Molecular markers for immunohistochemical typing of pancreatic ductal adenocarcinoma and application of molecular markers

Abstract

The invention discloses a group of molecular markers for pancreatic ductal adenocarcinoma immunohistochemical typing and application thereof. The molecular marker disclosed by the invention comprises HLA-1 (Human Leukocyte Antigen) and CXCL10 (CXCL10). Based on the clinical test result of the first application of an OPL scheme in the advanced PDAC patient and the contained multi-omics data set at present, the immunohistochemical molecule combination for predicting the treatment reaction is screened out, and implementation and verification are carried out in a pancreatic duct adenocarcinoma puncture sample before treatment; the actual treatment response of the patient is highly consistent with the prediction result of immunohistochemistry, which proves that the immunohistochemistry typing has prompting significance for the prognosis of the patient. According to the method, the defects that when disease typing and curative effect prediction are carried out through expression profile sequencing, the technology is complex, the price is high, time is consumed, and the method is difficult to apply clinically are overcome, and a basis is provided for precise treatment of PDAC.

Description

Technical Field

[0001] The invention relates to a group of molecular markers for immunohistochemical typing of pancreatic ductal adenocarcinoma and application thereof, belonging to the technical field of biomedical detection. Background Art

[0002] Pancreatic cancer is a type of malignant tumor originating from pancreatic ductal epithelium and acinar cells. Pancreatic ductal adenocarcinoma (PDAC) is the main type of pancreatic cancer, accounting for more than 90% of all pancreatic cancers. It is a highly malignant and difficult to cure tumor with an increasing incidence and high mortality. Combination chemotherapy is an important first-line treatment option for patients with advanced pancreatic cancer. However, almost all patients with advanced unresectable pancreatic ductal adenocarcinoma who receive combination chemotherapy will inevitably develop drug resistance and progression. For such patients who fail initial treatment, the current number of second-line alternatives is limited and the effect is poor, suggesting an urgent need for individualized treatment for patients with initial treatment. In recent years, with the development of high-throughput detection methods such as gene sequencing and multi-omics research, precision medicine has made great progress in the diagnosis and treatment of pancreatic cancer. At the same time, the progress of targeted therapy and immunotherapy has changed the survival status of patients with advanced cancer. Precision medicine provides decision-making basis for the selection of treatment options for many cancers from multiple aspects such as the exploration of key therapeutic targets, the screening of suitable populations, the sensitivity of therapeutic drugs and dynamic monitoring of treatment. However, in previous studies, PDAC was found to lack molecular site mutations that effectively predict treatment efficacy. Mismatch repair deficiency (dMMR) and microsatellite high instability (MSI-H) can lead to increased tumor mutation load and are important indicators for predicting the efficacy of immune checkpoint inhibitors (ICI). However, in pancreatic cancer, dMMR / MSI-H accounts for less than 1% of the population, and pancreatic cancer is generally resistant to ICI treatment with an extremely low response rate, which makes it difficult for researchers to find reliable therapeutic targets for all PDAC, and clinical trials for the treatment of all PDAC patients have also benefited little.

[0003] Around this issue, the inventors observed the efficacy and immunological changes of PDAC patients after local treatment by combining immune targeted therapy. Multi-omics sequencing of paired samples before and after treatment revealed differences in immunological characteristics and microenvironment gene spectrum expression between responding and non-responding patients. The results showed that patients who responded to the oncolytic virus combined with anti-PD-1 and lenvatinib regimen (OPL treatment regimen) showed a significant upregulation of antigen presentation-related genes at the baseline level and were in a state of obvious immune activation. At this time, CCL4 derived from tumor cells with high expression of HLA-I molecules promoted the recruitment of CXCL10-positive macrophages and monocytes through CCR5. These macrophages and monocytes exhibit an inflammatory phenotype, which contributes to the recruitment and activation of tumor-reactive T cells, not only effectively reshaping the originally poorly infiltrated tumor stroma, but also successfully recruiting effector T cells into the tumor to perform killing functions in the context of vascular normalization.

[0004] However, there is currently no recognized method for predicting and grouping whether patients with advanced PDAC should undergo the OPL regimen (a regimen that uses oncolytic viruses combined with immunotherapy (anti-PD-1) and targeted therapy (lenvatinib)) as the preferred second-line treatment option. Therefore, if a typing system based on immunohistochemistry is constructed to replace the complex and expensive expression profiling sequencing technology, it will help to achieve simple and efficient typing of pancreatic cancer in clinical practice, reveal the individual's genomic characteristics and the corresponding precision treatment strategies, and achieve personalized precision treatment for pancreatic cancer. Summary of the invention

[0005] The purpose of the present invention is to provide a group of molecular markers for immunohistochemical typing of pancreatic ductal adenocarcinoma and their applications, wherein the molecular markers are CXCL10 and HLA-1, and immunohistochemical typing of pancreatic cancer patients is performed based on CXCL10 and HLA-1, and then prognosis is predicted based on the typing results, which helps to achieve precise treatment for patients.

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

[0007] In a first aspect, the present invention provides a group of molecular markers for immunohistochemical typing of pancreatic ductal adenocarcinoma, wherein the molecular markers are CXCL10 and HLA-1 genes or proteins thereof.

[0008] Preferably, the pancreatic ductal adenocarcinoma immunohistochemical typing is performed according to the immunohistochemical scores of CXCL10 and HLA-1, specifically including the following four types:

[0009] Type I: HLA-1 positive and CXCL10 positive;

[0010] Type II: HLA-1 positive, CXCL10 negative;

[0011] Type III: HLA-1 negative, CXCL10 positive;

[0012] Type IV: HLA-1 negative and CXCL10 negative;

[0013] The determination of HLA-I and CXCL10 positivity was achieved through immunohistochemical scoring, which included both the proportion of stained cells (%) and the grading of staining intensity; for the quantification of HLA-1 expression, the cell proportion was the percentage of cells with positive staining, and the intensity level was divided into 0 (negative), 1 (weakly positive), and 2 (strongly positive), where strong and weakly positive refer to whether the staining of the tumor cell membrane is stronger or weaker than that of lymphocytes and endothelial cells, respectively; for the quantification of CXCL10 expression, the staining intensity was divided into 0 (negative), 1 (light yellow), 2 (brown), and 3 (tan), and the cell proportion was the proportion of cells with positive CXCL10 staining in monocytes; the final score was the product of the two variables of the proportion of stained cells and the staining intensity;

[0014] Among them, HLA-1 positivity was defined as a score greater than 1.2, and CXCL10 positivity was defined as a score greater than 0.6.

[0015] Preferably, the molecular marker HLA-1 is screened by the following steps:

[0016] Step 1: Evaluate the efficacy of PDAC patients treated with oncolytic virus combined with anti-PD-1 and lenvatinib. Patients with the best response of complete remission (CR) or partial remission (PR) are classified as responders, and the rest of the patients who did not experience effective tumor regression after treatment are classified as non-responders.

[0017] Step 2: Using the single-cell dataset, the single-cell samples obtained from the biopsy samples of the above-mentioned PDAC patients were clustered, and malignant epithelial cells were defined in the epithelial cell clusters;

[0018] Step 3: Use the MHC-1 antigen presentation core gene set HLA-1, HLA-C, TAP2, NLRC5, TAP1, PSMB9, PSMB8, HLA-B and B2M to establish an MHC score to score tumor cells and define the MHC high group (MHC hi ) and low group (MHC lo );

[0019] Step 4: Compare MHC hi Groups and MHC loThe differentially expressed genes of tumor cells were grouped, a gene set was assembled, and an EPi_MHC feature score was established. The patient's baseline EPi_MHC score was correlated with prognosis using bulk RNA data. This confirmed that high expression of MHC-1 in tumor cells was associated with effective anti-tumor response after treatment, and the HLA-1 molecule was screened out as one of the candidate molecules for predicting efficacy.

[0020] In the step 1, evaluation and definition are performed according to RECIST v1.1, the solid tumor response evaluation criteria.

[0021] Preferably, in the step 2: the molecular markers used for the clustering are obtained by the FindAllMarkers function of the Seurat package in the R language, and the definition of the malignant epithelial cells is implemented by the InferCNV package;

[0022] In the step 3: MHC high score (MHC hi ) group and low score (MHC lo The cutoff value of the group is the median expression of the core gene set, among which the expression level above the median is defined as MHC hi Group, below the median expression level is defined as MHC lo Group.

[0023] In the step 4: the analysis of the differentially expressed genes is implemented using the R language DESeq2 package, the initial screening of the differentially expressed genes is based on Log2|Fold change|>1 and corrected P value<0.05, the EPi_MHC feature gene set is established with the top 25 differentially expressed genes, the bulk RNA gene expression data set is paired and normalized using the z-scored method as the EPi_MHC feature score, the baseline EPi_MHC median score is used as the grouping boundary, the log-rank method is used to analyze and compare the progression-free survival and overall survival of the EPi_MHC high group and the low group, and a P value <0.05 is considered to be statistically different.

[0024] Preferably, the molecular marker CXCL10 is screened by the following steps:

[0025] Step 1: Use the single-cell dataset to group myeloid cells and define two groups of cells, Macro_CXCL10 and InflaMono_CXCL10, in the macrophage and monocyte groups;

[0026] Step 2: The above two groups of cells were compared with other macrophages / monocytes for differentially expressed gene analysis. The top 40 upregulated characteristic gene sets were composed into the MoMac_CXCL10 scoring gene set. The patient's baseline MoMac_CXCL10 score was correlated with the prognosis using bulk RNA data. This confirmed that the high expression of CXCL10 in mononuclear macrophages was associated with an effective anti-tumor response after treatment, and the CXCL10 molecule was screened out as one of the candidate molecules for predicting efficacy.

[0027] Preferably, in the step 1: the molecular markers used for the clustering are obtained by the FindAllMarkers function of the Seurat package in the R language;

[0028] In the step 2: the analysis of the differentially expressed genes is implemented using the R language DESeq2 package, the initial screening of the differentially expressed genes is based on Log2|Fold change|>1 and the corrected P value <0.05, the EPi_MHC feature gene set is established with the top 40 differentially expressed genes, and the MoMac_CXCL10 feature score is calculated using the ssGSEA function of the R language GSVA package; the MoMac_CXCL10 median score of the baseline sample is used as the grouping boundary, and the log-rank method is used to analyze and compare the progression-free survival and overall survival of the MoMac_CXCL10 high group and the low group, and a P value <0.05 is considered to be statistically different.

[0029] In a second aspect, the present invention provides use of the molecular marker or its detection reagent described in the first aspect in preparing a pancreatic ductal adenocarcinoma immunohistochemical typing detection kit.

[0030] In a third aspect, the present invention provides use of the molecular marker or a detection reagent thereof described in the first aspect in preparing a kit for evaluating the prognosis of pancreatic ductal adenocarcinoma.

[0031] Preferably, the prognosis assessment of pancreatic ductal adenocarcinoma includes evaluating the efficacy of an oncolytic virus combined with targeted immunotherapy for patients with pancreatic ductal adenocarcinoma (especially patients with advanced pancreatic ductal adenocarcinoma), such as an OPL treatment regimen: an oncolytic virus combined with immunotherapy (anti-PD-1) and targeted therapy (Lenvatinib).

[0032] Preferably, the immunohistochemical typing that can benefit from the treatment regimen is: HLA-1 positive and CXCL10 positive.

[0033] The present invention screened out CXCL10 and HLA-1 as molecules predicting the efficacy of the OPL regimen in patients with pancreatic ductal adenocarcinoma. Analysis showed that when tumor samples were subjected to immunohistochemical staining at baseline, the positive rates of HLA-1 and CXCL10 molecules were highly correlated with the response rate of the oncolytic virus combined with targeted immune regimen, which can effectively predict whether the patient responds to the treatment regimen, help to effectively screen out PDAC patients who can benefit from the OPL treatment regimen clinically, and achieve precise treatment for patients.

[0034] Therefore, the effective population should be accurately grouped before treatment to achieve the goal of precision medicine.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] Based on the multi-omics data set constructed from samples of advanced PDAC patients who currently receive the first OPL regimen as second-line treatment, the present invention screens out HLA-I and CXCL10 as molecular markers for immunohistochemical typing, constructs an immunohistochemical sorting scheme, and implements and verifies it in existing PDAC tissue samples; the results of the immunohistochemical typing are highly consistent with the clinical results, and the immunohistochemical typing has a suggestive significance for the patient's prognosis, indicating potential treatment opportunities, and providing a decision-making basis for the precise treatment of PDAC patients; the present invention solves the problems of complex operation and high cost of typing by sequencing technology, large batch effect, etc., and the typing results are highly consistent and reproducible, which is suitable for clinical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The present invention is a flowchart of the molecular screening of immunohistochemical typing of advanced progressive pancreatic ductal adenocarcinoma in Example 1.

[0038] Figure 2 Kaplan–Meier curves of progression-free survival (left) and overall survival (right) for the baseline population in Example 1 of the present invention, which were grouped according to EPi_MHC scores (n=26); the data were tested by the log-rank method.

[0039] Figure 3 The Kaplan–Meier curves of progression-free survival (left) and overall survival (right) were made for the baseline population in Example 1 of the present invention based on the MoMac_CXCL10 score (n=26). The data were tested by the log-rank method.

[0040] Figure 4 Schematic diagram of immunohistochemical staining of HLA-I and CXCL10 molecules in Example 1 of the present invention, scale bar: 100 μm (25 μm in the large figure)

[0041] Figure 5 The Kaplan–Meier curves of progression-free survival (left) and overall survival (right) are shown for the baseline population in the embodiment of the present invention based on whether HLA-I and CXCL10 immunohistochemical staining are double positive or not (n=22), and the data are tested by the log-rank method.

[0042] Figure 6 For the comparison of clinical response rates between the immunohistochemical staining double-positive group and other groups in Example 2 of the present invention, the data were tested by Fisher's method. DETAILED DESCRIPTION

[0043] In order to make the present invention more clearly understood, preferred embodiments are described in detail below with reference to the accompanying drawings.

[0044] The experimental methods in the following examples without specifying specific conditions were carried out according to conventional methods and conditions, or selected according to the product instructions; the materials, reagents, etc. used were conventional commercially available products unless otherwise specified.

[0045] Example 1 Screening of molecular markers for immunohistochemical typing of pancreatic ductal adenocarcinoma

[0046] In this example, the genes screened for PDAC immunohistochemical typing are HLA-I and CXCL10. The specific screening and verification process includes the following steps (process as shown in FIG. Figure 1 shown):

[0047] 1) Single-cell mRNA sequencing data of PDAC patients from the Affiliated Cancer Hospital of Fudan University were analyzed, including 14 pairs of paired samples before and after treatment with oncolytic virus combined with anti-PD-1 and lenvatinib and 5 baseline samples. The R language Harmony package was used to filter low-quality cells and eliminate batch effects, and the main cell types and subclusters were identified by classic marker genes.

[0048] 2) According to RECIST1.1 criteria, patients were divided into groups based on efficacy. Among them, patients with the best overall response of CR (complete remission) or PR (partial remission) were defined as the response group (4 patients), and patients with SD (stable disease) or PD (progressive disease) were defined as the non-response group (15 patients).

[0049] 3) The epithelial cell population was defined using the FindAllMarkers function of the Seurat package in R, and the malignant epithelial cells were further defined using the InferCNV package.

[0050] 4) Using the MHC-1 antigen presentation core gene set (HLA-1, HLA-C, TAP2, NLRC5, TAP1, PSMB9, PSMB8, HLA-B, B2M), an MHC score was established to score tumor cells. The boundary between the MHC high score (MHC-hi) group and the low score (MHC-lo) group was the median expression of the core genes in the baseline sample. Among them, those above the median were defined as the MHC-hi group, and those below the median were defined as the MHC-lo group.

[0051] 5) Compare the differentially expressed genes of tumor cells in the MHC-hi group and the MHC-lo group, and construct a gene set to establish the EPi_MHC feature score, which is implemented using the R language DESeq2 package. The differential gene initial screening is based on Log2|Fold change|>1 and the corrected P value <0.05, and the EPi_MHC feature gene set is established with the top 25 differential genes. Pairs of differential gene expression are taken and standardized by the z-scored method as the EPi_MHC feature score. The results are shown in Table 1.

[0052] Table 1. EPi_MHC signature gene set

[0053]

[0054]

[0055] 6) Using the median EPi_MHC score as the grouping boundary, the baseline EPi_MHC score of the patient was associated with prognosis, and the log-rank method was used to compare the progression-free survival and overall survival of the high- and low-EPi_MHC groups. A P value < 0.05 was considered statistically significant. The significant extension of progression-free survival and overall survival confirmed that the high expression of MHC-1 genes in tumor cells is associated with the successful production of effective anti-tumor response after treatment ( Figure 2 ), screened out human HLA-1 molecule as one of the candidate molecules for predicting therapeutic efficacy.

[0056] 7) The myeloid cells defined in step 1) are grouped again, and two groups of cells, Macro_CXCL10 and InflaMono_CXCL10, are defined in the macrophage and monocyte groups.

[0057] 8) Comparison of differentially expressed genes between Macro_CXCL10 and InflaMono_CXCL10 and other macrophage / monocyte populations was performed using the DESeq2 package in R. The top 40 upregulated signature genes were synthesized into the MoMac_CXCL10 scoring gene set, and the results are shown in Table 2.

[0058] Table 2. EPi_MHC signature gene set

[0059]

[0060]

[0061] 9) The MoMac_CXCL10 feature score was calculated using the ssGSEA function of the GSVA package in R. The median MoMac_CXCL10 score of the baseline sample was used as the grouping cutoff, and samples above and below this cutoff were defined as the MoMac_CXCL10 high group and low group, respectively.

[0062] 10) The patient's baseline MoMac_CXCL10 score was associated with prognosis. The log-rank method was used to compare the progression-free survival and overall survival of the MoMac_CXCL10 high- and low-groups. P values ​​< 0.05 were considered statistically significant. The results showed that high expression of CXCL10 in mononuclear macrophages was associated with effective anti-tumor response after treatment ( Figure 3 ), and screened out CXCL10 as one of the candidate molecules for predicting therapeutic efficacy.

[0063] 11) 27 baseline paraffin-embedded tumor tissues were selected for immunohistochemical testing and evaluation of HLA-1 and CXCL10 protein expression. For HLA, a marker that stains on tumor cells, the percentage of positive tumor cells in the total tumor cells multiplied by the staining intensity was used as the immunohistochemical score; for CXCL10, a secretory marker, the percentage of positive monocytes in the total monocytes multiplied by the staining intensity was used as the immunohistochemical score ( Figure 4 ).

[0064] 12) The patients were divided into three categories: low, medium and high expression, with 40% and 60% of the immunohistochemical scores of the two markers respectively as the boundaries of low expression and high expression. Patients with high expression of both HLA-1 and CXCL10 were defined as the double positive group. After removing patients with medium expression intensity of any marker from the whole, the remaining single positive and double negative expression patients were defined as other groups. The survival differences between the double positive group and other groups were further compared. The results are shown in Table 3. The log-rank method was used to analyze and compare the progression-free survival and overall survival of the two groups. A P value <0.05 was considered to be statistically different. The results showed that immunohistochemical staining of HLA-1 and CXCL10 on the patient's tumor samples before treatment can quickly and efficiently group patients with advanced progressive pancreatic ductal adenocarcinoma according to the positive expression of marker molecules, screen out patients who can effectively respond to the OPL regimen, and significantly improve the prognosis of such patients ( Figure 5 ).

[0065] Table 3. HLA-1 and CXCL10 immunohistochemical scores and groups

[0066]

[0067]

[0068] II. Reagents and equipment required for immunohistochemical typing

[0069] 1. Antibodies for immunohistochemistry are shown in Table 4.

[0070] Table 4. Immunohistochemistry antibodies

[0071] molecular Antibody code company Dilution multiple HLA-1 Cat#ab225636 Abcam 1:5000 CXCL10 Cat#10937-1-AP Proteintech 1:800

[0072] 2. Immunohistochemistry experimental platform: Ventana ULTRA detection system.

[0073] Example 2 Immunohistochemical typing verification

[0074] The immunohistochemical typing genes screened in Example 1 were validated in a cohort of 27 patients with advanced pancreatic ductal adenocarcinoma from the Affiliated Cancer Hospital of Fudan University, verifying the consistency between the immunohistochemical typing results and the efficacy of the OPL regimen and the clinical significance of the immunohistochemical typing.

[0075] 1) Baseline tumor tissue samples from 27 PDAC patients before treatment were obtained.

[0076] 2) Immunohistochemical staining of HLA-I and CXCL19 was performed on paraffin-embedded sections using the Ventana ULTRA immunohistochemical detection system of the Department of Pathology, Affiliated Cancer Hospital of Fudan University. The antibodies used are shown in Table 4.

[0077] 3) Organize pathology experts from Huashan Hospital Affiliated to Fudan University to read the immunohistochemical staining results and evaluate the immunohistochemical staining scores of each indicator of each sample. The determination of HLA-I and CXCL10 positivity is achieved by immunohistochemical scoring. The scoring system includes both the proportion of stained cells (%) and the grading of staining intensity. For HLA-1, the cell proportion is the percentage of cells with positive staining, and the intensity level is divided into 0 (negative), 1 (weak positive), and 2 (strong positive), where strong / weak positive refers to whether the staining of the tumor cell membrane is stronger / weaker than that of lymphocytes and endothelial cells. For the quantification of CXCL10 expression, the staining intensity is divided into 0 (negative), 1 (light yellow), 2 (brown), and 3 (tan), and the cell proportion is the proportion of cells with positive CXCL10 staining in myeloid cells. The final score is the product of the two variables of proportion and intensity. Among them, HLA-1 positivity is defined as a score greater than 1.2, and CXCL10 positivity is defined as a score greater than 0.6.

[0078] 4) According to the above-mentioned immunohistochemical scoring method, 27 samples in this example were subjected to immunohistochemical typing and screening, and the specific typing results are as follows: HLA-I and CXCL10 double positive group: 5 people. Others: 17 people.

[0079] 5) To investigate the prognostic significance of immunohistochemical typing: the typing was associated with progression-free survival, overall survival, and efficacy evaluation (response and non-response). The results of survival analysis showed that there was a significant difference in progression-free survival between the two groups (log-rank; P = 0.02; Figure 5 ), and the response of the two groups to the OPL regimen was significantly different (Fisher test; P = 0.03; Figure 6 The progression-free survival and objective response rate of the double-positive group were significantly better than those of other groups, suggesting that HLA-I and CXCL10 are effective molecular markers for predicting efficacy and distinguishing beneficiary populations.

[0080] The above description is only a preferred embodiment of the present invention and is not any formal or substantial limitation of the present invention. It should be pointed out that a person skilled in the art can make several improvements and supplements without departing from the present invention, and these improvements and supplements should also be regarded as the protection scope of the present invention.

Claims

1. A group of molecular markers for immunohistochemical typing of pancreatic ductal adenocarcinoma, characterized in that: The molecular markers are CXCL10 and HLA-1 genes or their proteins.

2. The molecular marker according to claim 1, characterized in that The immunohistochemical typing of pancreatic ductal adenocarcinoma is performed according to the immunohistochemical scores of CXCL10 and HLA-1, and specifically includes the following four types: Type I: HLA-1 positive and CXCL10 positive; Type II: HLA-1 positive, CXCL10 negative; Type III: HLA-1 negative, CXCL10 positive; Type IV: HLA-1 negative and CXCL10 negative; The determination of HLA-I and CXCL10 positivity was achieved through immunohistochemical scoring, which included both the proportion of stained cells (%) and the grading of staining intensity; for the quantification of HLA-1 expression, the cell proportion was the percentage of cells with positive staining, and the intensity level was divided into 0 (negative), 1 (weakly positive), and 2 (strongly positive), where strong and weakly positive refer to whether the staining of the tumor cell membrane is stronger or weaker than that of lymphocytes and endothelial cells, respectively; for the quantification of CXCL10 expression, the staining intensity was divided into 0 (negative), 1 (light yellow), 2 (brown), and 3 (tan), and the cell proportion was the proportion of cells with positive CXCL10 staining in monocytes; the final score was the product of the two variables of the proportion of stained cells and the staining intensity; Among them, HLA-1 positivity was defined as a score greater than 1.2, and CXCL10 positivity was defined as a score greater than 0.

6.

3. The molecular marker according to claim 1, characterized in that The molecular marker HLA-1 is screened by the following steps: Step 1: Evaluate the efficacy of PDAC patients treated with oncolytic virus combined with anti-PD-1 and lenvatinib. Patients with the best response of complete remission (CR) or partial remission (PR) are classified as responders, and the rest of the patients who did not experience effective tumor regression after treatment are classified as non-responders. Step 2: Using the single-cell dataset, the single-cell samples obtained from the biopsy samples of the above-mentioned PDAC patients were clustered, and malignant epithelial cells were defined in the epithelial cell clusters; Step 3: Use the MHC-1 antigen presentation core gene set HLA-1, HLA-C, TAP2, NLRC5, TAP1, PSMB9, PSMB8, HLA-B and B2M to establish an MHC score to score tumor cells and define the MHC high group (MHC hi ) and low group (MHC lo ); Step 4: Compare MHC hi Groups and MHC lo The differentially expressed genes of tumor cells were grouped, a gene set was assembled, and an EPi_MHC feature score was established. The patient's baseline EPi_MHC score was correlated with prognosis using bulk RNA data. This confirmed that high expression of MHC-1 in tumor cells was associated with effective anti-tumor response after treatment, and the HLA-1 molecule was screened out as one of the candidate molecules for predicting efficacy.

4. The molecular marker according to claim 3, characterized in that In the step 2: the molecular markers used for the clustering are obtained by the FindAllMarkers function of the Seurat package in the R language, and the definition of the malignant epithelial cells is implemented by the InferCNV package; In the step 3: MHC high score (MHC hi ) group and low score (MHC lo The cutoff value of the group is the median expression of the core gene set, among which the expression level above the median is defined as MHC hi Group, below the median expression level is defined as MHC lo Group. In the step 4: the analysis of the differentially expressed genes is implemented using the R language DESeq2 package, the initial screening of the differentially expressed genes is based on Log2|Fold change|>1 and corrected P value<0.05, the EPi_MHC feature gene set is established with the top 25 differentially expressed genes, the bulk RNA gene expression data set is paired and normalized using the z-scored method as the EPi_MHC feature score, the baseline EPi_MHC median score is used as the grouping boundary, the log-rank method is used to analyze and compare the progression-free survival and overall survival of the EPi_MHC high group and the low group, and a P value <0.05 is considered to be statistically different.

5. The molecular marker according to claim 1, characterized in that The molecular marker CXCL10 was screened by the following steps: Step 1: Use the single-cell dataset to group myeloid cells and define two groups of cells, Macro_CXCL10 and InflaMono_CXCL10, in the macrophage and monocyte groups; Step 2: The above two groups of cells were compared with other macrophages / monocytes for differentially expressed gene analysis. The top 40 upregulated characteristic gene sets were composed into the MoMac_CXCL10 scoring gene set. The patient's baseline MoMac_CXCL10 score was correlated with the prognosis using bulk RNA data. This confirmed that the high expression of CXCL10 in mononuclear macrophages was associated with an effective anti-tumor response after treatment, and the CXCL10 molecule was screened out as one of the candidate molecules for predicting efficacy.

6. The molecular marker according to claim 5, characterized in that In the step 1: the molecular markers used for the clustering are obtained by using the FindAllMarkers function of the Seurat package in the R language; In the step 2: the analysis of the differentially expressed genes is implemented using the R language DESeq2 package, the initial screening of the differentially expressed genes is based on Log2|Fold change|>1 and the corrected P value <0.05, the EPi_MHC feature gene set is established with the top 40 differentially expressed genes, and the MoMac_CXCL10 feature score is calculated using the ssGSEA function of the R language GSVA package; the MoMac_CXCL10 median score of the baseline sample is used as the grouping boundary, and the log-rank method is used to analyze and compare the progression-free survival and overall survival of the MoMac_CXCL10 high group and the low group, and a P value <0.05 is considered to be statistically different.

7. Use of the molecular marker or the detection reagent thereof according to any one of claims 1 to 6 in the preparation of a pancreatic ductal adenocarcinoma immunohistochemical typing detection kit.

8. Use of the molecular marker or the detection reagent thereof according to any one of claims 1 to 6 in the preparation of a kit for evaluating the prognosis of pancreatic ductal adenocarcinoma.

9. The use according to claim 8, characterized in that: The prognosis assessment of pancreatic ductal adenocarcinoma includes evaluating the efficacy of an oncolytic virus combined with targeted immune therapy for patients with pancreatic ductal adenocarcinoma.

10. The use according to claim 8, characterized in that: The immunohistochemical typing that can benefit from the treatment regimen is: HLA-1 positive and CXCL10 positive.

Images