that: The acute myeloid leukemia prognosis assessment model uses the expression level of the biomarker described in claim 1 as an input variable, and calculates the risk score according to the following formula: Risk score = e^((-0.08062789)*log2(MPO relative expression)+(0.05275965)*log2(OPTN relative expression)+(0.14933799)*log2(GZMB relative expression)+(0.04348332)*log2(GZMH relative expression)+(-0.13506054)*log2(DNM1 relative expression)+0.6812835).

4. A system for evaluating the prognosis of acute myeloid leukemia, characterized in that: The system comprises a calculation unit, which calculates a risk score using the acute myeloid leukemia prognosis assessment model of claim 3; The system also includes a detection unit, an information acquisition unit and an evaluation unit; the detection unit is used to detect the expression level of the biomarker according to claim 1; the information acquisition unit is used to execute the operation of acquiring the test information of the subject, and the test information includes the expression level of the biomarker; the evaluation unit is used to judge the risk probability of the prognosis of acute myeloid leukemia in the subject according to the calculation result of the calculation unit, and give prevention and treatment suggestions.

5. The system according to claim 4, characterized in that The expression level of the biomarker is determined by fluorescent quantitative PCR.

6. The system according to claim 5, characterized in that The primers involved in the fluorescence quantitative PCR assay include forward and reverse primers of the GZMB gene, forward and reverse primers of the OPTN gene, forward and reverse primers of the DNM1 gene, forward and reverse primers of the MPO gene, forward and reverse primers of the GZMH gene and forward and reverse primers of the GAPDH gene.

7. The system according to claim 5, characterized in that The fluorescent quantitative PCR assay involves a PCR amplification system comprising: 10 μl of 2×QuantiNova SYBR Green PCR master Mix, 0.7 pmol of forward primer, 0.7 pmol of reverse primer, 10 ng of template cDNA, and 2 μl of RNAase-Free water.

8. The system according to claim 5, characterized in that The fluorescent quantitative PCR assay involves a PCR amplification program of: 95°C×2min; 95°C×5s, 60°C×10s, 40 cycles; after 40 cycles, entering a 4°C state.

9. Use of the acute myeloid leukemia prognosis assessment model as described in claim 3 or the acute myeloid leukemia prognosis assessment system as described in any one of claims 4 to 8 in the preparation of an acute myeloid leukemia risk prognosis assessment product.

Images