Marker combination for diffuse large B-cell lymphoma immune microenvironment typing prediction based on pathological tissue staining and application of marker combination in immune microenvironment typing prediction model

Abstract

The invention relates to a marker combination for immune microenvironment typing prediction of diffuse large B-cell lymphoma based on pathological tissue staining and application of the marker combination in an immune microenvironment typing prediction model. The marker combination is composed of the following markers: CD3, CD4, CD34, cologen, FAP, CD8, CD68 and PD-L1. Compared with the prior art, the diffuse large B-cell lymphoma immune microenvironment subtypes can be obtained more quickly and with higher quality by utilizing the advantages of pathological tissue staining compared with traditional RNA sequencing.

Description

Technical Field

[0001] The present invention relates to the field of tumor immunotechnology, and in particular to a marker combination for predicting immune microenvironment typing of diffuse large B-cell lymphoma based on pathological tissue staining and its application in an immune microenvironment typing prediction model. Background Art

[0002] Diffuse Large B-cell Lymphoma (DLBCL) is a group of malignant and aggressive tumors originating from mature B cells. It is the most common non-Hodgkin's lymphoma, accounting for 30%-40% of all non-Hodgkin's lymphoma (NHL) cases. DLBCL has a large heterogeneity. 60% to 65% of patients can be cured by the first-line R-CHOP regimen, but 10% to 15% of patients still have primary drug resistance, and 30% of patients relapse after complete remission. Some relapsed patients can achieve complete remission again through second-line chemotherapy or autologous transplantation, but this part of patients only accounts for about 15%. For patients with relapse, refractory disease and relapse after transplantation, the total response rate to retreatment can only reach 20%, and the complete remission rate is only 7%. Therefore, for most relapsed and refractory patients, second-line chemotherapy is difficult to cure the tumor.

[0003] In the new era, lymphoma treatment first requires screening out high-risk patients and then conducting stratified treatment to reduce the occurrence of treatment failure. Currently, immunohistochemistry-based methods (such as the Hans method) are widely used in clinic to classify cases into GCB and non-GCB (non-GCB includes ABC subtypes and most unclassifiable cases), which only provide approximate results of gene expression profile analysis and have the risk of misclassification.

[0004] The biology and clinical manifestations of DLBCL are derived not only from molecular changes in DLBCL cells, but also from their interaction with the microenvironment. Data from lymphoma patients and animal models indicate that external stimuli provided by microenvironmental cells and extracellular matrix (ECM) contribute to the occurrence, development, and response to treatment of the disease. Existing articles have determined the characteristics of the DLBCL microenvironment through immune microenvironment cells, and divided DLBCL into four different subtypes: germinal center-like (GC), mesenchymal (MS), inflammatory (IN), and exhausted (DP), reflecting unique biological characteristics and clinical manifestations, and providing a series of new potential targets for innovative therapeutic interventions.

[0005] However, the application of this classification method to routine clinical practice has been hindered by the stringent quality requirements of the tissue for RNA sequencing, long processing time, and high testing costs. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a marker combination for predicting the immune microenvironment typing of diffuse large B-cell lymphoma based on pathological tissue staining and its application in the immune microenvironment typing prediction model.

[0007] The purpose of the present invention can be achieved by the following technical solutions:

[0008] One of the technical solutions of the present invention is to provide a marker combination for predicting the immune microenvironment typing of diffuse large B-cell lymphoma based on pathological tissue staining. The immune microenvironment typing of diffuse large B-cell lymphoma is GC, MS, IN and DP types, which are composed of the following markers: CD3, collagen, CD8, CD68 and PD-L1.

[0009] In some embodiments, the marker combination consists of the following markers: CD3, CD4, CD34, collagen, FAP, CD8, CD68 and PD-L1.

[0010] Among them, CD3 and CD4 are used as staining markers, which can be T cells and CD4+T cells;

[0011] Using CD34, collagen, and FAP as staining markers, it can be tumor-associated fibroblasts and endothelial cells;

[0012] Using CD8, CD68 and PD-L1 as staining markers, it can be CD8+T cells and macrophages.

[0013] A second technical solution of the present invention is to provide an application of the marker combination described in one of the above technical solutions in constructing a prediction model for the immune microenvironment typing of diffuse large B-cell lymphoma.

[0014] In some specific embodiments, the prediction criteria of the diffuse large B-cell lymphoma immune microenvironment typing prediction model are:

[0015] (1) First, determine the size of the collagen staining quantitative data and 5.7: if the collagen staining quantitative data is less than 5.7, then continue to determine the size of the CD68 staining quantitative data and 13; if the collagen staining quantitative data is greater than or equal to 5.7, then continue to determine the size of the CD3 staining quantitative data and 39;

[0016] (2) When the quantitative data of collagen staining is less than 5.7, the size of the quantitative data of CD68 staining and 13 is determined: if the quantitative data of CD68 staining is less than 13, the size of the quantitative data of CD3 staining and 32 is determined; if the quantitative data of CD68 staining is ≥13, it is IN type;

[0017] (3) When the quantitative data of collagen staining is less than 5.7 and the quantitative data of CD68 staining is less than 13, continue to judge the size of the quantitative data of CD3 staining and 32; if the quantitative data of CD3 staining is less than 32, continue to judge the size of the quantitative data of CD8 staining and 15; if the quantitative data of CD3 staining is ≥32, it is GC type;

[0018] (4) When the quantitative data of collagen staining is less than 5.7, the quantitative data of CD68 staining is less than 13, and the quantitative data of CD3 staining is less than 32, continue to judge the size of the quantitative data of CD8 staining and 15: if the quantitative data of CD8 staining is less than 15, it is DP type; if the quantitative data of CD8 staining is ≥15, it is IN type;

[0019] (5) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining and 39 are continued to be judged; if the quantitative data of CD3 staining is ≥39, the quantitative data of CD68 staining and 13 are continued to be judged; if the quantitative data of CD3 staining is <39, the quantitative data of collagen staining and 7.2 are continued to be judged;

[0020] (6) When the quantitative data of collagen staining is ≥5.7 and the quantitative data of CD3 staining is ≥39, continue to judge the size of the quantitative data of CD68 staining and 13: if the quantitative data of CD68 staining is <13, it is GC type; if the quantitative data of CD68 staining is ≥13, it is IN type;

[0021] (7) When the quantitative data of collagen staining is ≥5.7 and the quantitative data of CD3 staining is <39, continue to judge the size of the quantitative data of collagen staining and 7.2; if the quantitative data of collagen staining is <7.2, continue to judge the size of the quantitative data of CD68 staining and 19; if the quantitative data of collagen staining is ≥7.2, continue to judge the size of the quantitative data of CD8 staining and 29;

[0022] (8) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, and the quantitative data of collagen staining is <7.2, continue to judge the size of the quantitative data of CD68 staining and 19: if the quantitative data of CD68 staining is <19, continue to judge the size of the quantitative data of PD-L1 staining and 57; if the quantitative data of CD68 staining is ≥19, it is IN type;

[0023] (9) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, the quantitative data of collagen staining is <7.2, and the quantitative data of CD68 staining is <19, continue to judge the size of the quantitative data of PD-L1 staining and 57: if the quantitative data of PD-L1 staining is ≥57, it is IN type; if the quantitative data of PD-L1 staining is <57, continue to judge the size of the quantitative data of PD-L1 staining and 15;

[0024] (10) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, the quantitative data of collagen staining is <7.2, the quantitative data of CD68 staining is <19, and the quantitative data of PD-L1 staining is <57, continue to judge the size of the quantitative data of PD-L1 staining and 15: if the quantitative data of PD-L1 staining is <15, it is DP type; if the quantitative data of PD-L1 staining is ≥15, it is MS type;

[0025] (11) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, and the quantitative data of collagen staining is ≥7.2, continue to judge the size of the quantitative data of CD8 staining and 29: if the quantitative data of CD8 staining is ≥29, it is IN type; if the quantitative data of CD8 staining is <29, it is MS type.

[0026] In some specific embodiments, the method for constructing a prediction model for immune microenvironment typing of diffuse large B-cell lymphoma comprises the following steps:

[0027] S1. Diffuse large B-cell lymphoma tissue samples were subjected to immunohistochemistry and Mallory's trichrome staining;

[0028] S2, scanning and reading the slides to obtain quantitative data of marker combination staining;

[0029] S3. Based on the quantitative data of marker combination staining, a decision tree was constructed using the rpart() function of the R software “RPART” package.

[0030] In some specific embodiments, in step S1, the immunohistochemistry process includes: dewaxing; hydration; high-pressure heat repair; quenching endogenous peroxidase; blocking; primary antibody incubation; secondary antibody incubation; counterstaining; dehydration; transparency; and sealing.

[0031] In some specific embodiments, in step S1, the Mallory's trichrome staining process includes: dewaxing; hydration; liquid immersion; liquid immersion; differentiation; dehydration; transparency; and sealing.

[0032] In some specific embodiments, in step S2, the scanning and reading of the film include: marking positive cells, quantifying the stained area, and obtaining quantitative data of marker combination staining.

[0033] In some specific embodiments, the positive cells are characterized as follows: the positive cell nuclei are blue and the positive parts are brown by immunohistochemical staining; the positive cell nuclei are red and the fibers are dark blue by Mallory's trichrome staining.

[0034] Exemplarily, positive cells were marked and staining areas were quantified using QuPath and Aperio eSlide Manager software, and quantification was performed.

[0035] In some specific embodiments, in step S3, the staining quantitative data samples obtained after scanning and reading are randomly divided into a training set and a validation set, and the rpart function of the R software "RPART" package is used to recursively divide the training set and the validation set, respectively, to construct a decision tree, and obtain a diffuse large B-cell lymphoma immune microenvironment typing prediction model.

[0036] Exemplarily, the set.seed() function of R software was used to randomly divide the samples into a training set and a validation set in a ratio of 6:4, and the rpart function of the "RPART" package of R software was used to recursively divide the training set and the validation set, respectively. The CP value was selected as 0.015504, and a decision tree was constructed to obtain a diffuse large B-cell lymphoma immune microenvironment typing prediction model.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The present invention adopts a DLBCL microenvironment classification method based on pathological tissue staining and digital pathology, simplifies high-throughput transcriptome classification, and improves the objectivity and repeatability of the analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Schematic diagram of the relationship between marker combination and immune microenvironment typing of diffuse large B-cell lymphoma.

[0040] Figure 2 LME typing obtained by RNA sequencing, PFS and OS under RCHOP regimen.

[0041] Figure 3 The LME classification obtained by the method of the present invention, PFS and OS under RCHOP regimen treatment. DETAILED DESCRIPTION

[0042] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0043] In the following examples, unless otherwise specified, raw materials or processing techniques are conventional commercially available raw materials or conventional processing techniques in the art.

[0044] Embodiment 1:

[0045] 1. Specimen collection

[0046] In this experiment, diffuse large B-cell lymphoma tissues were collected from 315 patients with diffuse large B-cell lymphoma (DLBCL).

[0047] 2. Dyeing

[0048] 2.1 The specific steps of immunohistochemistry are as follows:

[0049] (1) Place the slices in xylene and soak them in xylene I, xylene II, and xylene III for 10 min each (divide the same xylene into three bottles and soak them in sequence).

[0050] (2) The sections were sequentially placed in anhydrous ethanol, 90% ethanol, 70% ethanol, and 50% ethanol, each soaked once for 5 min each time; then washed and soaked in TBS three times, each time for 3 min.

[0051] (3) High-pressure thermal repair: Preheat the repair solution, put in the slices, place in a pressure cooker and heat until full pressure is reached. Continue heating for 1.5 minutes, turn off the power, and cool to room temperature.

[0052] Repair fluid includes: ①TRIS-EDTA repair fluid: merchant biosharp, item number BL617A, for CD8, CD68, PD-L1, CD4, CD34, FAP

[0053] ② Sodium citrate antigen retrieval solution: biosharp, product number BL619A, for CD3

[0054] (4) Inactivation of endogenous peroxidase: Soak in 3% hydrogen peroxide solution for 15 min.

[0055] (5) Blocking: Block with 10% goat serum + TBS (same species as the secondary antibody) at room temperature for 30 min. Do not wash after blocking and discard the blocking solution.

[0056] (6) Primary antibody incubation: Prepare antibody working solution with 10% goat serum and incubate overnight at 4°C. Wash with TBST 3 times, 3 min each time.

[0057] (7) Secondary antibody incubation: Incubate with mouse and rabbit mixed secondary antibodies for 1 h at room temperature. Wash three times with TBST, 3 min each time.

[0058] (8) Color development: Prepare DAB color development solution and use it immediately. Observe the coffee-colored background and specific positive signal under a microscope and quickly place it in tap water to stop color development.

[0059] (9) Restaining: Soak the sections in hematoxylin and stain for 1 min at room temperature. Soak in TBS and wash 5 times. Differentiate in 1% hydrochloric acid ethanol for 5 seconds, soak in TBS to reverse blue. Observe the nuclear staining under a microscope.

[0060] (10) Dehydration: Place in 70% ethanol, 95% ethanol, 100% ethanol II, and I for 5 min each (the same ethanol is divided into two bottles and soaked in sequence).

[0061] (11) Transparency: Soak the sample in xylene I and II for 5 min each (divide the same xylene into two bottles and soak them in sequence).

[0062] (12) Sealing: Neutral resin sealing.

[0063] Among them, germinal center-like (GC type) is mainly composed of T cells and CD4+T cells, and CD3 and CD4 can be used as staining markers;

[0064] Mesenchyme (MS type) is mainly composed of tumor-associated fibroblasts and endothelial cells, and CD34, collagen, and FAP can be used as staining markers;

[0065] Inflammatory (IN type) cells are mainly CD8+T cells and macrophages, and CD8, CD68 and PD-L1 can be used as staining markers.

[0066] 2.2 The specific steps of Mallory's trichrome staining (using Shanghai Hongqiao Lexiang kit) are as follows:

[0067] (1) Place the slices in xylene and soak them in xylene I, xylene II, and xylene III for 10 min each (divide the same xylene into three bottles and soak them in sequence).

[0068] (2) The sections were sequentially placed in anhydrous ethanol, 90% ethanol, 70% ethanol, and 50% ethanol, each soaked once for 5 min each time; then washed and soaked in distilled water for 3 times, each time for 3 min.

[0069] (3) Soak in solution 1 overnight.

[0070] (4) Rinse with tap water for 2 minutes and soak once in distilled water.

[0071] (5) Soak in solution 2 for 20 min. Do not rinse with water, and then soak in solution 3 for 20 min.

[0072] (6) Without washing with water, differentiate directly with 95% ethanol for 5-10 seconds, blow dry with a hair dryer, and observe under a microscope.

[0073] (7) Dehydrate with anhydrous ethanol for 2 min.

[0074] (8) Soak in xylene I and II for 5 minutes each until transparent (divide the same xylene into two or three bottles and soak in sequence).

[0075] (9) Neutral resin sealing.

[0076] 3. Scanning and reading

[0077] 3.1 Scan

[0078] An Aperio GT 450 scanner was used to acquire digital images of whole-slide stained DLBCL tissue sections.

[0079] 3.2 Film reading

[0080] 3.2.1 Confirm the image scanning quality

[0081] Confirm that the focus is clear. Samples that are blurred and cannot be analyzed normally need to be rescanned. Confirm that the scan is complete. Samples that fail to include all tissues in the scan range need to be rescanned.

[0082] 3.2.2 Use QuPath and Aperio eSlide Manager software to mark positive cells and quantify the stained area.

[0083] The analysis area is circled, and the rules are as follows:

[0084] (1) Areas with little or no blue or red nuclear staining: Since the cell nuclei are stained with hematoxylin in immunohistochemistry, they appear blue; the cell nuclei appear red after Mallory's trichrome staining with two-component acid fuchsin solution.

[0085] No blue or red nuclear staining, all should be discarded. A small amount of / sporadic blue or red nuclear staining, if it is blood or muscle tissue, all should be discarded;

[0086] (2) elimination of non-cellular components with blue and red nuclear staining;

[0087] (3) Folded area: The folded area caused by tissue detachment in the stained scan image was removed;

[0088] (4) Eliminate scattered areas outside the overall sample;

[0089] (5) Removal of normal epithelial tissue.

[0090] Marker positive cells:

[0091] Immunohistochemistry DAB and hematoxylin staining showed blue nuclei and brown positive sites. Different colors represented different areas and positively stained cells were identified by qupath software. Mallory staining showed red nuclei and dark blue fibers. Different color areas were identified by Aperio eSlide Manager software.

[0092] Quantify the stained area:

[0093] Both softwares automatically quantify cell nuclei, positive cells and areas by color.

[0094] The two softwares quantify the positive cells or positive areas of each tissue by identifying different colors in the stained tumor tissue.

[0095] 4. The present invention uses the rpart function of the “RPART” R software package (https: / / www.rdocumentation.org / packages / rpart / versions / 4.1.23) to recursively partition the quantitative data set and construct a classification model through a decision tree.

[0096] A total of 315 DLBCL samples with retained tumor tissue were randomly divided into training set: validation set ratio of 6:4.

[0097] Specifically:

[0098] The quantitative data of CD3, CD8, CD68, PD-L1 and Mallory staining of 315 patients were collected and sorted. The patients were randomly divided into training set and validation set at a ratio of 6:4 using the set.seed() function of R software. The training set was recursively divided using the rpart() function of the "RPART" package of R software. The CP (complexity parameter) value = 0.015504 was selected to construct a decision tree, i.e., a classification prediction model for the immune microenvironment of diffuse large B-cell lymphoma. The same method was used to construct a decision tree for the validation set. Figure 1 As shown in the figure, specific typing can be performed according to the cutoff value of the corresponding marker on the decision tree, and the prediction criteria are:

[0099] (1) First, determine the size of the collagen staining quantitative data and 5.7: if the collagen staining quantitative data is less than 5.7, then continue to determine the size of the CD68 staining quantitative data and 13; if the collagen staining quantitative data is greater than or equal to 5.7, then continue to determine the size of the CD3 staining quantitative data and 39;

[0100] (2) When the quantitative data of collagen staining is less than 5.7, the size of the quantitative data of CD68 staining and 13 is determined: if the quantitative data of CD68 staining is less than 13, the size of the quantitative data of CD3 staining and 32 is determined; if the quantitative data of CD68 staining is ≥13, it is IN type;

[0101] (3) When the quantitative data of collagen staining is less than 5.7 and the quantitative data of CD68 staining is less than 13, continue to judge the size of the quantitative data of CD3 staining and 32; if the quantitative data of CD3 staining is less than 32, continue to judge the size of the quantitative data of CD8 staining and 15; if the quantitative data of CD3 staining is ≥32, it is GC type;

[0102] (4) When the quantitative data of collagen staining is less than 5.7, the quantitative data of CD68 staining is less than 13, and the quantitative data of CD3 staining is less than 32, continue to judge the size of the quantitative data of CD8 staining and 15: if the quantitative data of CD8 staining is less than 15, it is DP type; if the quantitative data of CD8 staining is ≥15, it is IN type;

[0103] (5) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining and 39 are continued to be judged; if the quantitative data of CD3 staining is ≥39, the quantitative data of CD68 staining and 13 are continued to be judged; if the quantitative data of CD3 staining is <39, the quantitative data of collagen staining and 7.2 are continued to be judged;

[0104] (6) When the quantitative data of collagen staining is ≥5.7 and the quantitative data of CD3 staining is ≥39, continue to judge the size of the quantitative data of CD68 staining and 13: if the quantitative data of CD68 staining is <13, it is GC type; if the quantitative data of CD68 staining is ≥13, it is IN type;

[0105] (7) When the quantitative data of collagen staining is ≥5.7 and the quantitative data of CD3 staining is <39, continue to judge the size of the quantitative data of collagen staining and 7.2; if the quantitative data of collagen staining is <7.2, continue to judge the size of the quantitative data of CD68 staining and 19; if the quantitative data of collagen staining is ≥7.2, continue to judge the size of the quantitative data of CD8 staining and 29;

[0106] (8) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, and the quantitative data of collagen staining is <7.2, continue to judge the size of the quantitative data of CD68 staining and 19: if the quantitative data of CD68 staining is <19, continue to judge the size of the quantitative data of PD-L1 staining and 57; if the quantitative data of CD68 staining is ≥19, it is IN type;

[0107] (9) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, the quantitative data of collagen staining is <7.2, and the quantitative data of CD68 staining is <19, continue to judge the size of the quantitative data of PD-L1 staining and 57: if the quantitative data of PD-L1 staining is ≥57, it is IN type; if the quantitative data of PD-L1 staining is <57, continue to judge the size of the quantitative data of PD-L1 staining and 15;

[0108] (10) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, the quantitative data of collagen staining is <7.2, the quantitative data of CD68 staining is <19, and the quantitative data of PD-L1 staining is <57, continue to judge the size of the quantitative data of PD-L1 staining and 15: if the quantitative data of PD-L1 staining is <15, it is DP type; if the quantitative data of PD-L1 staining is ≥15, it is MS type;

[0109] (11) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, and the quantitative data of collagen staining is ≥7.2, continue to judge the size of the quantitative data of CD8 staining and 29: if the quantitative data of CD8 staining is ≥29, it is IN type; if the quantitative data of CD8 staining is <29, it is MS type.

[0110] Table 1 shows the classification accuracy in the training set and the validation set

[0111] Table 1

[0112] Model Training Error Test Error RPART model 16.3％ 18.4％

[0113] The previous article (Kotlov, N., A. Bagaev, MV Revuelta, JM Phillip, MT Cacciapuoti, Z. Antysheva, V. Svekolkin, E. Tikhonova, N. Miheecheva, N. Kuzkina, et al., (2021). Clinical and Biological Subtypes of B-cell Lymphoma Revealed by Microenvironmental Signatures. Cancer Discov 11, 1468-1489.10.1158 / 2159-8290. CD-20-0839.) analyzed the immune microenvironment cells of DLBCL based on RNA sequencing results, and divided DLBCL into GC, MS, IN and DP types according to the characteristics of the immune microenvironment.

[0114] Applying the method in this article, typing based on RNA sequencing was also performed, and the consistency rate between the training set and the validation set in the immune microenvironment classification prediction model based on pathological tissue staining of the present invention and the typing obtained by RNA sequencing was compared, and the accuracy rate was greater than 80%.

[0115] Embodiment 2:

[0116] The LME typing of 181 DLBCL samples obtained by RNA sequencing was further compared. Figure 2 ) and tissue staining LME typing ( Figure 3 )PFS (Progression-Free-Survival, representing the time from the start of treatment to disease progression or death of cancer patients) and OS (Overall Survival) under R-CHOP regimen treatment.

[0117] The longer the PFS time, the longer the patient's survival period, and the shorter the time, the shorter the survival period.

[0118] Generally speaking, the longer the OS time, the better the efficacy of the anti-tumor drug.

[0119] The results showed that the p values ​​of PFS and OS among the four groups were all less than 0.05, indicating that there was no significant difference in the prognosis between the two classification methods. This shows that the LME classification algorithm based on pathological tissue staining of the present invention can effectively replicate the LME classification method based on transcriptomes, distinguish the immune microenvironment typing of DLBCL, and provide a reliable basis for individualized treatment.

[0120] The above description of the embodiments is to facilitate the understanding and use of the invention by those skilled in the art. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative work. Therefore, the present invention is not limited to the above embodiments, and improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope of the present invention should be within the scope of protection of the present invention.

Claims

1. A marker combination for predicting the immune microenvironment typing of diffuse large B-cell lymphoma based on pathological tissue staining, wherein the immune microenvironment typing of diffuse large B-cell lymphoma is GC, MS, IN and DP, characterized in that: It is composed of the following markers: CD3, collagen, CD8, CD68 and PD-L1.

2. The marker combination according to claim 1, characterized in that: The marker combination consists of the following markers: CD3, CD4, CD34, collagen, FAP, CD8, CD68 and PD-L1.

3. Use of the marker combination as described in claim 1 or 2 in constructing a prediction model for immune microenvironment typing of diffuse large B-cell lymphoma.

4. The use according to claim 3, characterized in that: The prediction criteria of the diffuse large B-cell lymphoma immune microenvironment typing prediction model are: (1) First, determine the size of the collagen staining quantitative data and 5.7: if the collagen staining quantitative data is less than 5.7, then continue to determine the size of the CD68 staining quantitative data and 13; if the collagen staining quantitative data is greater than or equal to 5.7, then continue to determine the size of the CD3 staining quantitative data and 39; (2) When the quantitative data of collagen staining is less than 5.7, the size of the quantitative data of CD68 staining and 13 is determined: if the quantitative data of CD68 staining is less than 13, the size of the quantitative data of CD3 staining and 32 is determined; if the quantitative data of CD68 staining is ≥13, it is IN type; (3) When the quantitative data of collagen staining is less than 5.7 and the quantitative data of CD68 staining is less than 13, continue to judge the size of the quantitative data of CD3 staining and 32; if the quantitative data of CD3 staining is less than 32, continue to judge the size of the quantitative data of CD8 staining and 15; if the quantitative data of CD3 staining is ≥32, it is GC type; (4) When the quantitative data of collagen staining is less than 5.7, the quantitative data of CD68 staining is less than 13, and the quantitative data of CD3 staining is less than 32, continue to judge the size of the quantitative data of CD8 staining and 15: if the quantitative data of CD8 staining is less than 15, it is DP type; if the quantitative data of CD8 staining is ≥15, it is IN type; (5) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining and 39 are continued to be judged; if the quantitative data of CD3 staining is ≥39, the quantitative data of CD68 staining and 13 are continued to be judged; if the quantitative data of CD3 staining is <39, the quantitative data of collagen staining and 7.2 are continued to be judged; (6) When the quantitative data of collagen staining is ≥5.7 and the quantitative data of CD3 staining is ≥39, continue to judge the size of the quantitative data of CD68 staining and 13: if the quantitative data of CD68 staining is <13, it is GC type; if the quantitative data of CD68 staining is ≥13, it is IN type; (7) When the quantitative data of collagen staining is ≥5.7 and the quantitative data of CD3 staining is <39, continue to judge the size of the quantitative data of collagen staining and 7.2; if the quantitative data of collagen staining is <7.2, continue to judge the size of the quantitative data of CD68 staining and 19; if the quantitative data of collagen staining is ≥7.2, continue to judge the size of the quantitative data of CD8 staining and 29; (8) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, and the quantitative data of collagen staining is <7.2, continue to judge the size of the quantitative data of CD68 staining and 19: if the quantitative data of CD68 staining is <19, continue to judge the size of the quantitative data of PD-L1 staining and 57; if the quantitative data of CD68 staining is ≥19, it is IN type; (9) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, the quantitative data of collagen staining is <7.2, and the quantitative data of CD68 staining is <19, continue to judge the size of the quantitative data of PD-L1 staining and 57: if the quantitative data of PD-L1 staining is ≥57, it is IN type; if the quantitative data of PD-L1 staining is <57, continue to judge the size of the quantitative data of PD-L1 staining and 15; (10) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, the quantitative data of collagen staining is <7.2, the quantitative data of CD68 staining is <19, and the quantitative data of PD-L1 staining is <57, continue to judge the size of the quantitative data of PD-L1 staining and 15: if the quantitative data of PD-L1 staining is <15, it is DP type; if the quantitative data of PD-L1 staining is ≥15, it is MS type; (11) When the quantitative data of collagen staining is ≥5.7, the quantitative data of CD3 staining is <39, and the quantitative data of collagen staining is ≥7.2, continue to judge the size of the quantitative data of CD8 staining and 29: if the quantitative data of CD8 staining is ≥29, it is IN type; if the quantitative data of CD8 staining is <29, it is MS type.

5. The use according to claim 3, characterized in that: The method for constructing the diffuse large B-cell lymphoma immune microenvironment typing prediction model comprises the following steps: S1. Diffuse large B-cell lymphoma tissue samples were subjected to immunohistochemistry and Mallory's trichrome staining; S2, scanning and reading the slides to obtain quantitative data of marker combination staining; S3. Based on the quantitative data of marker combination staining, a decision tree was constructed using the rpart() function of the R software "RPART" package.

6. The use according to claim 5, characterized in that: In step S1, the immunohistochemistry process includes: dewaxing; hydration; high-pressure heat repair; quenching endogenous peroxidase; blocking; primary antibody incubation; secondary antibody incubation; counterstaining; dehydration; transparency; and sealing.

7. The use according to claim 5, characterized in that: In step S1, the Mallory's trichrome staining process includes: dewaxing; hydration; liquid immersion; liquid immersion; differentiation; dehydration; transparency; and sealing.

8. The use according to claim 5, characterized in that: In step S2, the scanning and reading of the film include: marking positive cells, quantifying the stained area, and obtaining quantitative data of marker combination staining.

9. The use according to claim 8, characterized in that: The positive cell characteristics are: the positive cell nucleus is blue and the positive part is brown after immunohistochemical staining; the positive cell nucleus is red and the fiber is dark blue after Mallory's trichrome staining.

10. The use according to claim 5, characterized in that: In step S3, the staining quantitative data samples were randomly divided into a training set and a validation set. The rpart function of the R software "RPART" package was used to recursively divide the training set and the validation set, respectively, and a decision tree was constructed to obtain a prediction model for the immune microenvironment typing of diffuse large B-cell lymphoma.

Images