A Model and Construction Method for Prognostic and Therapeutic Adaptability Assessment of Hepatocellular Carcinoma Based on mRNA Vaccine Antigens

Abstract

This invention discloses a prognostic and therapeutic suitability assessment model for hepatocellular carcinoma (HCC) based on mRNA vaccine antigens, and its construction method. First, the differences in gene expression between normal tissues and HCC tumors are analyzed to understand the mutations and genomic structural changes in HCC patients. Then, genes related to the level of antigen-presenting cell infiltration, as well as genes significantly related to overall survival and disease-free survival, are further selected from anomalously expressed and mutated genes to obtain candidate mRNA vaccine neoantigen targets. Based on the expression levels of these targets, patients are immunophenotyped to assess the patient population suitable for mRNA vaccines. Simultaneously, the relationship between target expression levels and patient prognosis is quantified to predict the probability of HCC patients achieving 3-year and 5-year overall survival. This invention can objectively and accurately assess treatment resistance and tumor immune status in HCC, improving the predictive accuracy of HCC treatment prognosis.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of biological medicine, and relates to prognosis evaluation of hepatocellular carcinoma, in particular to a hepatocellular carcinoma prognosis and treatment suitability evaluation model based on mRNA vaccine antigens and a construction method. BACKGROUND

[0002] Hepatocellular carcinoma (HCC) has the characteristics of high malignancy and extremely poor prognosis, and the survival time of patients after the appearance of symptoms is usually less than 1 year. The current relatively mature HCC treatment methods include chemotherapy using sorafenib, vascular cannulation, radiofrequency ablation, surgical resection and liver transplantation. However, drug resistance, shortage of liver transplant donors and side effects are still obstacles in the treatment of HCC.

[0003] Therapeutic cancer vaccines aim to stimulate the adaptive immune system of patients against specific tumor antigens to control the speed of tumor progression, and even eradicate micro residual lesions, which is expected to have a synergistic effect with existing long-term cancer treatment immunotherapy. Therapeutic cancer vaccines include tumor or immune cell-based vaccines, peptide-based vaccines, virus vector-based vaccines and nucleic acid-based vaccines. Due to difficulties in antigen prediction and poor immunogenicity, the clinical transformation of mRNA (Messenger RNA) vaccines is plagued. Therefore, in-depth understanding of tumor heterogeneity and tumor immune microenvironment, identification and verification of predictive biomarkers are crucial for identifying HCC patients suitable for mRNA vaccination and prognosis prediction. SUMMARY

[0004] In order to judge the overall survival of hepatocellular carcinoma patients, evaluate patients suitable for mRNA vaccine, and assess the potential of tumor cell immune escape and the degree of immune cell infiltration in the tumor, and improve the prediction ability of HCC immunotherapy response, the present application provides a hepatocellular carcinoma prognosis and treatment suitability evaluation model based on mRNA vaccine antigens and a construction method. By selecting new antigen targets of hepatocellular carcinoma mRNA vaccines, the model is constructed to divide the immune subtypes of patients, improve the accuracy of screening patients suitable for mRNA vaccine treatment, predict the potential of tumor cell immune escape and the degree of immune cell infiltration in the tumor, evaluate the overall survival and prognosis of hepatocellular carcinoma patients, thereby reducing the incidence of adverse events and improving the efficacy of immunotherapy.

[0005] The hepatocellular carcinoma prognosis and treatment suitability evaluation model based on mRNA vaccine antigens first divides patients into different immune subtypes by analyzing the expression levels of candidate mRNA vaccine new antigen target genes of patients using consistency clustering method. The suitability of mRNA vaccine therapy for patients is evaluated based on immune subtypes. Secondly, the probability values of the overall survival time of hepatocellular carcinoma patients reaching 3 years and 5 years are predicted by combining the expression levels of candidate mRNA vaccine new antigen target genes with the linear relationship with prognosis.

[0006] The candidate mRNA vaccine neoantigen target gene includes CCNB1, CDC25C, PTTG1, CHEK1, KPNA2, MKI67, KIF2C, MCM3, EZH2, CDT1, PES1, PRC1, PPM1G, CDK1, NEK2, TRIP13, TUBG1, AURKA and G6PD.

[0007] The method for constructing a liver cancer prognosis and treatment suitability evaluation model based on mRNA vaccine antigens specifically comprises the following steps:

[0008] Step 1, collect liver cancer tumor samples and paracancer normal tissue samples, extract tissue RNA for purification and sequencing, and randomly divide them into a training set and a validation set.

[0009] Step 2, analyze the gene expression difference between the paracancer normal tissue samples and the liver cancer tumor samples, and collect the abnormally expressed genes.

[0010] Step 3, understand the mutations and genomic structural changes of liver cancer patients, and collect the mutant genes.

[0011] Step 4, further select genes related to antigen presenting cell (APC) infiltration level and genes with significant relationship with overall survival (OS) and recurrence-free survival (RFS) of patients from the liver cancer abnormally expressed genes collected in step 2 and the mutant genes collected in step 3, respectively, and mark them as genes affecting APC infiltration and genes related to survival. Select the intersection of mutant genes, abnormally expressed genes, genes related to survival and genes affecting APC infiltration as the candidate mRNA vaccine neoantigen target.

[0012] Step 5, based on the expression level of the candidate mRNA vaccine neoantigen target, perform consistent clustering analysis on HCC patients, and divide the HCC patients into different immune subtypes. Compare the overall survival of HCC patients in different immune subtypes to verify the clustering scheme.

[0013] Step 6, compare the tumor immune microenvironment characteristics between immune subtypes, select the immune subtype mainly composed of immunosuppressive immune cells, and take the patients in this immune subtype as the model output suitable for mRNA vaccine treatment.

[0014] As a preferred, the immune microenvironment characteristics include the expression levels of immune checkpoint and immunogenic cell death related genes, tumor mutation burden (TMB) and infiltration abundance of immune cells.

[0015] Step 7: Construct a Nomogram model using multivariate Cox regression analysis. Use the expression level of candidate mRNA vaccine neoantigen targets in hepatocellular carcinoma (HCC) patients as the input factor for the Nomogram model. Based on the input clinical prognostic data, quantify the linear relationship between the target and the prognosis of HCC patients. The model is used to predict the probability of overall survival of 3 years and 5 years based on the expression level of candidate mRNA vaccine neoantigen targets in HCC patients.

[0016] The probability value is the value on the probability axis that corresponds perpendicularly to the position of the sum of the point values ​​corresponding to the input factors in the Nomogram model on the total point value axis.

[0017] The present invention has the following beneficial effects:

[0018] (1) Based on patient prognosis and the infiltration of antigen-presenting cells, potential antigens of hepatocellular carcinoma (HCC) can be identified, thereby screening patients suitable for mRNA vaccination, improving treatment efficiency and individualization, and providing theoretical support for promoting the application of mRNA vaccines in the field of anti-cancer.

[0019] (2) Gain a deeper understanding of the tumor immune microenvironment (TIME), identify differences in the tumor microenvironment among different immune subtypes, and improve the accuracy of predicting the efficacy of mRNA-based immunotherapy.

[0020] (3) By detecting the expression levels of mRNA vaccine candidate targets in hepatocellular carcinoma (HCC) tumor tissues and constructing a Nomogram model through multivariate Cox regression analysis, the relationship between targets and HCC patient prognosis can be quantified, enabling an objective assessment of overall survival in HCC patients. Compared with existing next-generation sequencing technologies for predicting patient prognosis, this method reduces the number of genes that need to be detected, improves efficiency, and lowers costs. Attached Figure Description

[0021] Figure 1 Chromosomal distribution of upregulated and downregulated genes in hepatocellular carcinoma tumor samples compared to normal samples;

[0022] Figure 2 The mutation count and frequency distribution of genes with genomic alterations in hepatocellular carcinoma tumor samples;

[0023] Figure 3 This is a screening diagram of potential neoantigens for mRNA vaccines in the examples;

[0024] Figure 4 Clustering results of immune subtypes in hepatocellular carcinoma patients based on potential neoantigens from mRNA vaccines;

[0025] Figure 5 To train the survival curves of patients with different immune subtypes;

[0026] Figure 6 Box plots for training the expression of immune checkpoint genes;

[0027] Figure 7 Box plots of expression of immunogenic cell death-related genes were used for training.

[0028] Figure 8 To train a violin plot of tumor mutational burden;

[0029] Figure 9 Box plots were used to train the differential infiltration of immune cells.

[0030] Figure 10 A violin diagram showing alpha-fetoprotein (AFP) levels for training purposes;

[0031] Figure 11 To validate the survival curves and violin plots of alpha-fetoprotein levels in patients with different immune subtypes;

[0032] Figure 12 The results of the immune score for the validation set;

[0033] Figure 13 For the prediction results and evaluation of the survival prediction model. Detailed Implementation

[0034] The present invention will be further explained below with reference to the accompanying drawings;

[0035] This invention proposes a prognostic and therapeutic suitability assessment model and construction method for hepatocellular carcinoma (HCC) based on mRNA vaccine antigens. The primary objective is to screen for neoantigen targets for hepatocellular carcinoma mRNA vaccines, classify HCC patients into immune subtypes through tumor tissue analysis, thereby predicting the efficacy of mRNA vaccines and selecting suitable patient populations for mRNA vaccine therapy to aid clinical decision-making. The second objective is to predict the overall survival of HCC patients based on the screened neoantigen targets for hepatocellular carcinoma mRNA vaccines, using tumor tissue analysis. The third objective is to improve the accuracy of selecting suitable patients for mRNA vaccine therapy by classifying patient samples into immune subtypes, comparing different immune characteristics among these subtypes, predicting the degree of immune cell infiltration within the tumor and the potential for tumor cell immune escape, thereby reducing the incidence of adverse events and improving the efficacy of immunotherapy.

[0036] This embodiment uses sequencing data from the Cancer Genome Atlas Database and the International Cancer Genome Consortium Database as training and validation sets, respectively. Data analysis was performed using R software, employing paired / unpaired Student's t-tests, one-way ANOVA, or two-way ANOVA. Statistical significance was set at p < 0.05, and all p-values ​​were two-tailed. This demonstrates the method for establishing a prognostic and therapeutic suitability assessment model for hepatocellular carcinoma based on mRNA vaccine antigens. The model was then evaluated against tumor TNM staging, histological grade, and alpha-fetoprotein levels on the validation set, demonstrating the effectiveness of the proposed model. The specific steps are as follows:

[0037] Step 1: Select pathological diagnoses of HCC from the Cancer Genome Atlas database, as well as patient data containing survival status and overall survival. Use sequencing data of hepatocellular carcinoma tumor tissue from patients as hepatocellular carcinoma tumor samples, and sequencing data of adjacent normal tissue as normal samples, to obtain a total of 328 training samples.

[0038] Step 2: Gene expression differences between normal samples and hepatocellular carcinoma tumor samples were analyzed using the limma R package. Using |log2FC| values ​​> 1 and q values ​​< 0.01 as screening criteria, 1475 genes abnormally overexpressed in hepatocellular carcinoma were identified, such as... Figure 1 As shown.

[0039] Step 3: Considering the close correlation between tumor mutational burden (TMB) and copy number alterations (CNA) and patient prognosis and immunotherapy efficacy, the cBio Cancer Genome Portal was used to understand mutations and genomic structural changes in liver cancer patients, evaluating 13,640 mutated genes that may encode tumor-associated antigens. In the HCC cohort, the highest proportion of low mutation numbers (40–50) was observed, suggesting that HCC may have low immunogenicity. Figure 2 As shown in (a) and (b) in the figure.

[0040] Step 4: Using the Tumor Immunological Estimation Resource (TIMER), select genes associated with antigen-presenting cell (APC) invasion levels from mutated genes and abnormally highly expressed genes in hepatocellular carcinoma; these are designated as genes affecting APC invasion. Then, using the Gene Expression Analysis Interactive Platform (GEPIA), select genes from mutated genes and abnormally highly expressed genes in hepatocellular carcinoma that have a significant relationship with overall survival (OS) and disease-free survival (RFS); these are designated as survival-related genes.

[0041] like Figure 3As shown, the intersection of mutated genes, abnormally expressed genes, survival-related genes, and genes affecting APC invasion was selected to obtain 19 overexpressed and mutated genes that are significantly associated with patient survival. These genes are CCNB1, CDC25C, PTTG1, CHEK1, KPNA2, MKI67, KIF2C, MCM3, EZH2, CDT1, PES1, PRC1, PPM1G, CDK1, NEK2, TRIP13, TUBG1, AURKA, and G6PD. These 19 genes were selected as candidate neoantigen targets for mRNA vaccines.

[0042] Step 5: Convert the expression levels of candidate mRNA vaccine neoantigen targets in the training set into matrix form. Set the number of cluster repetitions (reps = 100) to ensure the stability of the clustering results. The sample proportion (pItem = 0.8) indicates that 80% of the samples are randomly selected for clustering in each repetition. Set pFeature = 1, selecting k-means as the clustering algorithm. Use the R package "ConsensusClusterPlus" to perform consensus clustering analysis on HCC patients. By comprehensively analyzing the consensus matrix and the consensus cumulative distribution function, HCC patients are divided into immune subtypes 1, 2, and 3. The results are as follows: Figure 4 As shown.

[0043] The overall survival of HCC patients with different immune subtypes in the training set was compared using the Kaplan-Meier method and log-rank test. The results are as follows: Figure 5 As shown, patients with immune subtype 1 had significantly longer overall survival (P<0.001), while patients with immune subtype 3 had a poorer prognosis. This indicates that patients with different immune subtypes have different prognoses, suggesting that this immune subtype classification scheme is reasonable.

[0044] Step 6: Significant differences exist in tumor mutational burden, immune checkpoint expression levels, and expression of immunogenic cell death-related genes among different immune subtypes. These differences in the tumor microenvironment are crucial for predicting the efficacy of mRNA-based vaccine-based immunotherapy. Immune scores are calculated in the training set to determine the immune mechanisms of different immune subtypes.

[0045] (1) By comparing the expression levels of 47 immune checkpoint-related genes in HCC patients with different immune subtypes through differential analysis, it was found that 36 genes showed significant differences in different subtypes, such as PD-1, PD-L1, BTNL2, CD160, CD200R1, CD244, and CD27. Furthermore, a large proportion of these genes were highly expressed in immune subtype 3, such as... Figure 6 As shown in the figure. Box plot of expression of immunogenic cell death-related genes. Figure 7As shown, 18 immunogenic cell death-related genes showed significant differences across different subtypes.

[0046] (2) The total number of somatic nonsynonymous mutations in the coding region is calculated as the tumor mutational burden (TMB), including missense, nonsense, splice site, and frameshift mutations. The results are as follows: Figure 8 As shown, immune subtype 3 has the highest tumor mutation burden.

[0047] (3) Based on transcriptome data, the CIBERSORT algorithm was used to quantify the infiltration abundance of 23 immune cells in the tumor immune microenvironment. The results are as follows: Figure 9 As shown, it can be clearly observed that cytotoxic cells such as CD8+ T cells, γδ T cells (Tgd) and natural killer cells are estimated to be higher in patients with immune subtype 1, while suppressive immune cells, including immature dendritic cells (iDCs) and CD56 bright natural killer (NK) cells, are at higher levels in patients with immune subtype 3.

[0048] The results above lead to the conclusions that: Immunoassay subtype 1 is primarily infiltrated by cytotoxic immune cells, representing an immunologically "hot" subtype, indicating intolerance to mRNA vaccine therapy and suggesting avoidance of this treatment; while immunoassay subtype 3 is dominated by immunosuppressive immune cells, representing a relatively "cold" subtype, showing higher adaptability and making it an ideal candidate for mRNA vaccine therapy. Immunoassay subtype 2 is in an intermediate state, requiring a comprehensive assessment of its treatment suitability in conjunction with other clinical factors. Testing the alpha-fetoprotein levels of different immunoassay subtypes, such as... Figure 10 As shown, the different characteristics between subtypes can also be explained by the alpha-fetoprotein level. Therefore, it can be inferred that these three immune subtypes represent different immune mechanisms of HCC, indicating that the model proposed in this invention can distinguish the immune subtypes of different HCC patients. The final model output is that patients belonging to immune subtype 3 are suitable for the use of mRNA vaccines.

[0049] Step 7: Keeping all parameter settings unchanged, the 203 samples from the validation set of the International Cancer Genome Consortium database were divided into immune subtype 1, immune subtype 2, and immune subtype 3. Overall survival was compared among the different immune subtypes using survival analysis. Results are as follows: Figure 11 As shown in (a) of the data, the validation set yields the same conclusions as the training set: patients with immune subtype 1 have significantly longer overall survival (p<0.0001), and patients with immune subtype 3 have a poorer prognosis. The alpha-fetoprotein levels of different immune subtypes are shown in Figure (a). Figure 11 As shown in (b) of the diagram.

[0050] Perform immune scoring on the validation set. For example... Figure 12As shown in (a), 20 immune checkpoint-related genes differ among different immune subtypes, possibly due to differences in ethnicity and overall cohort prognosis. Figure 12 As shown in (b) of the figure, ten immunogenic cell death-related genes, including CALR, EIF2A, EIF2AK1, EIF2AK2, EIF2AK3, HMGB1, ZFNAR2, P2RX7, P2RY2, and TLR3, showed significant differences among different immune subtypes. Figure 12 As shown in (c), the infiltration level of immune cells in the validation set is consistent with the overall trend of the training set, and the difference is negligible.

[0051] Step 8: A survival prediction model was constructed based on 19 candidate mRNA vaccine neoantigen targets. Multivariate Cox regression analysis was performed, and a nomogram was plotted using the R package "rms". Clinical data was used to calibrate the model's predictions for consistency at 500 and 1000 days. The results are as follows: Figure 13 As shown in the calibration curves, the proposed nomogram demonstrates good adaptability in survival prediction. The coefficients corresponding to the 19 candidate mRNA vaccine neoantigen targets are fitted and are shown in Table 1.

[0052] Table 1

[0053]

[0054]

[0055] The model calculates the probability values ​​of overall survival of HCC patients reaching 3 years and 5 years based on the expression levels of candidate mRNA vaccine neoantigen target genes and the correlation coefficients in Table 1.

[0056] A prognostic and therapeutic suitability assessment model for hepatocellular carcinoma based on mRNA vaccine antigens uses the expression levels of candidate mRNA vaccine neoantigen targets in patients as input. First, the expression profiles of target genes are clustered using the R package "ConsensusClusterPlus," classifying patients into three different immune subtypes. Patients belonging to immune subtype 3 are output as ideal candidates for mRNA vaccine therapy. Second, combining the linear relationship between the expression levels of candidate mRNA vaccine neoantigen target genes and prognosis shown in Table 1, the model calculates and outputs the probability values ​​for overall survival of hepatocellular carcinoma patients reaching 3 and 5 years.

[0057] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for constructing a liver cancer prognosis and treatment suitability evaluation model based on mRNA vaccine antigens, characterized in that: Specifically comprising the following steps: Step 1, collect hepatocellular carcinoma tumor samples and paracancer normal tissue samples, extract tissue RNA for purification and sequencing; Step 2, analyze the gene expression difference between paracancer normal tissue samples and hepatocellular carcinoma tumor samples, collect abnormally expressed genes; Step 3, collect mutant genes of hepatocellular carcinoma patients; Step 4, select genes related to antigen presenting cell infiltration level and genes with significant relationship with overall survival and disease-free survival of patients from abnormally expressed genes and mutant genes, respectively, as genes affecting APC infiltration and genes related to survival; Select the intersection of mutant genes, abnormally expressed genes, genes related to survival and genes affecting APC infiltration as candidate mRNA vaccine neoantigen targets; Step 5, based on the expression level of the candidate mRNA vaccine neoantigen target, perform consistency clustering analysis on HCC patients, and divide HCC patients into different immune subtypes; compare the overall survival of HCC patients in different immune subtypes to verify the division scheme of immune subtypes; Step 6, compare the tumor immune microenvironment characteristics between immune subtypes, select the immune subtype mainly composed of immunosuppressive immune cells, and take the patients in this immune subtype as the model output suitable for mRNA vaccine treatment patients; Step 7, by multivariate Cox regression analysis and construction of Nomogram model, taking the expression level of the candidate mRNA vaccine neoantigen target of hepatocellular carcinoma patients as the input factor of the Nomogram model, according to the input clinical prognosis data, quantifying the linear relationship between the target and the prognosis of HCC patients, for the model to predict the probability value of the overall survival time reaching 3 years and 5 years according to the expression level of the candidate mRNA vaccine neoantigen target of HCC patients. 2.The method for constructing a liver cancer prognosis and treatment suitability evaluation model based on mRNA vaccine antigens according to claim 1, wherein: The candidate mRNA vaccine neoantigen target is CCNB1, CDC25C, PTTG1, CHEK1, KPNA2, MKI67, KIF2C, MCM3, EZH2, CDT1, PES1, PRC1, PPM1G, CDK1, NEK2, TRIP13, TUBG1, AURKA and G6PD. 3.The method for constructing a hepatocellular carcinoma prognosis and treatment suitability evaluation model based on mRNA vaccine antigens according to claim 1 or 2, wherein: Set the clustering repetition number reps = 100, the sample proportion pItem = 0.8, pFeature = 1, select k-means as the clustering algorithm, use the R package "ConsensusClusterPlus" to perform consistency clustering analysis on HCC patients, and select to divide HCC patients into 3 different immune subtypes by comprehensive analysis of consistency matrix and consistency cumulative distribution function. 4.The method for constructing a liver cancer prognosis and treatment suitability evaluation model based on mRNA vaccine antigens according to claim 3, wherein: By Kaplan-Meier method and log-rank test, compare the overall survival of HCC patients in different immune subtypes to verify the clustering scheme. 5.The method for constructing a liver cancer prognosis and treatment suitability evaluation model based on mRNA vaccine antigens according to claim 1, wherein: The immune microenvironment characteristics include the expression levels of immune checkpoint and immunogenic cell death related genes, tumor mutation burden and infiltration abundance of immune cells. 6.The method for constructing a prognosis and treatment suitability evaluation model of hepatocellular carcinoma based on mRNA vaccine antigens according to claim 1 or 5, wherein: Compare the tumor immune microenvironment characteristics between immune subtypes, including: (1) comparing the expression levels of immune checkpoint and immunogenic cell death related genes of HCC patients in different immune subtypes by differential analysis; (2) calculating the total number of somatic non-synonymous mutations in the coding region as the tumor mutation burden (TMB), including missense, nonsense, splice site and frameshift mutations; (3) based on the transcriptome data, the CIBERSORT algorithm is used to quantify the infiltration abundance of immune cells in the tumor immune microenvironment.

7. The method for constructing a prognostic and therapeutic suitability assessment model for hepatocellular carcinoma based on mRNA vaccine antigens as described in claim 2, characterized in that: Through multivariate Cox regression, the coefficients corresponding to the candidate mRNA vaccine neoantigen targets are fitted as shown in Table 1: The model calculates the probability value of the overall survival time of the HCC patient reaching 3 years and 5 years according to the expression level of the candidate mRNA vaccine neoantigen target gene of the patient and the related coefficients in Table 1. 8.The method for constructing a liver cancer prognosis and treatment suitability evaluation model based on mRNA vaccine antigens according to claim 6, wherein: The probability value of the overall survival time of the patient reaching 3 years and 5 years is the position of the point value corresponding to the input factor in the total point value axis on the value of the corresponding probability axis.

9. A method of predicting based on a mRNA vaccine antigen a prognosis and treatment suitability assessment model for hepatocellular carcinoma, characterized in that: The model is established by the method of any one of claims 1, 2, 5, 7, 8, the input of the model is the expression level of the candidate mRNA vaccine neoantigen target gene of the HCC patient; the model first uses the consistency clustering method to divide the patients into different immune subtypes, and outputs the patients in the immune subtype mainly composed of immunosuppressive immune cells as the patients suitable for mRNA vaccine treatment; then, combining the linear relationship between the expression level of the candidate mRNA vaccine neoantigen target gene and the prognosis, the probability value of the overall survival time of the hepatocellular carcinoma patient reaching 3 years and 5 years is output.

Images