System for evaluating the clinical prognosis of acute myeloid leukemia by integrating multiple fatty acid metabolism genes

Abstract

The present invention belongs to the field of biomedicine, and specifically relates to a prognostic model for predicting the overall survival rate of AML patients based on 9 fatty acid metabolism-related genes. The present invention analyzed the gene expression profiles of 354 AML patients from TCGA and VIZOME, determined the prognosis-related genes based on the fatty acid metabolism-related genes of the TCGA dataset using univariate Cox regression analysis and survival analysis, and established a prognostic model with 9 fatty acid-related genes. This model can evaluate the prognosis of AML patients more specifically and sensitively. Then, using a nomogram, the fatty acid metabolism gene prognostic model, age, and cytogenetic risk degree were integrated into a scoring system to more accurately predict the survival of AML patients, with broad clinical application prospects.

Description

Technical Field

[0001] The present invention belongs to the field of biomedicine, and particularly relates to a prognostic model for predicting the overall survival rate of patients with acute myeloid leukemia based on 9 fatty acid metabolism-related genes. Background Art

[0002] Acute myeloid leukemia (AML) is a hematopoietic tumor characterized by clonal proliferation, differentiation arrest, and inhibition of apoptosis of malignant stem cells or progenitor cells. AML has a poor clinical prognosis and a high mortality rate, and its overall 5-year survival rate is less than 50%. The standardized chemotherapy regimen is the main treatment method, but sometimes it cannot achieve a complete remission (CR), and relapse may also occur after CR. Although current treatment regimens with new targeted drugs such as demethylating agents and BCL2 inhibitors have been proposed, the treatment effect is still not very satisfactory. Therefore, for patients with acute myeloid leukemia, there is an urgent clinical need for effective prognostic judgment indicators and disease monitoring indicators to provide effective treatment and prognostic prediction programs.

[0003] In recent years, metabolic reprogramming has been recognized as a fundamental feature of tumors. The previously proposed Warburg effect indicates that cancer cells mainly obtain energy for rapid proliferation through aerobic glycolysis; recent studies have found that mitochondrial metabolism also plays a crucial role in tumor growth, and fatty acids play a key role as the main raw material for mitochondrial metabolism. Existing studies have shown that fatty acid metabolism plays a key role in the survival of AML cells, especially leukemia stem cells, and in the resistance to targeted drugs such as BCL2 inhibitors. Fatty acid metabolism-related genes are closely related to the occurrence and development of AML, but there has been no report on constructing a model based on fatty acid metabolism-related genes to stratify the prognosis of AML patients.

[0004] Most existing studies use single genes as prognostic indicators, and there are few multi-gene prognostic models constructed in AML. In particular, there is no relevant research on fatty acid prognostic models in AML, and we further integrate the fatty acid metabolism model with classical prognostic factors (age and cytogenetic risk level) to more accurately comprehensively evaluate the prognosis of AML patients. Summary of the Invention

[0005] In view of the above problems, the present invention provides the application of a kit for detecting the expression levels of 9 fatty acid metabolism genes in the preparation of a product for diagnosing or assisting in the diagnosis of the overall survival rate of patients with acute myeloid leukemia (AML). The risk score based on 9 fatty acid metabolism genes can well divide AML patients into high-risk and low-risk groups, which may help in the selection of clinical treatment plans.

[0006] To achieve the above object, the present invention provides the following technical solutions.

[0007] The present invention provides an overall survival prognosis model for patients with acute myeloid leukemia, characterized in that the model is based on 9 fatty acid metabolism genes.

[0008] Further, the 9 fatty acid metabolism genes are MLYCD, CYP4F2, SLC25A1, PLA2G4A, ACBD4, ACOT7, ACSF2, CBR1, and ACSL5.

[0009] Further, the prognosis model performs risk scoring by detecting the expression levels of 9 fatty acid metabolism genes in samples of patients with acute myeloid leukemia: Risk score = (0.299 * expression level of SLC25A1) - (1.090 * expression level of MLYCD) - (0.394 * expression level of CYP4F2A) + (0.474 * expression level of PLA2G4A) + (0.488 * expression level of ACBD4) + (0.538 * expression level of ACOT7) + (0.566 * expression level of ACSF2) + (0.632 * expression level of CBR1) + (0.750 * expression level of ACSL5).

[0010] Further, the above-mentioned overall survival prognosis model for patients with acute myeloid leukemia is used in the preparation of products for diagnosing or assisting in diagnosing the overall survival of patients with acute myeloid leukemia.

[0011] The present invention also provides a method for constructing a nomogram for acute myeloid leukemia based on a prognosis model of 9 fatty acid metabolism genes, age, and cytogenetic risk level, characterized in that the method comprises the following steps:

[0012] S1 Screen 9 fatty acid metabolism genes with prognostic value for acute myeloid leukemia: MLYCD, CYP4F2, SLC25A1, PLA2G4A, ACBD4, ACOT7, ACSF2, CBR1, and ACSL5;

[0013] S2 Establish an overall survival prognostic risk scoring model for acute myeloid leukemia based on the expression levels of the 9 fatty acid metabolism genes described in S1: Risk score = (0.299 * expression level of SLC25A1) - (1.090 * expression level of MLYCD) - (0.394 * expression level of CYP4F2A) + (0.474 * expression level of PLA2G4A) + (0.488 * expression level of ACBD4) + (0.538 * expression level of ACOT7) + (0.566 * expression level of ACSF2) + (0.632 * expression level of CBR1) + (0.750 * expression level of ACSL5);

[0014] S3 Generate a nomogram by adding up the scores corresponding to the risk score, age, and cytogenetic risk degree of each patient to obtain the total score.

[0015] Furthermore, the method for screening 9 fatty acid metabolism genes with prognostic value for acute myeloid leukemia in S1 includes the following steps:

[0016] (1) Download the RNA sequencing datasets and clinical data of the TCGA AML database and the VIZOME AML database;

[0017] (2) Obtain fatty acid metabolism-related genes from Gene Set Enrichment Analysis, perform prognostic analysis on fatty acid metabolism-related genes using Cox regression analysis and survival analysis, with P < 0.05 as the cut-off value for screening prognostic-related genes, and screen out 9 fatty acid metabolism genes with prognostic value.

[0018] Furthermore, the method for establishing an overall survival prognostic risk scoring model for acute myeloid leukemia in S2 includes the following steps:

[0019] Calculate the individualized risk score using the regression coefficients of the 9 fatty acid metabolism genes screened in S1 in the training set TCGA, and divide AML patients into high-risk and low-risk groups according to the risk score; use ROC to calculate the area under the curve AUC at multiple time points to evaluate the discrimination ability of the prognostic model; then verify the accuracy of the model in the test set VIZOME using the same risk score formula and cut-off value.

[0020] Furthermore, the 1-year, 2-year, and 3-year survival rates corresponding to the total score of the nomogram in S3 are respectively the predicted 1-year, 2-year, and 3-year survival rates of AML patients.

[0021] Further, the application of the nomogram constructed by the method in the preparation of a product for diagnosing or assisting in diagnosing the overall survival rate of patients with acute myeloid leukemia.

[0022] The beneficial effects of the present invention compared with the prior art.

[0023] 1. For the first time, a prognostic evaluation model is constructed using fatty acid metabolism-related genes that play important roles in AML. This model can evaluate the prognosis of AML patients more specifically and sensitively and has been verified in the test set.

[0024] 2. Using the nomogram, the fatty acid metabolism model, age, and cytogenetic risk are integrated into a scoring system to more accurately predict the survival of AML patients.

[0025] 3. The present invention for the first time proposes the concept of constructing a prognostic evaluation model using fatty acid metabolism genes. Compared with single-gene prognosis evaluation, it is more accurate and comprehensive; by combining traditional prognostic factors such as age and cytogenetic risk to jointly evaluate prognosis, it is more reliable. Brief Description of the Drawings

[0026] Figure 1 Screen fatty acid metabolism genes for constructing a risk model.

[0027] Figure 2 A risk score is established based on the regression coefficients of 9 genes, and patients are divided into low-risk and high-risk groups. Among them, Figure A shows the comparison of OS between the high-risk group and the low-risk group in the training set; Figure B shows the comparison of OS between the high-risk group and the low-risk group in the test set; Figure C shows the ROC curve analysis of the one-year survival period between the high-risk group and the low-risk group in the training set; Figure D shows the ROC curve analysis of the two-year survival period between the high-risk group and the low-risk group in the training set; Figure E shows the ROC curve analysis of the three-year survival period between the high-risk group and the low-risk group in the training set; Figure F shows the ROC curve analysis of the one-year survival period between the high-risk group and the low-risk group in the test set; Figure G shows the ROC curve analysis of the two-year survival period between the high-risk group and the low-risk group in the test set; Figure H shows the ROC curve analysis of the three-year survival period between the high-risk group and the low-risk group in the test set.

[0028] Figure 3 Univariate and multivariate Cox proportional hazards regression analyses are applied to the training set and the test set. Among them, Figure A shows the univariate Cox analysis of various prognostic indicators of AML patients in the training set; Figure B shows the univariate Cox analysis of various prognostic indicators of AML patients in the test set; Figure C shows the multivariate Cox analysis of various prognostic indicators of AML patients in the training set; Figure D shows the multivariate Cox analysis of various prognostic indicators of AML patients in the test set; Figure E shows the comparison of the ROC curve analysis of the three-year survival period of various prognostic indicators in the training set; Figure F shows the comparison of the ROC curve analysis of the three-year survival period of various prognostic indicators in the test set.

[0029] Figure 4 The nomogram predicts the prognosis of AML patients. Among them, in Figure A, scores are assigned according to the contribution of each factor to the prognosis; in Figure B, the calibration curve analyzes the matching of actual and expected survival rates; in Figure C, the calibration curve shows the consistency analysis of actual and expected survival rates. Detailed implementation manners

[0030] In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be further described below in conjunction with specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other. Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Therefore, the present invention is not limited by the specific embodiments disclosed in the following specification.

[0031] Example 1.

[0032] As Figures 1-4 shown, this embodiment provides a method for constructing a model for predicting the survival possibility of AML patients and an effect verification.

[0033] 1.1 Data download and preprocessing.

[0034] Download the RNA sequencing dataset and clinical data from TCGA. This dataset contains data of 155 AML patients with eligible clinical information. Download the RNA sequencing dataset and clinical data from VIZOME. This dataset contains data of 299 AML patients with eligible clinical information.

[0035] 1.2 Screening of fatty acid metabolism genes related to prognosis.

[0036] In order to construct a prognostic model related to fatty acid metabolism, 201 fatty acid metabolism-related genes were obtained from Gene Set Enrichment Analysis (GSEA) (https: / / www.gsea-msigdb.org / gsea / index.jsp).

[0037] Perform univariate Cox prognostic analysis on 201 genes and find that the risk factors of 37 genes are greater than 1. Perform Log-rank prognostic analysis on 37 prognosis-related genes, and use P<0.05 as the cut-off value for screening prognosis-related genes, and determine 9 fatty acid metabolism genes for constructing a prognostic model.

[0038] 1.3 Construction of a fatty acid metabolism prognostic model.

[0039] In the training set TCGA, we calculated the individualized risk score using the regression coefficient of each gene and divided AML patients into high-risk and low-risk groups based on the risk score. The area under the curve (AUC) at multiple time points was calculated using ROC (the receiver operating characteristic) to evaluate the discriminative ability of the prognostic model. The following model was proposed:

[0040] Risk score = (0.299 * expression level of SLC25A1) - (1.090 * expression level of MLYCD) - (0.394 * expression level of CYP4F2A) + (0.474 * expression level of PLA2G4A) + (0.488 * expression level of ACBD4) + (0.538 * expression level of ACOT7) + (0.566 * expression level of ACSF2) + (0.632 * expression level of CBR1) + (0.750 * expression level of ACSL5).

[0041] Then, the accuracy of the model was verified using the same risk score formula and cut-off value in the test set.

[0042] 1.4 Prognostic value of the risk score by multivariate regression analysis.

[0043] Clinical prognostic indicators of AML patients were collected from the TCGA database, including survival time, survival status, age, FLT3 mutation status, NPM1 mutation status, white blood cell count, and cytogenetic risk degree. Univariate and multivariate cox regression analyses were performed using clinical data and risk score to evaluate whether the prognostic value of the risk score was independent of other clinical prognostic indicators. A p < 0.05 value was considered statistically significant.

[0044] 1.5 Construction of a nomogram.

[0045] A nomogram was constructed using the R package based on age, cytogenetic risk degree, and risk score. Then, a calibration curve was plotted to evaluate the consistency between actual and expected survival times.

[0046] 1.6 Statistical analysis.

[0047] Statistical analysis was performed using R software (version 3.6.1; https: / / www.R-project.org), and statistical significance was set at p < 0.05. Survival analysis was performed using Kaplan-Meier curve analysis and univariate Cox regression analysis. Multivariate Cox regression analysis was used to determine independent prognostic factors. Time-dependent ROC analysis was used to evaluate the accuracy of the prognostic prediction model. An AUC > 0.60 was considered acceptable for prediction.

[0048] 2. Results.

[0049] 2.1 Construction of a prognostic model for AML patients using genes related to fatty acid metabolism.

[0050] The TCGA dataset was used as the training set, and AML patients in the training set were used to construct the prognostic model. We used Cox proportional hazards regression analysis and Log-rank prognostic analysis on the training set to screen fatty acid metabolism genes closely related to prognosis, and finally selected 9 genes to construct the risk model ( Figure 1 )

[0051] Subsequently, the risk score of each patient in the training group was calculated based on the regression coefficients of the 9 genes. Using the median of the scores as the boundary, the patients were divided into high-risk and low-risk groups. Higher than the critical value was high risk, and lower than or equal to the critical value was low risk. Risk score = (0.299 * expression level of SLC25A1) - (1.090 * expression level of MLYCD) - (0.394 * expression level of CYP4F2A) + (0.474 * expression level of PLA2G4A) + (0.488 * expression level of ACBD4) + (0.538 * expression level of ACOT7) + (0.566 * expression level of ACSF2) + (0.632 * expression level of CBR1) + (0.750 * expression level of ACSL5). The risk score significantly stratified the training group, clearly dividing the patients into low-risk and high-risk groups. Our data showed that the OS of the high-risk group was significantly lower than that of the low-risk group (P < 0.0001) ( Figure 2 A). ROC curve analysis ( Figure 2 C, D, E) showed that at 1-year, 2-year, and 3-year follow-ups, the AUCs were 0.8297, 0.8392, and 0.8130, respectively, indicating acceptable discrimination.

[0052] 2.2 Validation of the prognostic model using the test group data.

[0053] We used the risk score constructed in the training set above to divide the patients in the test set into high-risk and low-risk groups. In the test group, we verified the clinical utility and discriminative ability of the model.

[0054] The risk score significantly stratified the test group, dividing the patients into low-risk and high-risk groups. The OS of the high-risk group was significantly lower than that of the low-risk group (P < 0.05) ( Figure 2 B). ROC curve analysis ( Figure 2F, G, and H) showed that at 1-year, 2-year, and 3-year follow-ups, the AUCs were 0.6560, 0.6649, and 0.6663, respectively, representing acceptable discrimination. The test set data verified that the prognostic model had high reliability.

[0055] 2.3 Risk score as an independent prognostic factor for AML patients.

[0056] To further explore whether the risk score could be used as an independent clinical prognostic factor, univariate and multivariate Cox proportional hazards regression analyses were applied to the training set and the test set. As Figure 3 shown in A and B, univariate analysis found that the risk score, age, and cytogenetic risk were risk factors for the overall survival of AML patients. As Figure 3 shown in C and D, multivariate analysis found that the risk score was an independent prognostic factor for the overall survival of AML patients (HR = 4.2 [2.6 - 6.8], p < 0.0001). Therefore, the risk score is an independent prognostic factor for AML patients. Moreover, compared with the other two independent prognostic indicators (age and cytogenetic risk), the risk score had relatively high efficacy in assessing the 3-year survival of AML patients ( Figure 3 E and F).

[0057] 2.5 Nomogram for predicting the prognosis of AML patients.

[0058] A nomogram is a powerful tool that has been used to quantitatively determine individual risks in clinical settings by integrating multiple risk factors. By combining the risk score, age, and cytogenetic risk, we generated a nomogram to predict the 1-year, 2-year, and 3-year prognoses. According to the scores corresponding to the risk score, age, and cytogenetic risk of each patient, the scores were added together to obtain the total score. In the nomogram, the 1-year, 2-year, and 3-year survival rates corresponding to the total score were the predicted survival rates of AML patients at 1 year, 2 years, and 3 years. As Figure 4 shown in A, scores were assigned to each factor according to its contribution to the prognosis. The calibration curve showed a match between the actual and expected survival rates ( Figure 4 B), including the 1-year, 2-year, and 3-year survival periods. The study found that the calibration curve showed a high consistency between the actual and expected survival rates ( Figure 4 C), indicating that our prediction model had high credibility.

Claims

1. A prognostic model for the overall survival rate of patients with acute myeloid leukemia, characterized in that, the model is based on 9 fatty acid metabolism genes; the 9 fatty acid metabolism genes are MLYCD, CYP4F2, SLC25A1, PLA2G4A, ACBD4, ACOT7, ACSF2, CBR1, and ACSL5; the prognostic model performs risk scoring by detecting the expression levels of 9 fatty acid metabolism genes in samples of patients with acute myeloid leukemia: Risk score = (0.299 * expression level of SLC25A1) - (1.090 * expression level of MLYCD) - (0.394 * expression level of CYP4F2A) + (0.474 * expression level of PLA2G4A) + (0.488 * expression level of ACBD4) + (0.538 * expression level of ACOT7) + (0.566 * expression level of ACSF2) + (0.632 * expression level of CBR1) + (0.750 * expression level of ACSL5).

2. Use of the prognostic model for the overall survival rate of patients with acute myeloid leukemia according to claim 1 in the preparation of a product for diagnosing or assisting in the diagnosis of the overall survival rate of patients with acute myeloid leukemia.

3. A method for constructing a nomogram for acute myeloid leukemia based on a prognostic model of 9 fatty acid metabolism genes, age, and cytogenetic risk degree, characterized in that, the method comprises the following steps: S1 Screen 9 fatty acid metabolism genes with prognostic value for acute myeloid leukemia: MLYCD, CYP4F2, SLC25A1, PLA2G4A, ACBD4, ACOT7, ACSF2, CBR1, and ACSL5; S2 Establish a prognostic risk scoring model for the overall survival rate of acute myeloid leukemia based on the expression levels of the 9 fatty acid metabolism genes described in S1: Risk score = (0.299 * expression level of SLC25A1) - (1.090 * expression level of MLYCD) - (0.394 * expression level of CYP4F2A) + (0.474 * expression level of PLA2G4A) + (0.488 * expression level of ACBD4) + (0.538 * expression level of ACOT7) + (0.566 * expression level of ACSF2) + (0.632 * expression level of CBR1) + (0.750 * expression level of ACSL5); S3 Generate a nomogram by adding up the scores corresponding to the risk score, age, and cytogenetic risk degree of each patient to obtain a total score.

4. The method for constructing a nomogram for acute myeloid leukemia based on a prognostic model of 9 fatty acid metabolism genes, age, and cytogenetic risk degree according to claim 3, characterized in that, the method for screening 9 fatty acid metabolism genes with prognostic value for acute myeloid leukemia in S1 comprises the following steps: (1) Download the RNA sequencing datasets and clinical data of the TCGA AML database and the VIZOME AML database; (2) Obtain fatty acid metabolism-related genes from Gene Set Enrichment Analysis, and perform prognostic analysis on fatty acid metabolism-related genes using Cox regression analysis and survival analysis. P < 0.05 is used as the cut-off value for screening prognostic-related genes, and 9 fatty acid metabolism genes with prognostic value are screened.

5. A method for constructing a nomogram for acute myeloid leukemia based on a prognostic model of 9 fatty acid metabolism genes, age, and cytogenetic risk degree according to claim 3, characterized in that, The method for establishing a prognostic risk score model for the overall survival rate of acute myeloid leukemia in S2 includes the following steps: Calculate the individualized risk score using the regression coefficients of the 9 fatty acid metabolism genes screened in S1 in the training set TCGA, and divide AML patients into high-risk and low-risk groups according to the risk score; Use ROC to calculate the area under the curve AUC at multiple time points to evaluate the discrimination ability of the prognostic model; then verify the accuracy of the model in the test set VIZOME using the same risk score formula and cut-off value.

6. A method for constructing a nomogram for acute myeloid leukemia based on a prognostic model of 9 fatty acid metabolism genes, age, and cytogenetic risk degree according to claim 3, characterized in that, The 1-year, 2-year, and 3-year survival rates corresponding to the total nomogram score in S3 are respectively the predicted 1-year, 2-year, and 3-year survival rates of AML patients.

7. Use of a method for constructing a nomogram for acute myeloid leukemia based on a prognostic model of 9 fatty acid metabolism genes, age, and cytogenetic risk degree according to claim 3 in the preparation of a product for diagnosing or assisting in diagnosing the overall survival rate of acute myeloid leukemia patients.

Images