Application of reagent for detecting predictive factor in preparation of product for predicting intervention effect of esophageal squamous cell carcinoma patient

Abstract

The invention discloses application of a reagent for detecting a predictive factor in preparation of a product for predicting an intervention effect of a patient with esophageal squamous cell carcinoma. The reagent for detecting the predictive factor comprises a reagent for detecting the content or expression level of a depletion marker of functional CD8 T cells in a sample and a reagent for detecting the content or expression level of immune checkpoint molecules of C1QC + macrophages in the sample. According to the method, key spatial features which are related to specific cell behaviors and influence treatment results are determined and layered; a powerful prediction model of drug response is developed by utilizing a machine learning technology, and the area under curve (AUC) can reach 97%.

Description

[0001] Related Applications

[0002] This application claims priority to Chinese patent application number CN2024111554527, filed on August 21, 2024, entitled “Application of reagents for detecting predictive factors in the preparation of products for predicting intervention effects in patients with esophageal squamous cell carcinoma”, the entire text of which is hereby incorporated by reference. Technical Field

[0003] The present application relates to the field of diagnosis or companion diagnosis, further to the application of predictive factors in predicting the intervention effect of patients with esophageal squamous cell carcinoma, and further to the application of reagents for detecting predictive factors in the preparation of products for predicting the intervention effect of patients with esophageal squamous cell carcinoma. Background Art

[0004] Esophageal squamous cell carcinoma (ESCC) is a global health challenge. The etiology of ESCC is multifactorial, with major risk factors including tobacco use, alcohol consumption, malnutrition, environmental toxins, and genetic susceptibility influencing its pathogenesis. Standard treatment includes surgery, chemotherapy, and radiotherapy tailored to the stage of the disease; however, the global five-year survival rate remains low at only 15%-25%. Neoadjuvant chemoimmunotherapy aims to reduce tumor burden and improve surgical outcomes by reactivating the immune response to tumor cells.

[0005] There is currently a lack of ways to predict the effects of interventions, making it difficult to improve treatments.

[0006] Therefore, there is an urgent need in this field to develop methods that can predict intervention outcomes. Summary of the invention

[0007] Based on this, it is necessary to provide at least one reagent for detecting predictive factors for use in the preparation of a product for predicting the effect of intervention on patients with esophageal squamous cell carcinoma.

[0008] Esophageal squamous cell carcinoma (ESCC) faces significant clinical and therapeutic challenges due to its aggressive nature and generally poor prognosis. A clinical trial was conducted to evaluate the efficacy of a pioneering neoadjuvant chemoimmunotherapy regimen containing a programmed death-1 (PD-1) blocker (Toripalimab), nano-albumin-bound paclitaxel (nab-paclitaxel), and an oral fluoropyrimidine derivative S-1 in patients with locally advanced ESCC. The study uniquely combined clinical results with advanced spatial proteomic profiling, using imaging mass cytometry (IMC) to elucidate dynamic changes in the tumor microenvironment (TME), focusing on resistance and response mechanisms. Among the 60 patients who received the combination therapy, 62% of patients showed a pathological response (MPR) and 29% of patients showed a complete pathological response (pCR). IMC analysis provided detailed regional assessments, revealing the expression levels of immune checkpoint-related molecules in CD8+T cells, CD4+T cells, and B cells at the epithelial-immune junction, as well as C1QC in the fibrosis-epithelial junction area. + Inflammatory macrophages are predictors of treatment outcomes. Through machine learning approaches, such as support vector machine (SVM) and random forest (RF) analysis, we identified key spatial features associated with drug resistance and developed a drug response prediction model that achieved an area under the curve (AUC) of 97%. These insights highlight the importance of integrating spatial proteomics in clinical trials to comprehensively resolve TME dynamics and pave the way for personalized and precise treatment strategies for ESCC. This holistic approach not only enhances our understanding of the mechanistic basis of drug resistance but also lays a solid foundation for optimizing ESCC therapeutic interventions.

[0009] Synergistic combinations of PD-1 blockers, taxanes (e.g., nab-paclitaxel), and S-1 present the potential to modulate antitumor immunity in the neoadjuvant ESCC setting. PD-1 inhibitors have shown significant promise by reactivating exhausted T cells, enabling them to effectively target and eliminate tumor cells. Nab-paclitaxel, a formulation of paclitaxel conjugated to albumin nanoparticles, is known for its enhanced permeability and retention effects, leading to increased drug accumulation at tumor sites. In addition, it has been observed to reduce the number of regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs) in the tumor microenvironment (TME), further promoting immune-mediated cancer cell attack. S-1 is an oral fluoropyrimidine derivative that combines tegafur with its metabolic modulator to enhance the efficacy of 5-fluorouracil (5-FU) while reducing its systemic toxicity. Its inclusion in the combination is based on its ability to inhibit thymidylate synthase, leading to DNA damage in proliferating tumor cells. In addition, 5-FU, the active ingredient in S-1, has been suggested to induce immunogenic cell death, potentially enhancing the efficacy of immunotherapeutic agents. By exploiting the individual mechanisms of action of these agents, we aim to dismantle the tumor's protective barrier to the immune system in a complementary manner, enhance the immunogenicity of tumor cells, and promote a more robust and durable immune response. This approach not only targets the primary tumor but also addresses systemic disease, providing a comprehensive strategy to combat ESCC in the neoadjuvant setting.

[0010] The TME of ESCC consists of a complex interaction of cancer cells, cancer-associated fibroblasts (CAFs), various immune cells including macrophages, T cells and MDSCs, and endothelial cells surrounded by a dynamic extracellular matrix (ECM). This complex network not only supports tumor growth and metastasis by secreting cytokines and growth factors, but also enhances tumor invasiveness and complicates treatment response through immune escape and ECM remodeling. Understanding these interactions is crucial for developing targeted therapies that effectively disrupt pathological processes within the ESCC TME. Traditional TME research methods, such as whole tissue analysis, often fail to capture the spatial heterogeneity that is critical for understanding tumor behavior and treatment response. These traditional techniques provide a generalized view that averages the cellular and molecular content of tissue samples, masking local interactions and microenvironmental information that significantly influence tumor behavior and treatment response. Imaging mass cytometry (IMC) enables detailed spatial proteomic profiling of the TME. This technology combines the principles of mass cytometry with imaging capabilities, allowing for the simultaneous detection of multiple proteins at high spatial resolution. IMC facilitates visualization of cellular and molecular patterns in their native tissue context, providing insights into the cellular architecture and functional status of the TME. By preserving spatial information and allowing cell-by-cell analysis, IMC helps delineate the heterogeneity and complexity of the TME, providing new avenues for targeted therapeutic intervention.

[0011] The primary goal of this study is to elucidate the spatial proteomic landscape of ESCC by using advanced IMC technologies, aiming to identify TME features that predict chemoimmunotherapy response. Second, we seek to use these insights to tailor targeted therapeutic strategies to improve the precision of interventions in clinical trials. Finally, we intend to develop a predictive tool that integrates spatial proteomic data with machine learning techniques to predict individual responses to chemoimmunotherapy and enhance personalized treatment plans for ESCC.

[0012] In a first aspect of the present application, there is provided an application of a reagent for detecting predictive factors in the preparation of a product for predicting the intervention effect on patients with esophageal squamous cell carcinoma, wherein the reagent for detecting predictive factors comprises: a reagent 1 for detecting the content of an exhaustion marker of functional CD8 T cells in a sample or its expression level and a reagent 2 for detecting the content of inflammatory macrophages in a sample.

[0013] In some embodiments, a reagent for detecting predictive factors is provided for use in preparing a product for predicting the intervention effect on patients with esophageal squamous cell carcinoma, wherein the reagent for detecting predictive factors comprises: a reagent 1 for detecting the content of an exhaustion marker of functional CD8T cells in a sample or its expression level and a reagent 2 for detecting the expression level of an immune checkpoint molecule of C1QC+ macrophages in a sample.

[0014] In some embodiments, the exhaustion marker includes at least one of PD-1, TIM3, CD38, CD39, CD73, TIGIT, VISTA, GITR, LAG3, and B7H4.

[0015] In some embodiments, the exhaustion marker comprises at least one of PD-1, TIM3, CD39, TIGIT, VISTA, GITR, LAG3, and CTLA4.

[0016] In some embodiments, the immune checkpoint molecule includes at least one of IDO1, PD-L1, PD-L2, CD47 and B7H4.

[0017] In some embodiments, the functional CD8 T cells include CD45RO + CD8 + T cells.

[0018] In some embodiments, the esophageal squamous cell carcinoma patient is a locally advanced esophageal squamous cell carcinoma patient.

[0019] In some embodiments, the inflammatory macrophages are selected from C1QC + 、IDO1 +The group consisted of macrophages with S100A9+ and macrophages with S100A9+ in the fibrotic area.

[0020] In some embodiments, the macrophages are selected from the epithelial / fibroblast junction C1QC + 、IDO1 + Groups of macrophages.

[0021] In some embodiments, the reagent 1 is one or more of a specific primer, a specific probe and an antibody for detecting the content of the depletion marker or its expression level.

[0022] In some embodiments, the reagent 2 includes a detection C1QC + 、IDO1 + One or more of specific primers, specific probes and antibodies for detecting the IDO1 content or expression level of macrophages.

[0023] In some embodiments, the reagent 2 comprises a method for detecting S100A9 in the fibrotic region. + One or more of specific primers, specific probes and antibodies for detecting the S100A9 content or expression level of macrophages.

[0024] In some embodiments, the reagent 2 includes a C1QC for detecting the epithelial-fibroblast junction. + 、IDO1 + One or more of specific primers, specific probes and antibodies for the content of immune checkpoint molecules in macrophages or their expression levels.

[0025] In some embodiments, detecting the amount of exhaustion markers of functional CD8 T cells or their expression levels in the sample is performed before intervention.

[0026] In some embodiments, the method for predicting the intervention effect in patients with esophageal squamous cell carcinoma further comprises the following characteristic analysis before intervention: CD4+ T / B cell ratio analysis and CD8+ T cell and B cell level analysis in tertiary lymphoid structures.

[0027] In some embodiments, the sample source of the predictor is tumor tissue from a patient with esophageal squamous cell carcinoma.

[0028] In some embodiments, the method for predicting the intervention effect in patients with esophageal squamous cell carcinoma further comprises the following characteristic analysis after intervention: frequency analysis of CD8+ T cells and B cells.

[0029] In some embodiments, the esophageal squamous cell carcinoma patient has undergone combined intervention with a PD-1 inhibitor, a taxane compound, and a fluoropyrimidine derivative.

[0030] In some embodiments, the PD-1 inhibitor is Toripalimab.

[0031] In some embodiments, the fluoropyrimidine derivative is S-1.

[0032] In some embodiments, the paclitaxel compound is nanoalbumin-bound paclitaxel.

[0033] In a second aspect of the present application, a kit for predicting the intervention effect of patients with esophageal squamous cell carcinoma is provided, which comprises a reagent 1 for detecting the content or expression level of an exhaustion marker of functional CD8 T cells in a sample and a reagent 2 for detecting the content of inflammatory macrophages in a sample.

[0034] In some embodiments, a kit for predicting the intervention effect of patients with esophageal squamous cell carcinoma is provided, which comprises a reagent 1 for detecting the content or expression level of an exhaustion marker of functional CD8 T cells in a sample and a C1QC + Reagent 2 for the expression level of immune checkpoint molecules in macrophages.

[0035] Wherein the reagent 1 and the reagent 2 are as defined in the first aspect above.

[0036] In the third aspect of the present application, a method for predicting and / or diagnosing the intervention effect of patients with esophageal squamous cell carcinoma is provided, comprising: detecting predictive factors before intervention, wherein the predictive factors include: the content of exhaustion markers of functional CD8 T cells in the sample or their expression level and the content of inflammatory macrophages in the sample.

[0037] In some embodiments, a method for predicting and / or diagnosing the intervention effect of patients with esophageal squamous cell carcinoma is provided, comprising: detecting predictive factors before intervention, wherein the predictive factors include: the content of the exhaustion marker of functional CD8 T cells in the sample or its expression level and C1QC + Expression levels of immune checkpoint molecules in macrophages.

[0038] In some embodiments, the predictive analysis method is a non-diagnostic and non-therapeutic method.

[0039] In some embodiments, the predictive analysis method is used for diagnosis and / or treatment.

[0040] In some embodiments, the predictors further include CD4+ T / B cell ratio and CD8+ T cell and B cell levels in tertiary lymphoid structures.

[0041] In some embodiments, the predictive analysis method further comprises an analysis of indicators after intervention, wherein the indicators after intervention comprise the frequencies of CD8+ T cells and B cells.

[0042] In some embodiments, the sample source of the predictor can be found in the description in the context.

[0043] In some embodiments, the detection of the predictive factor can be performed using the aforementioned reagents for detecting the predictive factor. The reagents for detecting the predictive factor can refer to the description in the above context, but are not limited thereto.

[0044] In a fourth aspect of the present application, a method for constructing a model for predicting the intervention effect of esophageal squamous cell carcinoma is provided, the method comprising:

[0045] Obtain detection data of predictive factors in the sample; the predictive factors at least include the content or expression level of exhaustion markers of functional CD8 T cells and inflammatory macrophages (such as C1QC + macrophages) content;

[0046] The intervention effect of the esophageal squamous cell carcinoma is predicted according to the detection data of the predictor.

[0047] In some embodiments, the construction method comprises:

[0048] Obtain detection data of predictive factors in the sample; the predictive factors at least include the content or expression level of exhaustion markers of functional CD8 T cells in the sample and C1QC + the content of immune checkpoint molecules in macrophages;

[0049] The intervention effect of the esophageal squamous cell carcinoma is predicted according to the detection data of the predictor.

[0050] In some embodiments, the exhaustion marker includes at least one of PD-1, TIM3, CD38, CD39, CD73, TIGIT, VISTA, GITR, LAG3, and B7H4.

[0051] In some embodiments, the exhaustion marker comprises at least one of PD-1, TIM3, CD39, TIGIT, VISTA, GITR, LAG3, and CTLA4.

[0052] In some embodiments, the functional CD8 T cells include CD45RO + CD8 + T cells.

[0053] In some embodiments, the esophageal squamous cell carcinoma patient is a locally advanced esophageal squamous cell carcinoma patient.

[0054] In some embodiments, the inflammatory macrophages are selected from the group consisting of C1QC+, IDO1+ macrophages, and S100A9+ macrophages in fibrotic areas.

[0055] In some embodiments, the macrophages are selected from the group of C1QC+ and IDO1+ macrophages at the epithelial-fibroblast junction.

[0056] In some embodiments, the detection data includes the content or expression level of an exhaustion marker of functional CD8 T cells, and the content of inflammatory macrophages.

[0057] In some embodiments, the detection data includes the content or expression level of the exhaustion marker of functional CD8 T cells, and C1QC + Expression levels of immune checkpoint molecules in macrophages.

[0058] In some embodiments, the content or expression level of the depletion marker is obtained by detecting one or more of specific primers, specific probes and antibodies for detecting the content or expression level of the depletion marker.

[0059] In some embodiments, the content of the inflammatory macrophages is obtained by detecting one or more of specific primers, specific probes and antibodies of markers of the inflammatory macrophages.

[0060] In some embodiments, by detecting C1QC at the epithelial-fibrous junction + One or more of the specific primers, specific probes and antibodies for immune checkpoint molecules in macrophages are used to obtain the C1QC + Content of immune checkpoint molecules in macrophages.

[0061] In some embodiments, the obtaining of detection data of exhaustion markers of functional CD8 T cells is performed before intervention.

[0062] In some embodiments, the following parameters are obtained before intervention: CD4+ T / B cell ratio and CD8+ T cell and B cell levels in tertiary lymphoid structures.

[0063] In some embodiments, the construction method further comprises obtaining the frequencies of CD8+ T cells and B cells after intervention.

[0064] In some embodiments, the intervention comprises a combined intervention of a PD-1 inhibitor, a taxane compound, and a fluoropyrimidine derivative.

[0065] In some embodiments, the PD-1 inhibitor is Toripalimab.

[0066] In some embodiments, the paclitaxel compound is nanoalbumin-bound paclitaxel.

[0067] In some embodiments, the fluoropyrimidine derivative is S-1.

[0068] In a fifth aspect of the present application, a system for predicting the intervention effect of patients with esophageal squamous cell carcinoma is provided, the system comprising:

[0069] A data processing module, used for analyzing the data of the content of the exhaustion marker of functional CD8 T cells in the received or input sample and the content of inflammatory macrophages in the sample to obtain an analysis result; and

[0070] The judgment and output module is used to judge whether the analysis result meets the preset judgment conditions to diagnose and / or predict the intervention effect of patients with esophageal squamous cell carcinoma, and output the judgment result.

[0071] In some embodiments, a system for predicting the effect of an intervention on a patient with esophageal squamous cell carcinoma is provided, the system comprising:

[0072] The data processing module is used to convert the content of the exhaustion marker of functional CD8 T cells in the received or input sample and the C1QC in the sample + Analyze the data of expression levels of immune checkpoint molecules in macrophages to obtain analysis results; and

[0073] The judgment and output module is used to judge whether the analysis result meets the preset judgment conditions to diagnose and / or predict the intervention effect of patients with esophageal squamous cell carcinoma, and output the judgment result.

[0074] In some embodiments, the determination and output module includes a machine learning model for performing the analysis.

[0075] In some embodiments, the machine learning model is selected from one or more of a generalized linear model (GLM), a random forest (RF), a neural network, and a shrinkage discriminant analysis.

[0076] In a sixth aspect of the present application, a computer-readable storage medium is provided, and when the computer program is executed by a processor, the functions of the system described in the fifth aspect can be implemented.

[0077] In a seventh aspect of the present application, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is used to execute the computer program to implement the functions of the system as described in the fifth aspect.

[0078] In the eighth aspect of the present application, a device for predicting the intervention effect of patients with esophageal squamous cell carcinoma is provided, which includes one or more of the system described in the fifth aspect, the computer-readable storage medium described in the sixth aspect, and the computer device described in the seventh aspect.

[0079] In a ninth aspect of the present application, a method for analyzing a tumor microenvironment (TME) of esophageal squamous cell carcinoma is provided, the method comprising:

[0080] The assay analyzes the spatial arrangement of immune cells in tertiary lymphoid structures and cytokine levels in fibrotic areas in esophageal squamous cell carcinoma tissues.

[0081] In some embodiments, the cytokines include at least S100A9+ inflammatory macrophages.

[0082] In some embodiments, the cytokines include at least C1QC + Expression level of IDO1 in macrophages.

[0083] In some embodiments, the tumor microenvironment (TME) is the TME of a patient with esophageal squamous cell carcinoma during intervention.

[0084] In some embodiments, the method further comprises detecting one or more of the following indicators:

[0085] Expression of exhaustion markers on functional CD8 T cells;

[0086] Q1C + 、IDO1 + The level of macrophages;

[0087] CD4+ T / B cell ratio.

[0088] In some embodiments, the method further comprises detecting one or more of the following indicators:

[0089] Expression of exhaustion markers on functional CD8 T cells;

[0090] Q1C + The expression level of IDO1 in macrophages;

[0091] CD4+ T / B cell ratio.

[0092] In some embodiments, the functional CD8 T cells are CD45RO + CD8 + T cells.

[0093] In some embodiments, the exhaustion marker includes at least one of PD-1, TIM3, CD38, CD39, CD73, TIGIT, VISTA, GITR, LAG3, and B7H4.

[0094] In some embodiments, the exhaustion marker comprises at least one of PD-1, TIM3, CD39, TIGIT, VISTA, GITR, LAG3, and CTLA4.

[0095] In some embodiments, the esophageal squamous cell carcinoma patient is a locally advanced esophageal squamous cell carcinoma patient.

[0096] In some embodiments, expression of an exhaustion marker on functional CD8 T cells is determined prior to intervention.

[0097] In some embodiments, CD4 + T / B cell ratio.

[0098] In some embodiments, CD4 + T / B cell ratio.

[0099] In some embodiments, imaging mass cytometry (IMC) is used to detect and analyze the spatial arrangement of the immune cells in the tertiary lymphoid structures and the cytokine levels in the fibrotic areas in the esophageal squamous cell carcinoma tissue.

[0100] In some embodiments, judging whether the cytokine content or expression level in the fibrosis area is a predictive factor for predicting the intervention effect of esophageal squamous cell carcinoma based on the detection and analysis results of imaging mass spectrometry cytometry;

[0101] Alternatively, it is determined based on the detection and analysis results of imaging mass spectrometry cytometry whether the cytokines in the fibrosis area are markers for predicting the intervention effect of esophageal squamous cell carcinoma.

[0102] In some embodiments, the immune cells are selected from the group consisting of CD8+ T cells and B cells.

[0103] In some embodiments, the method is used to analyze dynamic changes in the tumor microenvironment (TME).

[0104] In some embodiments, the method further comprises establishing a spatial proteomic map of the spatial organization of the tumor microenvironment (TME) during the intervention period based on the analysis results.

[0105] In some embodiments, the intervention comprises chemoimmunotherapy.

[0106] In some embodiments, the method further comprises identifying TME features that predict response to chemoimmunotherapy.

[0107] In some embodiments, the intervention comprises a combined intervention of a PD-1 inhibitor, a taxane compound, and a fluoropyrimidine derivative.

[0108] In some embodiments, the PD-1 inhibitor includes Toripalimab, and may further be Toripalimab.

[0109] In some embodiments, the paclitaxel compound is nanoalbumin-bound paclitaxel.

[0110] In some embodiments, the fluoropyrimidine derivative comprises S- 1. In some embodiments, the fluoropyrimidine derivative is S-1.

[0111] In some embodiments, the TME includes a complex interaction of cancer cells, cancer-associated fibroblasts (CAFs), various immune cells including macrophages, T cells and MDSCs, and endothelial cells surrounded by a dynamic extracellular matrix (ECM).

[0112] In some embodiments, the method further comprises: clustering and annotating cells based on the components of surrounding cells at each cell's spatial location, determining the neighborhood distribution of cells, and comparing the inter-group differences in neighborhood distribution of patients with different intervention effects.

[0113] In the tenth aspect of the present application, a method for treating esophageal squamous cell carcinoma is provided, the method comprising administering a drug combination to a subject in need thereof, the drug combination comprising a PD-1 inhibitor, a taxane compound, and a fluoropyrimidine derivative.

[0114] In some embodiments, the PD-1 inhibitor, the taxane compound, and the fluoropyrimidine derivative are administered separately or simultaneously.

[0115] In some embodiments, the pharmaceutical combination is in the form of a pharmaceutical composition, which further comprises a pharmaceutically acceptable carrier and / or excipient.

[0116] In some embodiments, the drug combination is administered by one or more of the following: oral, injection, implant, spray and / or inhalation; for example, parenteral, subcutaneous, intraperitoneal, intravenous, intralesional, intracerebral, intra-arterial interstitial infusion and / or implanted delivery device.

[0117] In some embodiments, the dosage form of the drug combination is one or more selected from the following: injection, oral solution, powder, tablet, granule, capsule, syrup, decoction, sustained-release preparation, enteric solvent, aerosol or suspension.

[0118] In some embodiments, the fluoropyrimidine derivative is S-1.

[0119] In some embodiments, the subject in need thereof is a patient with locally advanced esophageal squamous cell carcinoma.

[0120] In the eleventh aspect of the present application, there is provided use of a PD-1 inhibitor, a paclitaxel compound and a fluoropyrimidine derivative in the preparation of a medicament for treating esophageal squamous cell carcinoma.

[0121] In some embodiments, the drug is administered by one or more of the following: oral, injection, implant, spray and / or inhalation; for example, parenteral, subcutaneous, intraperitoneal, intravenous, intralesional, intracerebral, intra-arterial interstitial infusion and / or implanted delivery device.

[0122] In some embodiments, the dosage form of the drug is one or more selected from the following: injection, oral solution, powder, tablet, granule, capsule, syrup, decoction, sustained-release preparation, enteric solvent, aerosol or suspension.

[0123] In some embodiments, the fluoropyrimidine derivative is S-1.

[0124] The trial in one embodiment of the present application demonstrated the significant efficacy of a new neoadjuvant chemoimmunotherapy regimen in esophageal squamous cell carcinoma (ESCC), with 62% of patients achieving pathological remission (MPR) and 29% of patients achieving complete pathological remission (pCR). A spatial proteomic map of the complex spatial organization of the tumor microenvironment (TME) during treatment was established. Key spatial features associated with specific cell behaviors and affecting treatment outcomes were identified and stratified. Using machine learning technology, a powerful prediction model for drug response was developed with an area under the curve (AUC) of 97%. BRIEF DESCRIPTION OF THE DRAWINGS

[0125] In order to more clearly illustrate the technical solutions in the implementation methods and examples of the present application and to more completely understand the present application and its beneficial effects, the following is a brief introduction to the drawings required for use in the description of the implementation methods or examples. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. It should also be noted that the drawings are all drawn in a simplified form and are only used to conveniently and clearly assist in explaining the present application.

[0126] Figure 1AEnrollment plan for patient selection. This flow chart illustrates the process of evaluating and selecting patients for participation in a clinical trial, detailing the number of patients initially evaluated, exclusion criteria, and the final population of patients who met the inclusion criteria and received neoadjuvant chemoimmunotherapy.

[0127] Figure 1B Treatment planning and sample collection time points. This diagram shows a timeline of the treatment regimen that the patient received, emphasizing the pre- and post-treatment periods. Key time points for biopsy sample collection for molecular analysis are labeled to indicate when samples for IMC were obtained.

[0128] Figure 1C Representative CT images showing treatment response. This figure shows CT images of patients who showed different responses to neoadjuvant therapy, with images divided into those showing good response (grades 0 and 1) and poor response (grades 2 and 3), providing a visual record of changes in tumor size and morphology after treatment.

[0129] Figure 1D This figure summarizes the baseline demographic and clinical characteristics of the patient cohort, including tumor stage and lymph node involvement. It also includes the distribution of treatment outcomes according to RECIST 1.1 criteria, showing the response rates within the treatment groups.

[0130] Figure 1E and Figure 1F Figure 2 Kaplan-Meier survival curves showing the overall survival (OS) and progression-free survival (PFS) of 55 patients.

[0131] Figure 2A Schematic representation of the IMC and analysis workflow. This panel illustrates the experimental setup and subsequent data analysis pipeline in our study, from tissue sampling to IMC acquisition and the computational strategy used for data processing and image analysis.

[0132] Figure 2B Designed for the IMC Antibody Panel, this panel demonstrates markers used in research, showing lineage and major functional markers for cell type identification.

[0133] Figure 2C This panel shows representative single-color images from IMCs showing the expression of a specific marker in a tissue section. This visualization helps identify the spatial distribution and intensity of the marker in the cellular microenvironment.

[0134] Figure 2DHeatmap of major and detailed cell phenotypes, a two-layer heatmap where the top left section represents the major cell types identified in the TME and the other sections drill down into more detailed subtypes, highlighting their expression profiles in different samples or conditions. The color of each square represents the average expression level of the corresponding marker in the cell type.

[0135] Figure 2E The Unified Manifold Approximation and Projection (UMAP) visualization reduces multidimensional data to a two-dimensional space, facilitating the identification of clusters and patterns in the complex cellular landscape of the TME.

[0136] Figure 2F This panel provides a filled bar graph to compare the cell composition of samples before and after treatment in different treatment response groups. Each bar represents a sample, and the length of the colored segment represents the proportion of the corresponding cell type.

[0137] Figure 2G This panel provides a detailed visualization of the distribution of cell subtypes across response groups and before and after treatment, showing the distribution and frequency of specific cell subtypes (such as epithelial cells, lymphocytes, and myeloid cells), stratified by treatment response category (good vs. poor) and pre- and post-treatment time points. This detailed map is critical to elucidating cellular dynamics in treatment response and its potential relevance to clinical outcomes. To quantify changes, the difference in frequency change is expressed on a logarithmic scale (log2) using color coding from blue (decrease, -5) to red (increase, +5), providing a visual representation of the fluctuation in cell proportion. Statistical significance is indicated by the size of the squares, with larger squares indicating statistically significant changes (p-value < 0.05) and smaller squares indicating non-significant changes. Four comparative analyses were included, including comparison of good and poor response groups before treatment (0-1 before treatment vs. 2-3 before treatment), comparison of good and poor response groups after treatment (0-1 after treatment vs. 2-3 after treatment), longitudinal comparison of good response group from pre-treatment to post-treatment (0-1 after treatment vs. 0-1 before treatment), and longitudinal comparison of poor response group from pre-treatment to post-treatment (2-3 after treatment vs. 2-3 before treatment) to provide a comprehensive understanding of cellular changes. In addition, the left side of the figure shows the average frequency of each cell subtype in all samples, providing a quantitative baseline to measure changes. This multifaceted approach not only highlights the different effects of therapeutic interventions on various cell populations, but also helps identify potential cellular biomarkers to predict treatment efficacy.

[0138] Figure 3AThis panel shows the typical organization and cellular composition of TLS detected by time-of-flight mass cytometry (MCD) to demonstrate the organization and cellular composition of TLS by MCD and schematic diagrams. It includes a detailed schematic diagram highlighting the spatial arrangement of B cells and T cells in these structures. In addition, this panel describes how to identify TLS using a region recognition algorithm, with a focus on identifying densely organized B cells in these structures.

[0139] Figure 3B This figure shows the change in TLS area size, comparing the increase in good and poor response groups after treatment. Both groups showed significant increases after treatment compared to before treatment. Each point represents the average TLS cell frequency for a specific patient.

[0140] Figure 3C Middle panel 1 shows the cell composition within the TLS. The bar graph shows the proportion of key immune cell types within the TLS, including an increase in CD8+ T / B cells in the good responder group, highlighting their potential role in mediating positive treatment outcomes. The Y axis represents the overall frequency of cells within the TLS. Figure 3C Middle panel 2 is a statistical analysis of TLS composition. The Y axis represents the relative frequency of the corresponding cell types in the four groups. The t-test comparison shows the difference in the composition of immune cells in TLS between the good and poor response groups, highlighting the inhibitory effect of specific cell types in poor response patients. Each point represents the average frequency of the corresponding immune cell in TLS of a specific patient.

[0141] Figure 4A and Figure 4B This panel shows the expression changes of functional markers such as S100A9, PD-1, PD-L1, and HIF-1α in different cell populations in TLS, correlating these markers with the efficacy of response after treatment. Each point represents the mean expression level fold change of a marker in a specific cell type within a specific comparison group, the X-axis represents the log2 fold change, and the Y-axis represents the -log10 p value.

[0142] Figure 5 For environmental composition analysis, the composition of immune cells in the area surrounding TLS is shown, focusing on the frequency of activated CD45RO+ CD8+ T cells in the good response group, suggesting a favorable immune environment after treatment. The bar graph in the upper panel represents the overall frequency of immune cells, and the lower panel represents the comparison between different groups.

[0143] Figure 6 For the statistical evaluation of the environmental composition, the t-test comparison showed the difference in the proportion of immune cells in the TLS surrounding environment between the good and poor response groups after treatment, suggesting the influence of the local immune environment on the treatment outcome. Each point in the box plot represents the frequency of a certain sample.

[0144] Figure 7 This panel provides a detailed schematic diagram of the TME regions, including epithelial, immune, and fibrotic regions, as well as cross-regions such as epithelial / immune and epithelial / fibrotic regions. The parallel display of the IMC image and the regional definition schematic allows for intuitive depiction of the boundaries and features of each region, making it easier to understand the spatial analysis method.

[0145] Figure 8 This figure shows a quantitative analysis of the size of the zones within the TME before and after treatment, categorized by response group (good vs. poor), for comparative analysis of zone size before and after treatment. It highlights the significant reduction in epithelial area after treatment and provides a visual comparison to emphasize the impact of treatment outcomes on zone dynamics. Each point in the boxplot represents the zone size for a certain sample.

[0146] Fig. 9 This panel quantifies the cellular composition of different TME regions and compares the proportions of key immune and stromal cell types within different response groups before and after treatment. The focus is on identifying significant changes associated with treatment efficacy, such as an increase in effector cells and a decrease in suppressor cell populations.

[0147] Fig.10 This figure shows the changes in functional markers in the TME region, including immune checkpoints and inflammatory markers. It compares the changes in the status of the good and poor response groups before and after treatment, emphasizing the regulatory dynamics that potentially affect treatment outcomes. Each point represents the average expression level fold change of a marker in a cell type within a specific comparison group, the X-axis represents the log2 fold change, and the Y-axis represents the -log10 p value.

[0148] Fig.11 : Comparison of the composition of the surrounding environment of TME regions. This panel explores the changes in the composition of the surrounding environment of TME regions after treatment, detailing changes in cytokine profiles, growth factors, and other soluble mediators. It compares these environmental components between different regions and response groups, highlighting how the local biochemical environment modulates immune responses and affects clinical outcomes. Each square represents the log2 frequency fold change of the corresponding cell type, and only comparisons with t-test p-values ​​< 0.05 are shown.

[0149] Fig.12 This panel provides a schematic diagram of the method used to analyze cell interactions. A central cell is identified and its interactions with the ten nearest neighboring cells are evaluated to assess local cell communication dynamics, illustrating the analytical framework employed.

[0150] Fig.13 Figure 1 shows the interaction patterns in the immune region. This figure shows the enhanced interaction of T cells with other components in the immune region after treatment, suggesting an active immune response. Each square represents the log2 fold change of the corresponding cell-cell interaction, and only comparisons with t-test p value < 0.05 are shown.

[0151] Fig.14 This figure shows the T cell interactions that were significantly upregulated in the immune region in the good responder group (0-1 after treatment vs. 0-1 before treatment, orange), suggesting effective T cell infiltration and activation. The height of the bar corresponds to the log2 fold change of the interaction between the corresponding comparison groups.

[0152] Fig.15 This panel shows the enhanced interaction between T cells and fibroblasts after treatment within the epithelial-immune junction, highlighting the dialogue between epithelial and fibroblast cells in this mixed area. Each square represents the log2 fold change of the corresponding cell-cell interaction, and only comparisons with t-test p value < 0.05 are shown.

[0153] Fig.16 This figure demonstrates increased interactions between epithelial cells and T cells in the epithelial-immune junction in the good responder group, further emphasizing the importance of T cell presence and activity in mediating positive clinical outcomes. The height of the bar corresponds to the log2 fold change of the interaction between the corresponding comparison groups.

[0154] Fig.17 Feature selection process for SVM analysis. This panel outlines the rigorous process of selecting and optimizing the feature space from an initial set of over 8,000 features. The selection criteria included removing highly correlated features and subsequently eliminating features with too many zero values ​​in the sample, resulting in approximately 3,000 optimized features. This flowchart details each step of the feature reduction process to ensure the best input for the machine learning analysis.

[0155] Fig.18 This boxplot shows the importance of selected pre-treatment features identified by random forest analysis as key predictors of treatment response; the heatmap shows the means scaled by group. It highlights key features, including CD45RO + CD8 + T cells and CD4 in specific TME regions +PD-1 expression in T / B cells, these features are significantly enhanced in the good responder group. This visual representation emphasizes the predictive value of these immune markers in predicting treatment outcomes.

[0156] Fig.19 This line graph shows the relationship between the number of top features used in a prediction model and the resulting AUC value. It shows that 92% AUC can be achieved using only the first three features, and the AUC reaches a plateau of 97% after increasing to ten features. This analysis emphasizes the efficiency and predictive power of using a small number of highly informative features.

[0157] Fig. 20 The ROC curves are for the prediction effects of different models using only the top five most important features.

[0158] Fig.21 This box plot compares the importance of various post-treatment features, similar to Fig.18 Pre-treatment analysis in . It evaluates changes in feature importance after treatment and identifies new key therapeutic targets. This comparison helps understand the dynamic changes of TME in treatment response and its impact on ongoing treatment strategies.

[0159] Fig. 22 For the correlation between rash and treatment outcome in one embodiment of the present application, the distribution of the presence or absence of rash among patients showing good (0-1) and poor (2-3) responses is shown.

[0160] Fig.23 The distribution of LAMP3-positive DCs under different conditions in one embodiment of the present application shows the distribution of LAMP3-positive dendritic cells (DCs) under different conditions.

[0161] Fig.24A Ten cell clusters identified in the UMAP map: The UMAP map shows ten cell clusters identified from scRNA-seq data.

[0162] Fig. 24B for Fig.24A The top three cluster-specific signature genes of the ten cell clusters shown in Fig.24A The top three specific signature genes for each cell cluster are shown in .

[0163] Fig.24C The distribution patterns of the main groups between the normal group and the tumor group are shown.

[0164] Fig.24DSix macrophage clusters identified in the UMAP image are shown. Six macrophage clusters identified in the UMAP image are shown.

[0165] Fig.24E Specific signature genes of the six macrophage clusters are shown.

[0166] Fig.24F The distribution patterns of macrophage subsets between the normal group and the tumor group are shown. The distribution patterns of macrophage subsets between the normal group and the tumor group are shown.

[0167] Fig.25 The UMAP diagram of scRNA-seq data highlights the expression of marker gene S100A9. The UMAP diagram shows the expression of marker gene S100A9 in scRNA-seq data.

[0168] Fig.26 Figure 2 is the Kaplan-Meier survival curve based on the S100A9 macrophage GSVA score. Figure 2 is the Kaplan-Meier survival curve based on the S100A9 macrophage GSVA score. DETAILED DESCRIPTION

[0169] In order to facilitate the understanding of the present application, the present application will be described more fully below with reference to the relevant drawings. The preferred embodiments of the present application are given in the drawings. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive.

[0170] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0171] In this application, unless otherwise specified, "one or more" means any one of the listed items or any combination of the listed items. Similarly, "one or more" and the like, when used in other ways to mean "one or more", are also understood in the same way unless otherwise specified.

[0172] The term "combination" or the like as used in this application includes all suitable combinations of any two or more of the listed items.

[0173] In this application, the word "suitable" in "suitable combination", "suitable method", "any suitable method", etc. is based on the ability to implement the technical solution of this application, solve the technical problems of this application, and achieve the expected technical effects of this application.

[0174] In this application, "further", "further", "particularly", "for example", "such as", "example", "for example", etc. are used for descriptive purposes, indicating that the previous and subsequent different technical solutions are related in terms of the content covered, but should not be understood as a limitation on the previous technical solution, nor can it be understood as a limitation on the protection scope of this application. In this application, unless otherwise specified, A (such as B) means that B is a non-limiting example of A, and it can be understood that A is not limited to B.

[0175] In this application, "optionally", "optional", and "optional" mean optional, that is, any one of the two parallel solutions of "yes" or "no". If there are multiple "optional" in a technical solution, unless otherwise specified and there is no contradiction or mutual restriction, each "optional" is independent. Unless otherwise specified, the descriptions of "optionally include", "optionally contain", etc. in this application, taking "optionally include" as an example, mean "may include or not include".

[0176] The terms "contain", "include" and "include" used in this application are synonymous, which are inclusive or open-ended and do not exclude additional, uncited members or features. Members or features include materials or components, structures, elements, instruments, etc.; non-limiting examples of members or features include actions, conditions for the occurrence of actions, timing, states, etc.

[0177] The terms "and / or", "or / and", and "and / or" used in this application include any one of two or more related listed items, and also include any and all combinations of related listed items. "Any and all combinations" include any combination of two related listed items, any more related listed items, or all related listed items. For example, "A and / or B" includes three parallel solutions: A, B, and "a combination of A and B".

[0178] In this application, the technical features or technical solutions described in an open language include closed technical features or technical solutions composed of the listed contents, and also include open technical features or technical solutions containing the listed contents.

[0179] In the present application, exemplary descriptions such as "in some implementation modes (or examples)" and "in one implementation mode (or example)" may include but are not limited to the following meanings: these solutions may be combined with other solutions in a suitable manner to form new technical solutions.

[0180] In the present application, in "the first aspect", "the second aspect", "the third aspect", "the fourth aspect", etc., the terms "first", "second", "third", "fourth", etc. are used only for descriptive purposes and cannot be understood as indicating or implying relative importance or quantity, nor can they be understood as implicitly indicating the importance or quantity of the indicated technical features. Moreover, "first", "second", "third", "fourth", etc. only serve the purpose of non-exhaustive enumeration and description, and it should be understood that they do not constitute a closed limitation on quantity.

[0181] In the present application, when it comes to a numerical interval (i.e., a numerical range), unless otherwise specified, the distribution of the optional numerical values ​​in the numerical interval is considered to be continuous, and includes the two numerical endpoints (i.e., the minimum and maximum values) of the numerical interval, and each numerical value between the two numerical endpoints. Unless otherwise specified, when the numerical interval only refers to an integer in the numerical interval, it includes the two endpoint integers of the numerical range, and each integer between the two endpoints, which is equivalent to directly listing each integer. When multiple numerical ranges are provided to describe features or characteristics, these numerical ranges can be merged. In other words, unless otherwise specified, the numerical range disclosed herein should be understood to include any and all sub-ranges included therein. The "numerical value" in the numerical interval can be any quantitative value, such as a number, a percentage, a ratio, etc. "Numerical interval" allows broadly including numerical interval types such as percentage intervals, ratio intervals, and ratio intervals.

[0182] In this application, if there are multiple steps involved in the method flow, unless there is a clear different description in this document, there is no strict order restriction for the execution of these steps, and they can be executed in other orders than described. Moreover, any step can include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn, alternating or simultaneously with other steps or parts of sub-steps or stages of other steps.

[0183] Reagents for detecting predictive factors and their applications

[0184] The term "predictor" in this application refers to a parameter or indicator that can provide useful information for prediction, which can be the content or expression level of a marker, an analytical parameter (such as a ratio), etc., which can be used as a variable in a mathematical formula, algorithm, or decision support tool for predicting an outcome. A mathematical formula, algorithm, or decision support tool can use multiple predictors to predict an outcome.

[0185] The predictive factors in the present application include, but are not limited to, the content of exhaustion markers of functional CD8 T cells in the sample or their expression levels and the content of inflammatory macrophages in the sample.

[0186] The predictive factors in the present application include but are not limited to the content of exhaustion markers of functional CD8 T cells in the sample or their expression levels and the content of immune checkpoint molecules of C1QC+ macrophages in the sample.

[0187] In some embodiments of the present application, a reagent for detecting a predictive factor is provided, and the reagent for detecting a predictive factor comprises: a reagent 1 for detecting the content or expression level of an exhaustion marker of functional CD8 T cells in a sample and a reagent 2 for detecting the content of inflammatory macrophages in a sample.

[0188] In some embodiments, the reagents for detecting predictive factors include: reagent 1 for detecting the content of exhaustion markers or their expression levels of functional CD8 T cells in a sample and reagent 2 for detecting the content of immune checkpoint molecules of C1QC+ macrophages in a sample.

[0189] In some embodiments of the present application, a reagent for detecting a predictive factor is provided, wherein the reagent for detecting a predictive factor comprises: a reagent 1 for detecting the content or expression level of an exhaustion marker of functional CD8 T cells in a sample and a C1QC + Macrophage immune checkpoint molecule assay reagent 2.

[0190] In the present application, unless otherwise specified, detection reagents, such as reagents for detecting predictive factors, reagent 1 and reagent 2, can each independently be a single product or a product combination, and can be packaged as a single preparation or multiple preparations, can be ready-to-use, and can also allow related substances to be packaged separately and mixed as needed before use.

[0191] In some embodiments, the reagent 1 is one or more of a specific primer, a specific probe, and an antibody for detecting the content or expression level of the depletion marker. In some embodiments, the reagent 2 includes a C1QC + 、IDO1 + One or more of the specific primers, specific probes and antibodies for detecting the IDO1 content or expression level of macrophages. In some embodiments, the reagent 2 includes a specific primer for detecting S100A9 in the fibrosis area. + One or more of specific primers, specific probes and antibodies for detecting the S100A9 content or expression level of macrophages.

[0192] In some embodiments, the reagent 1 is one or more of a specific primer, a specific probe, and an antibody for detecting the content or expression level of the depletion marker. In some embodiments, the reagent 2 includes a C1QC +One or more of specific primers, specific probes and antibodies for the content of immune checkpoint molecules in macrophages or their expression levels.

[0193] Unless otherwise specified, the term "primer" in this application refers to a nucleic acid sequence that can form a base pair complementary to a template strand and acts as a starting point for replicating the template strand, for example, its length can be 7-50 bases. Primers are usually synthesized, but naturally occurring nucleic acids can also be used. The sequence of the primer does not necessarily need to be exactly the same as the sequence of the template, as long as it is sufficiently complementary and can hybridize with the template.

[0194] Unless otherwise specified, the term "probe" in this application refers to a nucleic acid fragment as short as a few to as long as hundreds of bases, such as RNA or DNA, which can establish specific binding with mRNA and can determine the presence of specific mRNA due to labeling. The probe can be prepared in the form of oligonucleotide probes, single-stranded DNA probes, double-stranded DNA probes, and RNA probes.

[0195] Unless otherwise specified, the term "antibody" in this application refers to a specific immunoglobulin for an antigenic site. The antibody in this application refers to an antibody that specifically binds to the TNFR1 protein described in this application, and the antibody can be prepared according to conventional methods in the art. The form of the antibody includes polyclonal antibodies or monoclonal antibodies, antibody fragments (such as Fab, Fab', F(ab')2 and Fv fragments), single-chain Fv (scFv) antibodies, multispecific antibodies (such as bispecific antibodies), monospecific antibodies, monovalent antibodies, chimeric antibodies, humanized antibodies, human antibodies, fusion proteins containing the antigen binding site of the antibody, and any other modified immunoglobulin molecules containing the antigen binding site, as long as the antibody exhibits the desired biological binding activity.

[0196] C1QC can be detected using real-time fluorescence quantitative reverse transcription polymerase chain reaction (RT-qPCR), western blotting, RNA sequencing, immunohistochemistry, flow cytometry, and in situ hybridization. + IDO1 content or expression level of macrophages.

[0197] Unless otherwise specified, the term "RNA sequencing" in this application refers to the determination of the base sequence of RNA through sequencing technology (such as single molecule sequencing technology, etc.).

[0198] Unless otherwise specified, the term "Western blotting" in this application refers to a method for detecting TNFR1 in a sample based on the specific binding of an antigen and an antibody.

[0199] Unless otherwise specified, the term "exhaustion marker" in this application refers to a marker expressed during T cell exhaustion, which is usually manifested as T cell functional defects, such as severe functional defects and loss of proliferation ability, etc. The exhaustion marker in this application may be a conventional exhaustion marker in the art, illustratively, including but not limited to at least one of PD-1, TIM3, CD38, CD39, CD73, TIGIT, VISTA, GITR, LAG3 and B7H4.

[0200] Unless otherwise specified, the term "immune checkpoint molecule" in this application refers to a molecule that is associated with T cells and regulates T cell responses. Exemplarily, the immune checkpoint molecules in this application include, but are not limited to, one or more of IDO1, PD-L1, PD-L2, CD47, and B7H4.

[0201] Unless otherwise specified, the functional CD8 T cells in this application include but are not limited to CD45RO + CD8 + T cells.

[0202] In some embodiments, the inflammatory macrophages are selected from C1QC + 、IDO1 + The group consisted of macrophages with S100A9+ and macrophages with S100A9+ in the fibrotic area.

[0203] In some embodiments, detecting the amount of exhaustion markers of functional CD8 T cells or their expression levels in the sample is performed before intervention.

[0204] In some embodiments, the method for predicting the intervention effect of esophageal squamous cell carcinoma patients further comprises the following characteristic analysis before intervention: CD4 + Analysis of T / B cell ratio and CD8 in tertiary lymphoid structures + Analysis of T cell and B cell levels.

[0205] In some embodiments, the sample source of the predictor is tumor tissue from a patient with esophageal squamous cell carcinoma.

[0206] Without wishing to be bound by any theory, it is believed that CD45RO + CD8 + Expression of exhaustion markers on T cells and CD4 at the epithelial-immune interface + T / B cell ratio, these characteristics were significantly higher in the good responder group.

[0207] For example, key predictors of good treatment outcome include pre-treatment CD45RO + CD8+ Expression of PD-1 on T cells, optimal CD4+ T / B cell ratio, and increase in CD8+ T cells and B cells in TLS after treatment.

[0208] For example, key predictors of good treatment outcome include pre-treatment CD45RO + CD8 + Expression of PD-1 on T cells, CD4 + T / B cell ratio and CD4 + T / B PD-1 expression.

[0209] In some embodiments, the method for predicting the intervention effect in patients with esophageal squamous cell carcinoma further comprises the following characteristic analysis after intervention: frequency analysis of CD8+ T cells and B cells.

[0210] Without wishing to be bound by any theory, it is believed that C1QC + The expression of immune checkpoint molecules in macrophages suggests that the intervention effect or therapeutic efficacy may be negative.

[0211] Exemplarily, the epithelial-fibroblast region C1QC + IDO1 expression in macrophages was higher in the less effective group, suggesting that this macrophage marker is a negative predictor of treatment efficacy.

[0212] In some embodiments of the present application, there is provided a use of the reagent for detecting the predictive factor in preparing a product for predicting the intervention effect on patients with esophageal squamous cell carcinoma.

[0213] If there is no special explanation, the esophageal squamous cell carcinoma patient in the present application is any patient suffering from any type of esophageal squamous cell carcinoma, and illustratively, may be a patient with locally advanced esophageal squamous cell carcinoma.

[0214] In some embodiments, the esophageal squamous cell carcinoma patient has undergone combined intervention with a PD-1 inhibitor, a taxane compound, and a fluoropyrimidine derivative.

[0215] Unless otherwise specified, the term "PD-1" in this application refers to programmed death receptor 1, a protein present on the surface of T cells. Cancer cells sometimes use the PD-1 pathway to escape the attack of the immune system. The function of PD-1 inhibitors is to restore the activity of T cells and enhance the immune system's ability to attack cancer cells by blocking the binding of PD-1 to its ligands (PD-L1 and PD-L2). PD-1 inhibitors are widely used in the treatment of various cancers, including melanoma, non-small cell lung cancer, kidney cancer, bladder cancer, head and neck cancer, etc. Common PD-1 inhibitors include but are not limited to nivolumab, tislelizumab injection (Tislelizumab), toripalimab (Toripalimab), durvalumab (durvalumab), avelumab (avelumab), atezolizumab (Atezolizumab), etc.

[0216] In the present application, the PD-1 inhibitor may be a common PD-1 inhibitor in the art, exemplarily Toripalimab.

[0217] Unless otherwise specified, the term "fluoropyrimidine derivatives" in this application refers to a class of compounds formed by introducing fluorine atoms into the molecule with pyrimidine as the basic structure. Exemplary fluoropyrimidine derivatives include, but are not limited to, 5-fluorouracil (5-FU), capecitabine, S-1, and the like.

[0218] The fluoropyrimidine derivative described in the present application may be S-1. S-1 is mainly composed of tegafur (FT), CDHP and oteracil potassium (Oxo).

[0219] The combined intervention of the PD-1 inhibitor, the taxane compound and the fluoropyrimidine derivative comprises administering to the subject a therapeutically effective amount of a drug combination, wherein the drug combination comprises the PD-1 inhibitor, the taxane compound and the fluoropyrimidine derivative.

[0220] The paclitaxel compound in the present application may be nano-albumin-bound paclitaxel.

[0221] The reagents for detecting predictive factors described in the present application may also be in the form of a kit.

[0222] In some embodiments of the present application, a kit for predicting the intervention effect of patients with esophageal squamous cell carcinoma is provided, which comprises a reagent 1 for detecting the content or expression level of an exhaustion marker of functional CD8 T cells in a sample and a reagent 2 for detecting the content of inflammatory macrophages in the sample.

[0223] In some embodiments of the present application, a kit for predicting the intervention effect of patients with esophageal squamous cell carcinoma is provided, which comprises a reagent 1 for detecting the content or expression level of a functional CD8 T cell exhaustion marker in a sample and a C1QC + Reagent 2 for measuring the expression level of immune checkpoint molecules in macrophages.

[0224] The reagents for detecting predictive factors described in the present application can be used directly or indirectly for predicting and / or diagnosing the intervention effects of patients with esophageal squamous cell carcinoma.

[0225] In some embodiments of the present application, a method for predicting and / or diagnosing the effect of intervention on patients with esophageal squamous cell carcinoma is provided, comprising: detecting predictive factors before intervention.

[0226] The predictive analysis method may be a non-diagnostic and non-therapeutic method, or may be used for diagnosis and / or treatment.

[0227] In some embodiments, the predictor also includes CD4 + T / B cell ratio and CD8 in tertiary lymphoid structures + T cell and B cell levels.

[0228] In some embodiments, the predictive analysis method further comprises an analysis of indicators after intervention, wherein the indicators after intervention include CD8 + The frequencies of T cells and B cells.

[0229] In some embodiments, the sample source of the predictor can be found in the description in the context.

[0230] In some embodiments, the detection of the predictive factor can be performed using the aforementioned reagents for detecting the predictive factor. The reagents for detecting the predictive factor can refer to the description in the above context, but are not limited thereto.

[0231] Treating esophageal squamous cell carcinoma

[0232] Some embodiments of the present application also provide a method for treating esophageal squamous cell carcinoma, the method comprising administering a drug combination to a subject in need thereof, the drug combination comprising a PD-1 inhibitor, a taxane compound, and a fluoropyrimidine derivative.

[0233] Among them, the PD-1 inhibitor, the paclitaxel compound and the fluoropyrimidine derivative are not limited in the form of administration, for example, they can be administered sequentially, separately or simultaneously. When administered alone, the dosage or precautions of each drug can refer to its instructions. It is also possible to choose from the following modes of administration, one or more of which can be selected: oral, injection, implantation, spray and / or inhalation; for example, parenteral, subcutaneous, intraperitoneal, intravenous, intralesional, intracerebral, intraarterial interstitial infusion and / or implanted delivery device.

[0234] The combination of the drug combination can be in the form of a single preparation or multiple preparations, can be ready-to-use, and can also allow the related substances to be packaged separately and mixed as needed before use. For example, it can be in the form of a pharmaceutical composition.

[0235] When it is a pharmaceutical composition, the pharmaceutical composition may optionally contain a pharmaceutically acceptable carrier and / or excipient. In some embodiments, the dosage form of the pharmaceutical combination is selected from one or more of the following: injection, oral solution, powder, tablet, granule, capsule, syrup, decoction, sustained-release preparation, enteric solvent, aerosol or suspension.

[0236] In some embodiments, the subject in need thereof is a patient with locally advanced esophageal squamous cell carcinoma.

[0237] Some embodiments of the present application also provide the use of a PD-1 inhibitor, nanoalbumin-bound paclitaxel, and a fluoropyrimidine derivative in the preparation of a drug for treating esophageal squamous cell carcinoma.

[0238] Wherein, the administration method and dosage form of the drug can refer to the above.

[0239] Some examples are provided below.

[0240] The embodiments of the present application will be described in detail below in conjunction with examples. It should be understood that these examples are only used to illustrate the present application and are not intended to limit the scope of the present application. The experimental methods for which the conditions are not specified in the following examples are preferably referred to the guidance given in the present application, and can also be based on the experimental manual or normal conditions in the art, or can also be based on the conditions recommended by the manufacturer, or with reference to experimental methods known in the art.

[0241] This study followed the ethical guidelines established by the ethics committee of a specific hospital. All participants signed written informed consent. In addition, this study was registered in the Clinical Trials Register.

[0242] CAP / NCCN Assessment

[0243] These guidelines recommend that the effectiveness of treatment be determined by pathological evaluation, which measures the proportion of tumor cells that survive treatment. After treatment, tumor shrinkage is systematically classified into several categories: complete response (no residual tumor, 0), partial response (PR) (residual tumor but significantly reduced, 1), stable disease (SD) (no significant change in tumor burden, 2), and progressive disease (PD) (increase in tumor size or number, 3). This classification helps standardize the evaluation and recording of treatment outcomes. Figure 1C Representative CT images showing treatment outcomes in patients with good (grades 0 and 1) and poor (grades 2 and 3) responses to treatment are shown.

[0244] IMC slice preparation

[0245] Sections were cut from formalin-fixed paraffin-embedded (FFPE) tissue samples from patients with hepatocellular carcinoma (HCC) using the HistoCore MULTICUT system. These sections were heated at 68°C for one hour to facilitate dewaxing. Subsequently, the sections were immersed in xylene twice at 68°C for 10 minutes each time to completely dewax. Rehydration was performed by sequentially immersing the sections in different concentrations of ethanol solutions (95%, 85%, and 75%) for 5 minutes each at room temperature. Antigen retrieval was performed by heating the sections in sodium citrate buffer at 100°C for 30 minutes. After cooling to room temperature, the sections were washed twice in PBS-TB (PBS containing 0.5% Tween-20 and 1% bovine serum albumin) for 5 minutes each. Blocking was performed using SuperBlock (37515, Thermo Fisher Scientific) for 30 minutes at room temperature. The sections were then washed three times in PBS-TB and then incubated with a custom antibody cocktail overnight at 4°C. Antibodies were conjugated to heavy metals using the Maxpar® X8 Antibody Labeling Kit (Fluidigm), as detailed in Table 1. After incubation, sections were washed three times in PBS-TB. Nuclear labeling was performed by labeling with Intercalator-Ir (201192B, Fluidigm) solution (1.25 µM) in PBS-TB for 30 min at room temperature, followed by two washes in PBS-TB and a final rinse with ddH2O. Group 0-1 included 7 samples, each of which had two regions of interest (ROIs) evaluated before and after the intervention. Group 2-3 included 10 samples; 9 of these samples had 2 ROIs evaluated each before treatment, and one sample had 4 ROIs evaluated. In the post-treatment period, there were 9 samples, each of which had 2 ROIs evaluated.

[0246] Table 1

[0247]

[0248] IMC single cell segmentation

[0249] Cell segmentation of IMC images uses a pre-trained TissueNet algorithm that has been described in the literature. The algorithm processes data from two different imaging channels. The nuclear channel is usually stained with DAPI to delineate the cell nucleus, and the second channel is chosen to highlight the cell membrane to mark the boundaries of individual cells.

[0250] IMC pretreatment and cell type identification

[0251] Initially, marker expression data were transformed using the arcsine function and truncated at the 1st and 99th percentiles to define the range. Normalization for each channel was performed based on the maximum and minimum values. To correct for batch effects, we used the Harmony algorithm in the R package (version 0.1.0). Cell clustering was subsequently performed using Rphenograph (version 0.99.1) using a parameter of 30 nearest neighbors. Initial clustering used lineage-specific markers including CD14, CD16, CD4, CD8, collagen I, CD31, E-cadherin, CD45, CD163, CD68, CD11b, CD20, CD11c, CD15, CD3, FAP, CD57, pan-cytokeratin, and αSMA to distinguish the major cell populations. The average expression of each marker in each cluster was displayed on a heat map. This expression pattern was then used for further annotation. Then, a secondary, finer clustering involved additional markers to refine lymphocytes (including FoxP3, CD4, CD8, CXCL13, CD127, CD45, CD38, CCR7, CD20, granzyme B, PD-1, Ki-67, CD45RA, CD3, CD45RO, CD57, and HIF1a), epithelial cells (including S100A9, HLAI, E-cadherin, IDO1, PD-L1, Ki-67, pan-cytokeratin, HIF1a), and myeloid cells (including S100A9, CD16, CD14, CD45, IDO1, PD-L1, Ki-67, CD68, HIF1a, CD11b, CD11c, SPP1, HLA-DR, CD208, C1QC, and CD163). Each subcluster was separated from the total population and re-clustered using the same approach. The resulting clusters are visualized in a heatmap to aid in cell type annotation.

[0252] TLS and environment detection

[0253] PatchDetection in the imcRtools package was used to detect the tissue structure formed by local B cell aggregation. The B cells in TLS analysis include CD8T&B cells, B cells, CD4T&B cells, CXCL13 +B cells and proliferating B cells. The maximum intercellular distance between B cells clustered in the TLS was defined as 20 µm, and the minimum number of cells in the TLS was 30. The TLS boundaries were defined by concave hulls. The environment was defined as the cell area located by extending the area boundaries by 20 µm.

[0254] Area and environment identification

[0255] The region identification adopts a strategy similar to that of TLS detection. Three main regions are defined: epithelial region including EMT, epithelial, proliferating epithelial, and IDO1 + S100A9 + Epithelial, IDO1 + Epithelial, S100A9 + Epithelial and PD-L1 + Epithelial. Fibrotic areas include FAP + Fibroblasts, αSMA + Fibroblasts and collagen I + Fibroblasts. Immune regions include all lymphocytes and myeloid cells. Regions were defined as areas containing at least 20 cells with a maximum inter-cell distance of 20 µm. Adjacent regions were merged together for downstream analysis. Regions were then expanded by 20 µm to form corresponding environmental regions, which is approximately equivalent to two cell diameters.

[0256] Computation of cell-cell interactions

[0257] Cell-to-cell interactions of each cell were captured by capturing a window of its ten nearest neighbors. These windows are represented as a composition matrix containing all defined cell types. The interaction strength of the central cell of each region of interest (ROI) with its neighbors was calculated by averaging all central cells. To identify statistically significant differences, comparisons between selected groups were performed using a two-tailed Student's t-test.

[0258] Multivariate modeling and variable importance

[0259] Four different sets of features were derived for each ROI, covering cell cluster frequency, regional characteristics, cell-cell interactions, and functional marker expression. In total, 8,401 features were derived for each ROI. Only features with values ​​greater than or equal to 0 in more than 10 ROIs were retained. We further reduced the feature space by identifying highly correlated variables (Pearson correlation coefficient > 0.9) and randomly selecting a representative variable. We used an established feature selection algorithm (implemented in the R package Boruta) to identify the most important features for treatment response prediction. This method repeatedly compares the true values ​​to randomly shuffled features to identify which features show importance more frequently than random data. This process was repeated 10,000 times to generate a binomial distribution of feature importance, and the final significant variables were identified based on a threshold of P < 0.01.

[0260] The AUC statistic was used to evaluate the predictive efficiency of different models. The dataset was randomly divided into a training set (70%) and a test set (30%). Four commonly used machine learning models, including random forest, generalized linear model, neural network, and shrinkage discriminant analysis model, were applied to the training set (model parameters were optimized using cross-validation). Predictions were made using independent test data. For each feature set, the process was repeated 100 times to calculate the average AUC. Additional features were added step by step, starting with the top three most important features.

[0261] Single-cell RNA sequencing data processing

[0262] The scRNA-seq dataset with accession number GSE145370 was downloaded from the GEO database. The UMI count matrix was processed using the Seuratv4.0.1 R package to create a Seurat object. Downstream analysis focused on qualified cells with gene counts between 200 and 6000 and a mitochondrial gene ratio of less than 25%.

[0263] To prepare the data for analysis, normalization was performed using the “NormalizeData” function, and the top 2000 highly variable genes were identified using the “FindVariableFeatures” function. To eliminate batch effects, the normalized data were integrated using the “FindIntegrationAnchors” and “IntegratedData” functions. The integrated data were then scaled and centered using the “ScaleData” function, and principal component analysis (PCA) was performed using the “RunPCA” function, selecting the top 20 principal components.

[0264] For visualization and clustering, t-SNE and graph-based clustering were performed using selected principal components. Cell types were manually identified based on the expression patterns of signature genes.

[0265] RNA-seq data of ESCC and corresponding clinical information were downloaded from the TCGA database. To assess the cell type characteristics in each sample, GSVA analysis (gene set variation analysis) was performed on TCGA samples using cluster-specific gene sets derived from scRNA-seq data. This analysis generated a cell type characteristic score for each sample, ranging from -1 to +1.

[0266] To evaluate the relationship between GSVA scores and survival outcomes, Kaplan-Meier survival curves were generated using the survival and survminer packages in R. GSVA scores were used as predictor variables and patient survival time was used as outcome variables. The log-rank test was used to determine the statistically significant differences between the survival curves. A significance level of P < 0.05 was used to determine whether the differences in the survival curves were statistically significant.

[0267] Statistical analysis

[0268] The sample size for the clinical study was carefully calculated based on data from a pilot study examining similar end points. We used these preliminary data to estimate the expected effect size for key outcomes, particularly differences in the tumor immune microenvironment after combining neoadjuvant chemotherapy and immunotherapy. To determine the required sample size, we performed a power analysis ensuring 80% power and a significance level of 0.05, which is standard for clinical studies, to detect clinically meaningful differences. The effect size was determined based on differences in immune cell infiltration and immune checkpoint expression levels observed in the pilot study. Using G Power software, we input these parameters and the estimated effect size to calculate the minimum necessary sample size. In addition, we accounted for potential withdrawals and non-compliance by increasing the sample size by 20%. This adjustment was designed to maintain the integrity and statistical validity of the study results under realistic conditions. Statistical tests were selected based on applicable data distribution and variation characteristics. Two-tailed t-tests were used to identify differences in cell proportions and functional marker expression between the two groups. P values ​​< 0.05 indicated statistical significance.

[0269] Example 1: Neoadjuvant chemoimmunotherapy enhances tumor response in esophageal squamous cell carcinoma (ESCC): clinical trial results

[0270] The primary objective of this study was to evaluate the efficacy and safety of neoadjuvant chemoimmunotherapy in patients with ESCC, with a focus on surgical outcomes and molecular characteristics. During the trial, a total of 67 patients were evaluated for eligibility. 60 patients met the inclusion criteria and received neoadjuvant therapy ( Figure 1A). During treatment, one patient developed severe atrial fibrillation, which required a change in treatment regimen. In addition, three patients chose not to undergo surgery. After neoadjuvant therapy, 56 patients underwent surgery, and 55 achieved complete tumor resection (R0 resection) with no microscopic residual disease. Biopsy samples collected before and after treatment were comprehensively analyzed by IMC, targeting 39 specific proteins ( Figure 1B and Table 1).

[0271] Patient characteristics

[0272] All patients (n=55)

[0273] Age at diagnosis, years: 60.4 (45-75) Note: 60.4 is the mean age. The age range is in brackets

[0274] Male: 48 (87.27%) Note: 48 is the number, and the percentage in brackets is the proportion of all patients.

[0275] ECOG Performance Status:

[0276] 0:31 (56.36%)

[0277] 1:24 (43.64%)

[0278] Stooler's Dysphagia Score:

[0279] 0:12 (21.83%)

[0280] 1:19 (34.55%)

[0281] 2:21 (38.17%)

[0282] 3:3 (5.45%)

[0283] Histology:

[0284] Squamous cell carcinoma: 55 (100%)

[0285] Overall clinical stage:

[0286] I: 1 (1.82%)

[0287] II: 7 (12.73%)

[0288] III: 27 (49.09%)

[0289] IV: 20 (36.36%)

[0290] Preoperative T status:

[0291] cT1: 3 (5.45%)

[0292] cT2: 9 (16.36%)

[0293] cT3: 24 (43.64%)

[0294] cT4: 19 (34.55%)

[0295] Preoperative N status:

[0296] cN0: 8 (14.55%)

[0297] cN1: 25 (45.45%)

[0298] cN2: 17 (30.91%)

[0299] cN3: 5 (9.09%)

[0300] Preoperative M status:

[0301] cM0: 55 (100%)

[0302] Tumor grade (differentiation):

[0303] Good: 5 (9.09%)

[0304] Good-moderate: 2 (3.64%)

[0305] Moderate: 31 (56.36%)

[0306] ·Medium-Poor: 4 (7.27%)

[0307] Poor: 13 (23.64%)

[0308] Tumor location:

[0309] Central: 16 (29.09%)

[0310] ·Middle and lower part: 21 (38.17%)

[0311] ·Lower: 18 (32.73%)

[0312] Tumor response after neoadjuvant therapy was evaluated according to the CAP / NCCN guidelines. The guidelines recommend that the effectiveness of treatment be determined by pathological evaluation, which assesses the proportion of tumor cells that survive after treatment. After treatment, tumor shrinkage was systematically classified into several categories: complete response (no residual tumor, 0), partial response (PR) (there is residual tumor but it is significantly reduced, 1), stable disease (SD) (no significant change in tumor burden, 2), and progressive disease (PD) (increase in tumor size or number, 3). This classification helps standardize the evaluation and recording of treatment outcomes. Figure 1CRepresentative CT images of patients with good (grades 0 and 1) and poor (grades 2 and 3) responses to treatment are shown, providing visual insights into the various treatment outcomes observed. In addition, the baseline demographic and clinical characteristics of the patients are summarized below. The median age of the cohort was 60.4 years, with an age range of 45 to 75 years. The majority of patients were male (87.27%), primarily diagnosed in stage III-IV, without metastases but with significant lymph node involvement (85.45%). According to RECIST 1.1 criteria, treatment outcomes were as follows: 60.0% of patients had partial response (PR), 36.36% had stable disease (SD), and 3.64% had progressive disease (PD), resulting in an overall disease control rate of 96.36% for patients, as detailed in Figure 1C and Table 2. A significant portion of the cohort experienced significant tumor regression after treatment: 52.73% (n=29) achieved pathological response (MPR), 29.09% (n=16) achieved complete pathological response (pCR), and 70.91% (n=39) showed at least 50% pathological regression, see Figure 1D .

[0313] The baseline clinical characteristics of the study population were as follows:

[0314] Summary of treatment effects

[0315] All patients (n=55)

[0316] Radiological reactions:

[0317] Partial response (PR): 33 (60.00%)

[0318] Stable disease (SD): 20 (36.36%)

[0319] Progressive disease (PD): 2 (3.64%)

[0320] Pathological remission:

[0321] Yes: 29 (52.73%)

[0322] No: 26 (47.27%)

[0323] Complete pathological response:

[0324] Yes: 16 (29.09%)

[0325] No: 39 (70.91%)

[0326] Pathological reduction ≥50%:

[0327] Yes: 39 (70.91%)

[0328] No: 16 (29.09%)

[0329] CAP / NCCN narrowed categories:

[0330] 0:15 (27.27%)

[0331] 1:14 (25.45%)

[0332] 2:13 (23.64%)

[0333] 3:13 (23.64%).

[0334] Table 2: Treatment effects of all patients

[0335]

[0336] Table 3: Overview of adverse events

[0337]

[0338] When patients were stratified by clinical outcome and clinical characteristics of each group were analyzed, no significant enrichment of specific clinical characteristics was observed between response groups. Treatment-related adverse events (AEs) are detailed in Table 3. Major AEs included alopecia (93.33%), muscle aches (73.33%), fatigue (58.33%), anemia, leukopenia / neutropenia, and peripheral sensory neuropathy (51.67% each). Eleven (18.33%) serious adverse events (SAEs) were reported, of which 10 (16.67%) were grade 3 and 1 (1.67%) was grade 4. The most common serious adverse events were leukopenia and neutropenia. No treatment-related deaths were reported (Table 3). Clearly, the majority of these effects can be attributed to the chemotherapy component of the combination therapy. However, side effects typical of immunotherapy, such as rash, were observed in 33.67% of patients. Previous studies support the view that although immune activation is effective against tumor cells, it may also induce collateral tissue reactions, leading to adverse events. However, in the current trial, immune-related adverse events were not associated with a favorable response ( Fig. 22 In addition, the Kaplan-Meier curves show the progression-free survival (PFS) and overall survival (OS) of 55 patients ( Figure 1E and Figure 1F ).

[0339] In conclusion, the results of this study highlight the potential of neoadjuvant chemoimmunotherapy to improve surgical and molecular outcomes in patients with ESCC, in line with the goal of evaluating the efficacy and safety of this therapeutic approach.

[0340] Example 2: Comprehensive characterization of the spatial characteristics of the ESCC immune microenvironment at single-cell resolution

[0341] To dissect the spatial heterogeneity of the ESCC immune microenvironment while preserving its architecture, a 40-marker IMC antibody panel (39 protein markers and DNA) was established for ESCC samples ( Figure 2A and Figure 2B The 40-marker IMC antibody panel covers markers for epithelial, endothelial, stromal, and various immune cells; the chemokine CXCL13; the proliferation marker Ki-67; and markers associated with immune checkpoints, including PD-1, programmed death ligand 1 (PD-L1), and IDO1 (Table 1). Each antibody was validated by immunohistochemistry (IHC) before binding to heavy metals, and each metal-binding antibody was further validated by IMC ( Figure 2C ).

[0342] In the spatial and cellular profiling of the ESCC immune microenvironment, multiple cellular components were identified, including pan-keratin + (Pan-CK + ) Epithelial cells, CD31 + Endothelial cells, αSMA + Myofibroblasts, collagen I + Fibroblasts and FAPs + Fibroblasts. Immune cell landscape including CD11b + 、CD15 + Granulocytes, CD14+ monocytes, and CD68 + Macrophages, and cells containing CD20 + B cells and CD3 + T cell lymphocytes ( Figure 2D and Figure 2E ).

[0343] In the epithelial cell population, pan-CK was further observed + E-Cad + Epithelial cell subpopulations, indicating that the subpopulation is in epithelial-mesenchymal transition (EMT), as well as other phenotypic subpopulations such as Ki67 + Proliferation, IDO1 + 、IDO1 + S100A9 + and PD-L1 + Epithelial cells, respectively, reflect the proliferation, inflammation, and immune escape status of their corresponding cell subsets. IDO1 and S100A9, two inflammatory markers, and PD-L1, a key immunosuppressive molecule, are widely documented in the literature.

[0344] Further subdivision of lymphocyte subsets revealed B cell, T cell, and natural killer (NK) cell subsets. Specifically, B cells include CD20+ Cells, CXCL13 + B cells (associated with recruitment of T cells and B cells), and Ki67 for proliferation + B cells. T cells are expressed by CD45RO + CD4 + Activated T cells, Foxp3 + Tregs and CD38 + Foxp3 + Tregs. In addition, granzyme B was also identified + Effector CD8 + T cells, further indicating functional specialization within the T cell population. Notably, the spatial proximity of different cell subsets suggests potential microenvironmental interactions that complicate the cell segmentation process.

[0345] The macrophage population is very diverse and includes CD163 + Macrophages, CD11b + Macrophages, C1QC + Macrophages (indicates resident macrophage population), PD-L1 + Macrophages, CD11c + HLA-DR + Antigen presenting macrophages, SPP1 + and IDO + SPP1 + Macrophages and LAMP3 + Dendritic cells (DCs) and proliferation of Ki67 + Macrophages. The expression of SPP1 and IDO1 suggests their roles in immune regulation and inflammation, whereas LAMP3 marks mature DCs, essential for efficient antigen presentation.

[0346] In the study, the cell distribution in different treatment response groups was analyzed in detail, e.g. Figure 2F The differences before and after treatment were evaluated, and the results were further stratified into good response (CAP / NCCN grade 0-1) and poor response (CAP / NCCN grade 2-3) categories.

[0347] In the epithelial cell population: Post-treatment analysis showed that most epithelial cell subsets were significantly reduced in the good response group (0-1 after treatment vs. 0-1 before treatment), which is consistent with expectations because chemotherapy and immunotherapy mainly target tumor epithelial cells. In contrast, in the poor response group, only a small number (3 / 7) of epithelial cell subsets showed a decrease, especially proliferative and inflammatory epithelial cells ( Figure 2G Interestingly, before treatment, S100A9 +The levels of epithelial cells were higher in the good responder group (0-1 before treatment vs. 2-3 before treatment), which may indicate that patients with inflammatory epithelial cells have a better prognosis when receiving adjuvant therapy.

[0348] In the fibroblast population: the only differential subpopulation was αSMA + Myofibroblasts, which are increased in patients who show a favorable response after adjuvant therapy, suggest a potential role for myofibroblasts in mediating favorable treatment outcomes.

[0349] Among the lymphocyte populations, most lymphocyte subsets did not change significantly between the good and poor responders at the pre- and post-treatment time points. The only exception was CD8 + The proportion of T&B cells was significantly increased compared with the poor group, which may indicate specialized immune features such as tertiary lymphoid structures (TLS). In contrast, lymphocyte subsets in both response groups generally increased after treatment compared with those before treatment. In particular, B cells, proliferating B cells, and CXCL13 + B cells increased only after treatment in the good responder group (0-1 after treatment vs. 0-1 before treatment), emphasizing the importance of these specific B cell subsets in mediating effective treatment responses. Tregs increased after treatment in the poor responder group (2-3 after treatment vs. 2-3 before treatment), possibly leading to immunosuppression.

[0350] Among myeloid cell populations: Although many myeloid cell subsets decreased in both responder groups after treatment, this trend was less pronounced in the poor responder group. In particular, four myeloid cell subsets, including granulocytes and macrophages, decreased after treatment in the good responder group (0-1 after treatment vs. 0-1 before treatment), while only granulocytes decreased after treatment in the poor responder group (2-3 after treatment vs. 2-3 before treatment). This pattern suggests that a reduction in myeloid cell suppression may be necessary to achieve a good treatment response.

[0351] These findings highlight the complex interplay of various cell populations in determining cancer treatment outcomes and underscore the importance of personalized treatment strategies based on detailed cellular profiling.

[0352] Example 3: Association of cellular composition, functional markers and surrounding environment of tertiary lymphoid structures with clinical trial results

[0353] TLS, characterized by dense clusters of B and T cells, have been widely reported to play a key role in treatment response. To investigate this, we used a previously developed region recognition algorithm designed to identify tissue structures where B cells aggregate ( Figure 3A ).

[0354] Post-treatment analysis showed that TLS size increased significantly in both response groups; however, no size differences were observed between the good and poor response groups ( Figure 3B In the evaluation of the cell composition within TLS, CD8 + T / B cells increased significantly after treatment in the good responder group, emphasizing their potential role in mediating favorable treatment outcomes. In contrast, in the poor responder group, higher levels of collagen I + Fibroblasts and PD-L1 + Macrophages suggest an inhibitory effect on therapeutic efficacy ( Figure 3C ).

[0355] These findings were also reflected in the analysis of the expression levels of functional markers in TLS. After treatment in the good response group (0-1 after treatment vs 0-1 before treatment), S100A9 expression in proliferating macrophages was reduced, indicating reduced inflammatory activity, which may contribute to the treatment outcome. In addition, after treatment in the poor response group (2-3 after treatment vs 2-3 before treatment), PD-1 and CD11b in lymphocyte subsets were expressed. + Significant upregulation of PD-L1 and LAMP3 in macrophages + Increased HIF-1α in dendritic cells (DCs) suggests that increased immune checkpoint activity and hypoxia responses may hinder effective adjuvant therapy ( Figure 4A and Figure 4B ).

[0356] The analysis was extended to the environment surrounding the TLS and a previously published method was used to define the extent of the neighborhood ( Figure 5 After treatment, the good response group (0-1 after treatment) showed higher CD45RO + CD8 + The proportion of T cells indicates that activated cytotoxic T cells are beneficial to the treatment outcome. The proportion of natural killer cells (NK cells) in the group that responded well to treatment increased moderately compared with that in the poor response group after treatment (2-3 days after treatment). NK cells are a key component of the anti-tumor immune response, especially in the innate immune system. Previous studies have shown that anti-PD-1 therapy can enhance the anti-tumor activity of NK cells. Therefore, the increase in the number of NK cells observed in the group that responded well to treatment in combination treatment with PD-1 antibodies is consistent with established immune mechanisms and supports the effectiveness of this treatment strategy in enhancing the innate immune response against tumors ( Figure 6). After treatment, no significant difference in Treg proportions was observed between the good (0-1 after treatment) and poor response groups (2-3 after treatment). However, a significant increase in Tregs from pre-treatment to post-treatment was observed in the poor response group (2-3 before treatment to 2-3 after treatment). These findings suggest that it is not just the presence of Tregs, but their increase in proportion after treatment that may weaken the efficacy of immunotherapy. This emphasizes the key role of Tregs in regulating immune responses during PD-1 targeted therapy and highlights strategies to modulate the dynamics of Tregs to improve treatment outcomes.

[0357] Please note that IMC analysis does not include markers for all DC subtypes, so only LAMP3-positive DCs were identified in TLS. Therefore, although no significant changes in LAMP3-positive DC content were observed ( Fig.23 ), but the possibility that other DC subsets show significant changes cannot be excluded.

[0358] In conclusion, differential dynamic alterations of TLS and their surroundings, particularly with respect to immune cell populations and functional markers, significantly impact the outcome of chemoimmunotherapy, highlighting the complex interplay between TLS and therapeutic efficacy.

[0359] Example 4: Association of regional tumor immune microenvironment characteristics with clinical trial results

[0360] ESCC can be roughly divided into three major regions based on its tissue structural characteristics, including epithelial, immune, and fibrotic regions. Previous studies have shown that these regions show different trends in treatment response. First, the regional range was defined using a previously developed algorithm ( Figure 7 ) and further identified overlapping regions, such as the epithelial-immune interface and the epithelial-fibrotic interface.

[0361] First, the areas of these regions were evaluated in the different response groups and in samples before and after treatment. Notably, only the epithelial area was significantly reduced after treatment in both groups ( Figure 8 The areas of other regions did not show significant changes between the groups. In the epithelial region, CD4 + T / B and CD4 + T / CD8 + After treatment, the frequency of immunosuppressive cell populations such as αSMA + and FAP + Fibroblasts, Tregs, and SPP1 + Macrophages were decreased in the good responder group (0-1 after treatment vs 0-1 before treatment), indicating a reduction in immunosuppressive cells.

[0362] In the immune area, tight interactions of T / B cells were more prevalent in the good responder group (0-1 after treatment vs 2-3 after treatment). Epithelial cell and macrophage subsets decreased in the responder group (0-1 after treatment vs 0-1 before treatment), while T cell subsets increased. Although T cell subsets increased in both responder groups after treatment, immunosuppressive cells such as fibroblasts, Tregs, and macrophages decreased only in the good responder group after treatment.

[0363] In the fibrotic areas, Tregs were higher in the good responders before treatment (pre-treatment 0-1 vs pre-treatment 2-3), while T cell subsets decreased after treatment (post-treatment 2-3 vs pre-treatment 2-3 and post-treatment 0-1 vs pre-treatment 0-1). These results were significantly different from those in the epithelial and immune areas, which may be attributed to the disordered structure of the fibrotic areas. There were fewer significant changes in cellular composition in the interaction between the epithelial-immune junction and the epithelial-fibrotic junction ( Fig. 9 ).

[0364] For the expression levels of functional markers, most of the significant changes occurred in the good response group before and after treatment (0-1 after treatment vs 0-1 before treatment, orange), indicating that these significant changes may be associated with effective treatment response. Notably, most of the changes in functional markers involved the downregulation of immune checkpoints and inflammatory markers, emphasizing their key role in successful treatment outcomes ( Fig.10 In contrast, upregulation of immune checkpoints in the adverse response group (2-3 days after treatment vs 2-3 days before treatment) may lead to poor treatment response ( Fig.10 ).

[0365] Analysis of changes in the cell composition of the surrounding environment in different regions found that in patients who showed good response after treatment (0-1 after treatment vs 0-1 before treatment), CD8 + T cell populations were significantly enriched, including CD38 + T cells and CD45RO + CD8 + In contrast, Tregs were mainly increased in patients who showed poor response after treatment (2-3 days after treatment vs 2-3 days before treatment) ( Fig.11 ).

[0366] Analysis of the expression levels of cellular components and function-related markers in different regions of the TME and its surroundings highlighted clear associations between spatially specific cellular changes and clinical outcomes.

[0367] Example 5: Importance of T cell infiltration and interactions in immune regions and epithelial-immune junctions for good clinical responses

[0368] To elucidate the impact of cell-cell interactions on clinical treatment efficacy, an analytical approach adapted from a previously established method was implemented. First, a “central” cell was identified, and the ten nearest neighbor cells around it were designated as neighboring cells, and the composition of these 10 cells was calculated to describe the cell-cell interactions ( Fig.12 ). At the same time, the analysis calculates the cell interactions in different regions separately to obtain a tissue region-specific cell interaction matrix.

[0369] In the immune-only area, a significant increase in the interaction between T cells and other cells was observed after treatment, reflecting an active immune response after treatment ( Fig.13 ). To further decode these complex interaction patterns, significantly upregulated cell-cell interactions in the immune region were ranked according to their fold change. In the good responder group (0-1 after treatment vs 0-1 before treatment, orange), T cell-cell interactions with other cells increased significantly, indicating enhanced T cell infiltration, which is a positive indicator of immune activation and potential therapeutic efficacy.

[0370] In contrast, interactions between T cells and fibroblasts were prevalent at the epithelial-immune interface, demonstrating cross-border dialogue between epithelial and immune cells in this mixed cellular environment ( Fig.14 and Fig.15 Similarly, in the epithelial-immune region, the significant interaction between epithelial cells and T cells in the good responder group further suggests a robust T cell infiltration and a potentially effective immune response ( Fig.16 ).

[0371] These findings highlight the critical role of cell-cell interactions in responses to clinical treatments.

[0372] Example 6: Identifying key spatial features for predicting therapeutic efficacy in TME through machine learning and constructing a predictive model for treatment response

[0373] Given the complexity of the TME, it is critical to discern which spatial features predict treatment response. A comprehensive feature set including cell cluster frequency, TLS, regional characteristics, cell interactions, and expression profiles of major functional markers was analyzed. The initial feature set included more than 8,000 features, including detailed metrics from cluster frequency, TLS characteristics, and regional interactions.

[0374] To optimize the analysis, features with too many zero values ​​in the sample were first removed. Subsequently, multicollinearity was handled by excluding highly correlated features, choosing to retain one representative feature from each relevant group. This reduced the feature set to approximately 3,000, which were further evaluated by a random forest (RF) model to determine their importance in predicting treatment outcomes, as shown in Fig.17 Focusing on pre-treatment characteristics, we identified three indicators that are most important for predicting treatment response: CD45RO at the epithelial-immune interface + CD8 + PD-1 expression in T cells and CD4 in epithelial areas and immune-epithelial junction areas + T&B PD-1 expression levels. Notably, these signatures were more highly expressed in the good responders (pre-treatment 0-1 vs pre-treatment 2-3), highlighting PD-1 expression in T and B cells at the epithelial-immune interface as a key determinant of good treatment response. In contrast, C1QC in the epithelial-fibrotic region + IDO1 expression in macrophages was higher in the poor responder group, suggesting that this macrophage marker is a negative predictor of treatment efficacy ( Fig.18 ).

[0375] Four different algorithms were used to build prediction models and predict treatment outcomes: random forest, generalized linear model, neural network, and shrinkage discriminant analysis. Each algorithm showed an area under the curve (AUC) value higher than 95% in different models, indicating strong performance when using selected features ( Fig.19 ). It is noteworthy that an AUC of 95% to 100% can be achieved using the first five features, involving six markers: Pan-CK (epithelial marker), CD45 (universal immune marker), PD-1, CD4, CD8, and CD45RO (T cell markers), and CD20 (B cell marker). ROC analysis of the first five features showed that the random forest model gave the highest ROC of 0.94 ( Fig. 20 ).

[0376] To elucidate the mechanisms associated with treatment response, the random forest method was also used to perform a detailed analysis of the post-treatment samples. In the group of patients who showed a good treatment response, CD8 + The proportion of T cells and B cells was higher in the poor response group, indicating that strong immune activation in these areas is associated with good treatment outcomes. In contrast, in patients who showed poor response, the significantly upregulated feature was PD-L1 in the fibrotic area. +Expression of S100A9 in macrophages. This phenomenon emphasizes the negative impact of inflammatory macrophages during treatment. S100A9 expression as a key feature in the adverse response group highlights the significant role of fibrotic area-associated inflammatory phenotype in treatment resistance ( Fig.21 ).

[0377] To further clarify S100A9 + To investigate the role of macrophages and extend the observations to a larger cohort of ESCC patients, single-cell RNA sequencing analysis was performed using the publicly available dataset GSE145370. This dataset includes 113,240 high-quality cells successfully isolated from seven surgically resected ESCC tumors and corresponding normal tissues, and identified ten major cell types including T cells, macrophages, B cells, NK cells, tissue stem cells, granulocytes, dendritic cells, mast cells, plasmacytoid dendritic cells, and endothelial cells. Sequencing features can be found in Fig.24A , Fig. 24B , Fig.24C , Fig.24D , Fig.24E , Fig.24F , Fig.25 and Fig.26 .

[0378] Further re-clustering of the macrophage population enabled the identification of six distinct macrophage subsets. These subsets included S100A9 + Macrophages, C1QA / B + Macrophages, CD206 + Macrophages, CTSK + Resident macrophages, SPP1 + Macrophages and proliferating macrophages.

[0379] In addition, single-cell datasets were used to calculate S100A9 + The top 20 genes highly expressed in macrophages were selected as a gene set, and the GSVA (gene set variation analysis) score of TCGA samples on this gene set was calculated. + Survival analysis using macrophage GSVA scores showed that S100A9 + Macrophage content was significantly negatively correlated with patient survival outcomes. This analysis not only revealed heterogeneity within the macrophage population, but also highlighted potential biomarkers and therapeutic targets for ESCC, enhancing the understanding of the tumor microenvironment and its impact on patient prognosis.

[0380] In conclusion, multi-omics analyses suggest that specific spatial features within the TME play a key role in treatment outcome. In particular, CD45RO + CD8 +PD-1 expression on T cells and CD4 + T cell and B cell content is a strong predictor and determinant of good treatment response. Conversely, IDO1 expression in C1QC+ macrophages in the epithelial / fibrotic interface is an important marker for poor treatment outcome. These findings highlight the critical role of spatial biomarkers in enhancing cancer treatment strategies.

[0381] in conclusion

[0382] This invention demonstrated the remarkable efficacy of a neoadjuvant chemoimmunotherapy regimen combining PD-1 blockade, nanoalbumin-bound paclitaxel, and S-1 in ESCC, with a high proportion of patients achieving pathological remission (MPR) and complete pathological remission (pCR). This regimen significantly improved the therapeutic effect and modulated the TME to overcome drug resistance. By combining spatial proteomic analysis of IMC and machine learning, key spatial features within the TME that predict treatment response and resistance were revealed. Key predictors of good treatment outcomes include pre-treatment CD45RO + CD8 + Expression of PD-1 on T cells, CD4 + Expression of PD-1 on T / B cells and CD4 + T / B cell ratio. In contrast, C1QC + IDO1 expression in macrophages and S100A9 expression in macrophages in fibrotic areas were markers of resistance. In addition, a robust prediction model for treatment response was developed using machine learning techniques, with an area under the curve (AUC) of 97%. These findings highlight the importance of spatial proteomics in understanding complex tumor dynamics and effectively tailoring personalized treatment strategies.

[0383] The spatial proteomic data obtained from this study provide insights into the determinants of clinical treatment efficacy in ESCC. Previous studies have shown the importance of TLS in ESCC treatment response and prognosis, but this study is the first to elucidate the impact of spatial characteristics of cells and dynamic changes in functional proteins within TLS on clinical efficacy in patients. For example, CD8 + The proportion of T cells and B cells was significantly upregulated, suggesting that these cells play a key role in the treatment response. + Fibroblasts and PD-L1 +The increase in macrophages suggests their inhibitory effect on treatment. Fibroblasts show different effects in immunotherapy of different cancer types. Fibroblasts are usually divided into myofibroblasts (myCAFs), inflammatory fibroblasts (iCAFs) and antigen-presenting fibroblasts (apCAFs). This study first discovered that collagen I in TLS + The dual role of macrophages in immunotherapy has also been widely reported. The findings in this ESCC clinical trial showed that PD-L1 in TLS + Macrophages have a negative effect on therapeutic response and may affect the activity of B cells and T cells through immunosuppression, thereby inhibiting the effect of immunotherapy.

[0384] The TME of ESCC can be divided into epithelial, immune, and fibrotic regions, and the role of specific cell subsets within these regions plays a key role in treatment response. In the epithelial region, CD4 + T / B cells and CD4 + / CD8 + Higher frequencies of T cells are associated with good outcomes, as are immunosuppressive cells such as αSMA after treatment + and FAP + Fibroblasts, Tregs, and SPP1 + Macrophages were reduced, suggesting that a reduction in immunosuppressive factors promoted the therapeutic effect. Enhanced T / B cell interactions in immune regions were associated with good responders, emphasizing the importance of such cell interactions in effective immune responses.

[0385] In the fibrotic areas, the spatial structure of immune cells was more disordered, which is consistent with the previous finding that fibrosis may hinder the immune response. Traditionally, fibroblasts are regarded as negative factors in immune response. However, αSMA was observed in the good responder group. + Myofibroblasts increased in frequency after treatment, suggesting that these cells may play a complex role in regulating immunotherapy responses.

[0386] In conclusion, this study not only demonstrated the efficacy of neoadjuvant therapy combining PD-1 inhibitors and chemotherapy in ESCC, but also emphasized the potential of spatial proteomics in dissecting the dynamic changes of tumor immune microenvironment and overcoming drug resistance. Integrating the conclusions of this study with clinical practice can significantly improve the treatment outcomes of ESCC patients and establish a new paradigm for precision cancer treatment.

[0387] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0388] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims, and the description and drawings may be used to interpret the contents of the claims.

Claims

1. Use of a reagent for detecting a predictive factor in the preparation of a product for predicting the intervention effect on patients with esophageal squamous cell carcinoma, characterized in that: The reagents for detecting predictive factors include: Reagent 1 for detecting the content of an exhaustion marker of functional CD8 T cells in a sample or its expression level and reagent 2 for detecting the content of inflammatory macrophages in a sample.

2. The use according to claim 1, characterized in that The reagents for detecting predictive factors include: Reagent 1 for detecting the content or expression level of exhaustion marker of functional CD8 T cells in a sample and C1QC in the sample + Reagent 2 for the expression level of immune checkpoint molecules in macrophages.

3. The use according to claim 2, characterized in that The exhaustion marker includes at least one of PD-1, TIM3, CD39, TIGIT, VISTA, GITR, LAG3 and CTLA4.

4. The use according to claim 2, characterized in that The immune checkpoint molecules include at least one of IDO1, PD-L1, PD-L2, CD47 and B7H4.

5. The use according to claim 2, characterized in that The functional CD8 T cells include CD45RO + CD8 + T cells.

6. The use according to claim 2, characterized in that The sample sources of the predictive factors include tumor tissues from patients with esophageal squamous cell carcinoma.

7. The use according to any one of claims 1 to 6, characterized in that: The esophageal squamous cell carcinoma patients include patients with locally advanced esophageal squamous cell carcinoma.

8. The use according to any one of claims 1 to 6, characterized in that: The reagent satisfies one or more of the following conditions 1) to 2): 1) The reagent 1 includes one or more of a specific primer, a specific probe and an antibody for detecting the content of the depletion marker or its expression level; 2) The reagent 2 includes C1QC for detecting the epithelial-fibrous junction area + One or more of specific primers, specific probes and antibodies for the expression level of immune checkpoint molecules in macrophages.

9. The use according to any one of claims 1 to 6, characterized in that: The content or expression level of exhaustion markers of functional CD8 T cells in the samples was detected before intervention.

10. The use according to any one of claims 1 to 6, characterized in that: The method for predicting the intervention effect on patients with esophageal squamous cell carcinoma also includes the following characteristic analysis before intervention: Analysis of CD4+ T / B cell ratios and CD8+ T cell and B cell levels in tertiary lymphoid structures.

11. The use according to any one of claims 1 to 6, characterized in that: The patient with esophageal squamous cell carcinoma underwent combined intervention of PD-1 inhibitors, paclitaxel compounds and fluoropyrimidine derivatives.

12. The use according to claim 11, characterized in that The combined intervention meets one or more of the following conditions 1) to 3): 1) The PD-1 inhibitor is Teplizumab; 2) the fluoropyrimidine derivative is S-1; and 3) The paclitaxel compound is nano-albumin-bound paclitaxel.

13. A kit for predicting the intervention effect of patients with esophageal squamous cell carcinoma, characterized in that: The method comprises a reagent 1 for detecting the content or expression level of an exhaustion marker of functional CD8 T cells in a sample and a C1QC + Reagent 2 for measuring the expression level of immune checkpoint molecules in macrophages; The reagent 1 includes one or more of a specific primer, a specific probe and an antibody for detecting the content of the exhaustion marker or its expression level; The reagent 2 includes C1QC for detecting the junction area between epithelium and fiber + One or more of specific primers, specific probes and antibodies for the expression level of immune checkpoint molecules in macrophages.

14. A method for predicting and / or diagnosing the intervention effect of patients with esophageal squamous cell carcinoma for non-diagnostic and non-therapeutic purposes, characterized in that: The method comprises: detecting predictive factors before intervention, wherein the predictive factors include: the content or expression level of exhaustion markers of functional CD8 T cells in the sample and C1QC + Expression levels of immune checkpoint molecules in macrophages.

15. The method according to claim 14, characterized in that The predictive factors also include CD4 + T / B cell ratio and CD8 in tertiary lymphoid structures + T cell and B cell levels.

16. A system for predicting the effect of intervention on patients with esophageal squamous cell carcinoma, characterized in that: The system comprises: A data processing module, used for analyzing the data of the content of the exhaustion marker of functional CD8 T cells in the received or input sample and the content of inflammatory macrophages in the sample to obtain an analysis result; and The judgment and output module is used to judge whether the analysis result meets the preset judgment conditions to diagnose and / or predict the intervention effect of patients with esophageal squamous cell carcinoma, and output the judgment result.

17. The system of claim 16, wherein: The system comprises: The data processing module is used to convert the content of the exhaustion marker of functional CD8 T cells in the received or input sample and the C1QC in the sample + Analyze the data of expression levels of immune checkpoint molecules in macrophages to obtain analysis results; and The judgment and output module is used to judge whether the analysis result meets the preset judgment conditions to diagnose and / or predict the intervention effect of patients with esophageal squamous cell carcinoma, and output the judgment result.

18. A computer-readable storage medium, characterized in that: When the computer program is executed by a processor, the functions of the system according to claim 16 or claim 17 can be realized.

19. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to execute the computer program to implement the functions of the system according to claim 16 or claim 17.

20. A device for predicting the effect of intervention on patients with esophageal squamous cell carcinoma, characterized in that: It comprises one or more of the system of claim 16 or claim 17, the computer-readable storage medium of claim 18, and the computer device of claim 19.

Images