Gastric cancer prognosis prediction method, electronic equipment and program product

Abstract

The invention discloses a gastric cancer prognosis prediction method, prognosis of a gastric cancer patient is predicted through a trained gastric cancer prognosis model, and construction of the gastric cancer prognosis model is based on myeloid reaction scoring. The construction of the gastric cancer prognosis model comprises screening of gastric cancer myeloid immune markers including a plurality of myeloid cell markers and a plurality of immune checkpoints. And by using a Lasso-Cox regression model, concluding and screening a lambda value on the premise of selecting the optimal model, and determining the number of features of which the importance is not zero and the weight of each feature through the selected lambda value to obtain a gastric cancer myeloid response score MRS.

Description

Technical Field

[0001] The present invention belongs to the field of health data technology, and in particular relates to a gastric cancer prognosis prediction method, electronic equipment and program product. Background Art

[0002] Gastric cancer is a very common digestive tract malignancy, with the morbidity and mortality ranking fifth and fourth respectively among all malignant tumors. Currently, the TNM staging and histopathological typing based on the degree of invasion of the primary tumor, lymph node metastasis, and distant metastasis are the most commonly used indicators for judging the prognosis of gastric cancer patients and whether adjuvant treatment is needed.

[0003] For patients with advanced gastric cancer, the NCCN guidelines recommend routine use of adjuvant chemotherapy after surgery. [1-3] Despite this, the prognosis of gastric cancer patients with the same postoperative pathological stage and who receive the same treatment varies greatly. [4] All signs indicate that gastric cancer is a highly heterogeneous tumor, and the existing staging system can no longer provide sufficient information to assist clinicians in making medical decisions. [5,6] Therefore, it is urgent to develop a more efficient, accurate, and relatively simple and easy-to-operate prognostic model.

[0004] TNM staging refers to the tumor staging system, which consists of three main parameters: T (Tumor), N (Node), and M (Metastasis), which represent the primary tumor site, regional lymph node involvement, and the presence or absence of distant metastasis, respectively;

[0005] The NCCN Guidelines, whose full name is the Oncology Clinical Practice Guidelines of the National Comprehensive Cancer Network, are a series of normative guidance documents for cancer treatment and prevention. Summary of the invention

[0006] One of the embodiments disclosed herein is a gastric cancer prognosis analysis model for evaluating the myeloid immune status of gastric cancer based on a bioinformatics method. The construction of the prediction model includes screening gastric cancer myeloid immune markers, including multiple myeloid cell markers, and multiple immune checkpoints.

[0007] The outputs of the gastric cancer prognostic model include myeloid response score MRS,

[0008] MRS=-2.17e-04*CD14 PT+8.77e-07*CD15 PT+7.23e-05*CD206 PT+4.64e-06*SIGLEC9_PT,

[0009] Among them, CD14_PT, CD15_PT, and CD206_PT are myeloid cell markers taken from PT, and SIGLEC9_PT is an immune checkpoint taken from PT.

[0010] After obtaining MRS through the gastric cancer prognostic model, the optimal cutoff points for high-risk and low-risk groups were calculated, and Kaplan-Meier survival curves were drawn for survival analysis.

[0011] The prediction model was trained using the Lasso-Cox regression model to summarize and screen features to select the lambda value under the premise of the optimal model. The number of features with non-zero importance and the weight of each feature were determined by the selected lambda value to obtain the myeloid response score (MRS) of gastric cancer.

[0012] The disclosed embodiments develop a prognostic model, myeloid response score (MRS), that can predict the survival and recurrence of gastric cancer patients by analyzing myeloid cell subpopulations in gastric cancer, and establish a gastric cancer prognostic analysis method. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:

[0014] Figure 1 A schematic diagram of an MRS model according to one embodiment of the present invention.

[0015] Figure 2 A schematic diagram of the prediction performance of MRS according to one embodiment of the present invention.

[0016] in, Figure 2 A uses the training set. Figure 2 B uses the test set. DETAILED DESCRIPTION

[0017] A tumor is an extremely complex ecosystem that contains malignant epithelial cells, immune cells, and stromal cells. The three are closely linked and work together to maintain the ecological balance of the tumor. [7] Among them, myeloid immune cells have attracted more and more attention. Myeloid immune cells are a very important part of immune cells. They infiltrate the core area of ​​the tumor, frequently communicate with malignant epithelial cells, and play a very important role in regulating the transformation of inflammation to cancer and tumor angiogenesis. [8,9] .

[0018] Fortunately, many therapeutic drugs targeting myeloid immune cells are currently undergoing preclinical and clinical trials.

[10] . However, only a few of these drugs have been proven to have significant therapeutic effects. One of the very important reasons is that myeloid immune cells are heterogeneous. Specifically, myeloid immune cells are mainly composed of mast cells, plasma cell-derived dendritic cells, myeloid dendritic cells, monocytes and macrophages.

[0019] In the past decade, with the help of flow cytometry, the heterogeneity of myeloid immune cells has been initially revealed. Scientists have found that monocytes highly express CD14 and CD16 on their surface.

[11] Macrophages have the highest heterogeneity and are involved in many aspects of tumor development.

[12] Generally speaking, macrophages can be divided into two categories: the first category is pro-inflammatory phenotype macrophages, namely M1 macrophages; the second category is anti-inflammatory phenotype macrophages, namely M2 macrophages.

[13] However, a large number of basic studies have reported that the macrophage phenotype in vivo is actually more complex, including pro-angiogenic phenotype, phagocytic phenotype, metabolic abnormality phenotype, proliferation phenotype, etc.

[14] Dendritic cells play the role of antigen presentation

[15] , recruiting T cells to kill tumor cells. Previous studies have reported two different types of myeloid dendritic cells, XCR + CADM1 + cDC1s and CD1A + CD172A + cDC2s, respectively, and CD8 + T cells, CD4 + T cell interactions induce immune responses

[16] .

[0020] On the other hand, immunotherapy has become a new pillar of cancer treatment. Immune checkpoint inhibitors, such as anti-PD-1 / anti-PD-L1, have greatly benefited patients with melanoma, non-small cell lung cancer, and other types of tumors. According to research progress, myeloid immune cells expressing immune checkpoints have become a new target for immunotherapy. The more myeloid immune cells expressing immune checkpoints infiltrate, the worse the patient's prognosis. [17,18] .

[0021] The present invention believes that although there is no literature research involving the role of myeloid immune cells expressing immune checkpoints in the prognosis of gastric cancer, the research on this issue will encounter many difficulties, including the large number of myeloid immune cells infiltrating the tumor, the functional overlap and complementarity between different myeloid subpopulations, and the plasticity of myeloid immune cells. However, if we face up to the prognostic predictive value of the myeloid-related immune microenvironment of the tumor in gastric cancer and deeply explore the myeloid immunosuppressive regulatory network in the tumor microenvironment, it can help find more effective new targets, so as to apply them in clinical practice and serve more patients.

[0022] Therefore, for this purpose, the present invention quantifies the distribution of major myeloid subtypes in immune cells by immunohistochemical counting, describes the myeloid immune characteristics of gastric cancer, constructs a myeloid score for gastric cancer, and proposes a concise and efficient prognostic model for gastric cancer.

[0023] According to one or more embodiments, a method for constructing a gastric cancer prognosis analysis system for evaluating the myeloid immune status of gastric cancer based on a bioinformatics method comprises the following steps.

[0024] S1, Screening of myeloid immune markers in gastric cancer:

[0025] By sorting out the literature, 11 myeloid cell markers were screened out: CD11b, CD33, CD11c, CD14, CD16, CD68, CD86, CD163, CD206, CD15, CD66b, and 7 immune checkpoints: CD73, IDO, LAG3, PD-1, SIGLEC9, SIRPA, and TIM3.

[0026] The literature here includes at least:

[0027] [a1]YANG J, ZHANG L, YU C, et al. Monocyte and macrophagedifferentiation:circulation inflammatory monocyte as biomarker for inflammatory diseases[J]. Biomark Res, 2014, 2(1):1.

[0028] [a2]DENARDO DG,RUFFELL B.Macrophages as regulators of tumorimmunity and immunotherapy[J]. Nat Rev Immunol, 2019,19(6):369-82.

[0029] [a3]VOGEL D Y,GLIM J E,STAVENUITER AW,et al.Human macrophagepolarization in vitro:maturation and activation methods compared[J].Immunobiology,2014,219(9):695-703.

[0030] [a4]GINHOUX F,SCHULTZE J L,MURRAY P J,et al.New insights into themultidimensional concept of macrophage ontogeny,activation and function[J].Nat Immunol,2016,17(1):34-40.

[0031] [a5]MELLMAN I.Dendritic cells:master regulators of the immuneresponse[J].Cancer Immunol Res,2013,1(3):145-9.

[0032] [a6]BINNEWIES M,MUJAL AM,POLLACK J L,et al.Unleashing Type-2DendriticCells to Drive Protective Antitumor CD4(+)T Cell Immunity[J].Cell,2019,177(3):556-71e16.

[0033] [a7]GENTLES AJ,NEWMAN AM,LIU C L,et al.The prognostic landscape ofgenes and infiltrating immune cells across human cancers[J].Nat Med,2015,21(8):938-45.

[0034] [a8] DE VISSER KE, JOYCE J A. The evolving tumor microenvironment: Fromcancer initiation to metastatic outgrowth[J]. Cancer Cell, 2023, 41(3):374-403.

[0035] [a9]SUN Z, ZHANG T, AHMAD MU, et al. Comprehensive assessment of immune context and immunotherapy response via noninvasive imaging in gastric cancer[J]. JClin Invest, 2024.

[0036] S2, gastric cancer tissue microarray preparation and myeloid cell staining:

[0037] All gastric cancer specimens were fixed with 4% neutral formalin and embedded in paraffin. Experienced pathologists manually marked the tumor area, invasion margin and normal area, and the manually marked area was cored, dotted and arrayed to prepare tissue microarrays. Immunohistochemical staining was used to stain 11 myeloid cell markers: CD11b, CD33, CD11c, CD14, CD16, CD68, CD86, CD163, CD206, CD15, CD66b, and 7 immune checkpoints: CD73, IDO, LAG3, PD-1, SIGLEC9, SIRPA, TIM3 in the tissue microarray. The stained microarray was scanned at 400 times magnification, and the number of stained immune cells in the tissue was quantitatively counted using Image ProPlus 6.0 (Medicine Cybernetics), and their distribution in different tumor locations (primary tumor (PT), invasive margin (IM) and normal tissue (NT)) was counted.

[0038] S3, Method for establishing myeloid score in gastric cancer:

[0039] A clinical cohort of gastric cancer patients was randomly divided into a training set and a test set in a 1:1 ratio. LASSO-Cox regression modeling was used to screen 11 myeloid cell molecular markers and 7 immune checkpoints in three sites (tumor tissue (primary tumor, PT, primary tumor, the primary site of the tumor), (invasive margin, IM, the area where the tumor and host tissue meet), and normal tissue (normal tissue, NT, normal tissue, healthy tissue with no pathological changes in structure and function)), a total of 54 immune features, to construct a myeloid immune prognostic prediction model for predicting the prognosis of gastric cancer patients.

[0040] The Lasso regression model was used to screen the standardized "myeloid characteristics", and the consistency index (C-index) was used as the evaluation indicator. The ten-fold cross-validation strategy was used to select the optimal lambda value.

[0041] In the final optimized model, 4 of the 54 immune signature features were selected based on their best lambda values, namely CD14_PT, CD15_PT, CD206_PT, and SIGLEC9_PT. The myeloid response score (MRS) was calculated using the following formula:

[0042] MRS=-2.17e-04*CD14_PT+8.77e-07*CD15_PT+7.23e-05*CD206_PT+4.64e-06*SIGLEC9_PT..

[0043] The MRS of each case was calculated, and the optimal critical value was obtained through the surv_cutpoint function, which was divided into high-risk and low-risk groups, and the KM curve was drawn to improve the survival analysis. The surv_cutpoint function is a tool in the R language that is used to determine the optimal cutpoint of one or more continuous variables. The function uses the method of maximally selected rank statistics, and the cutpoint value given by it corresponds to the most significant relationship with the survival analysis. The KM curve, that is, the Kaplan-Meier survival curve, is used in survival analysis to estimate the patient's survival rate and draw a survival curve. The KM curve uses time as the horizontal axis and the survival rate as the vertical axis. The curve is gradually drawn by grouping the sample survival time data and calculating the proportion of survivors in each group. The curve accurately reflects the survival rate of the observed object at a specific time point, and can compare the effects of different groups, different treatment regimens and other factors on the survival rate.

[0044] In short, the embodiments of the present disclosure quantitatively evaluate the myeloid immune characteristics of gastric cancer by immunohistochemistry; use the Lasso regression model to summarize the screening features (C-index as an indicator) to select the lambda value under the premise of the optimal model. At the same time, the number of features whose importance is not zero and the weight of each feature are determined by the selected lambda value to obtain the myeloid response score (MRS) of gastric cancer. Here, C-index, the full name of concordance index, is called the consistency index. It is used to evaluate the discrimination between the model prediction value and the actual survival result in survival analysis. C-index is an important indicator to measure the predictive ability of the model, especially in the prediction accuracy evaluation of the prognosis model of tumor patients. Specifically, the calculation steps of C-index are as follows:

[0045] 1. Generate all possible case pairs. If there are n observed individuals, the number of all pairs is C(n,2) (number of combinations).

[0046] 2. Remove pairs in which both patients have not reached the event endpoint (such as death), or in which one patient's survival time is shorter than the other patient but the patient has not reached the event endpoint, because such pairs cannot determine which patient has the event first.

[0047] 3. Calculate the number of pairs among the remaining pairs whose predicted results are consistent with the actual results.

[0048] 4. Calculate C-index = number of consistent pairs / total number of valid pairs.

[0049] When the C-index value is between 0.5 and 1, 0.5 means that the model has no predictive ability and is no different from random guessing; 1 means that the model's prediction results are completely consistent with the actual results and have perfect predictive ability. C-index above 0.90 means that the model prediction is highly accurate.

[0050] The beneficial effects of the embodiments of the present disclosure include summarizing and screening the myeloid immune characteristics of gastric cancer, establishing MRS, and having good predictive efficacy for the prognosis of gastric cancer.

[0051] This paper incorporates 11 myeloid cell molecular markers and 7 immune checkpoints from three sites (primary tumor (PT), invasive margin (IM) and normal tissue (NT)), a total of 54 immune characteristics to construct a model, comprehensively explores the role of myeloid immune characteristics in gastric cancer, and finally constructs a simple and effective MRS.

[0052] According to one or more embodiments, a method for constructing a gastric cancer prognosis analysis system for evaluating the myeloid immune status of gastric cancer based on a bioinformatics method is described in detail as follows.

[0053] 1. Case collection and acquisition of myeloid immune characteristics.

[0054] This paper retrospectively collects more than 335 patients who underwent radical resection of gastric malignant tumors in Zhongshan Hospital Affiliated to Fudan University from January 2005 to December 2022, and whose postoperative pathological pathology confirmed gastric adenocarcinoma. The inclusion criteria of patients are: no other malignant tumors, complete clinical pathological imaging data and follow-up information, no preoperative chemotherapy, etc. The basic clinical information of the above patients was entered, such as age, gender, tumor characteristics (such as number, size and location, etc.), pathological characteristics (such as tumor differentiation, infiltration depth, vascular invasion, lymph node metastasis, MSI, Her2 status, EBV status, TPS score, etc.), blood biochemical indicators (such as blood routine, liver and kidney function, electrolytes, tumor markers) and other parameters; regular follow-up of patients, such as postoperative recovery, whether to receive adjuvant chemotherapy, what chemotherapy regimen and chemotherapy cycle to receive, whether there is postoperative recurrence, recurrence time, recurrence site and what treatment regimen to use after recurrence, the efficacy of recurrence treatment, postoperative survival status, death time and cause of death, etc. Tumors, invasion boundaries, and normal tissues of patients in the cohort were sampled and prepared into tissue microarrays, and the above areas were stained by immunohistochemistry. The distribution of major myeloid subtypes in immune cells was quantitatively obtained (including 11 myeloid cell markers: CD11b, CD33, CD11c, CD14, CD16, CD68, CD86, CD163, CD206, CD15, CD66b, and 7 immune checkpoints: CD73, IDO, LAG3, PD-1, SIGLEC9, SIRPA, TIM3).

[0055] A clinical cohort of 335 gastric cancer patients was randomly divided into a training set (n = 167) and a test set (n = 168) in a 1:1 ratio. The clinical and pathological characteristics of the two groups of patients were evenly distributed, as shown in Table 1.

[0056] Table 1 Clinical baseline characteristics of patients

[0057]

[0058]

[0059] 2. Model model construction.

[0060] Using LASSO-Cox regression modeling, 11 myeloid cell molecular markers and 7 immune checkpoints in 3 sites (primary tumor, PT), (invasive margin, IM) and normal tissue (normal tissue, NT) were screened to construct a prognostic model for gastric cancer patients. A prognostic risk score, namely myeloid response score (MRS), was established. This risk score is the model for predicting the prognosis of gastric cancer provided by the present invention, such as Figure 1 As shown, the calculation formula is as follows: MRS = -2.17e-04*CD14_PT+8.77e-07*CD15_PT+7.23e-05*CD206_PT+4.64e-06*SIGLEC9_PT.

[0061] Figure 1 The English terms involved include:

[0062] Figure A shows the changes in model coefficients under different numbers of parameters.

[0063] Number of remaining parameters——Number of remaining parameters,

[0064] Coefficients——Coefficients refer to the weight of each parameter in the model

[0065] log(λ)——The logarithm of λ, λ(lambda) is a parameter of regularization strength, which controls the complexity of the model.

[0066] Figure B shows the partial likelihood deviation for different λ values.

[0067] Partial Likelihood Deviance is an indicator to measure the goodness of model fit. The smaller the value, the better the model fit.

[0068] Figure 1 The curves in Figure 1 represent the coefficients of different features (such as TIM3_NT, CD206_IM, etc.) as the regularization parameter λ changes. As λ increases (from right to left), more parameter coefficients tend to 0, which means that the model becomes simpler. The red dots in Figure B represent the partial likelihood deviations of the model at different λ values, and the vertical dotted lines may represent the best λ value selected, which is usually the point that minimizes the deviation or the point with the least number of parameters within a certain deviation tolerance range.

[0069] Figure 1A gastric cancer prognostic model based on myeloid response score (MRS) was presented. The MRS model predicts the prognosis of gastric cancer by comprehensively analyzing 11 myeloid cell molecular markers and 7 immune checkpoints in the primary tumor (PT), invasive margin (IM) and normal tissue (NT) of gastric cancer patients, totaling 54 immune features.

[0070] The model building process includes:

[0071] S101. Inductive screening - Screening out myeloid immune features that are closely related to the prognosis of gastric cancer from numerous immune features.

[0072] S102. Establishment of MRS - Based on the screened features, a myeloid response score (MRS) model was constructed. S103. Predictive efficacy - The MRS model showed good ability to predict the prognosis of gastric cancer.

[0073] In this model, each immune signature has a corresponding coefficient, which is obtained through LASSO-Cox regression analysis. LASSO-Cox regression is a statistical method used for variable selection and regularization in regression analysis to avoid overfitting.

[0074] Among them, CD14_PT, CD15_PT, CD206_PT and SIGLEC9_PT represent the expression levels of the corresponding immune markers in tumor tissues. This score can be used to assess the prognostic risk of gastric cancer patients. The higher the score, the better the prognosis may be.

[0075] 3. Model prognostic analysis.

[0076] The optimal cutoff value of the model is obtained by using the function disclosed in this paper. Gastric cancer patients are divided into two groups with high and low MRS scores, such as Figure 2 As shown in the figure, the Kaplan-Meier curve suggests that MRS based on IHC can fully predict the survival of GC patients after radical resection.

[0077] Figure 2 The analysis results of two sets of data (A and B) are shown. Figure 2 A uses the training set. Figure 2 B uses a test set, which includes ROC curves, survival curves (Kaplan-Meier curves) and corresponding statistical data. The English terms include:

[0078] 1. True positive rate (TPR) - True rate, also known as sensitivity or recall, indicates the proportion of samples that are correctly identified as positive among all samples that are actually positive.

[0079] 2. False positive rate (FPR) - False positive rate, which indicates the proportion of samples that are mistakenly identified as positive among all samples that are actually negative.

[0080] 3.AUC (Area Under Curve)——The area under the curve is used to measure the performance of the ROC curve. The closer the value is to 1, the better the model's discrimination ability.

[0081] 4.95% CI (95% Confidence Interval)——95% confidence interval, indicating the credibility of the statistical estimate, is used here to indicate the confidence interval of AUC.

[0082] 5. Overall Survival (OS) refers to the time from the start of treatment to the patient's death from any cause.

[0083] 6.Disease-Free Survival (DFS)——Disease-free survival rate refers to the time from the start of treatment to the patient's disease recurrence or death.

[0084] 7. Recurrence-Free Survival——Recurrence-free survival rate refers to the time from the start of treatment to the recurrence of the patient's disease.

[0085] 8. Number at risk: The number of patients who are still in the study at a specific point in time.

[0086] 9.P-value——P value is used to test the significance of statistical hypotheses. The smaller the P value, the more significant the result.

[0087] Figure 2 In Figures A and B, the left graphs show the ROC curve, which is used to evaluate the performance of the classification model. The closer the curve is to the upper left corner, the better the model's ability to distinguish. The AUC value and its 95% confidence interval provide a quantitative indicator of the model's performance. The middle graphs of Figures A and B show the Kaplan-Meier survival curve, which is used to compare the overall survival rates of the two groups of patients (MRS^high and MRS^low). The rate of decline and shape of the curve can reflect the survival differences between different groups. The right graphs of Figures A and B show the recurrence-free survival curve, which is used to compare the recurrence-free survival of the two groups of patients after treatment. The rate of decline and shape of the curve can also reflect the recurrence-free survival differences between different groups. A P value less than 0.0001 indicates that the survival difference between the two groups is highly statistically significant.

[0088] In the disclosed embodiment, paraffin sections or tissue chips are prepared according to standards for tissue specimens of gastric cancer patients, and immunohistochemical staining is performed on the markers in MRS to quantify the immunohistochemical results of myeloid markers in each case. The immunohistochemical quantitative results are brought into the MRS calculation formula to calculate the MRS of the case and perform prognosis prediction.

[0089] References Related to This Disclosure

[0090] [1]SUNG H,FERLAY J,SIEGEL RL,et al.Global Cancer Statistics 2020:GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36Cancersin185Countries[J].CACancer J Clin,2021,71(3):209-49.

[0091] [2]GROUP G,PAOLETTI

[0092] [3] BANG YJ, KIM YW, YANG HK, et al. Adjuvant capecitabine andoxaliplatin for gastric cancer after D2 gastrectomy (CLASSIC): a phase 3 open-label, randomized controlled trial [J]. Lancet, 2012, 379(9813): 315-21.

[0093] [4]NOH SH, PARK SR, YANG HK, et al.Adjuvant capecitabine plusoxaliplatin for gastric cancer after D2 gastrectomy(CLASSIC):5-year follow-upof an open-label,randomised phase 3trial[J].Lancet Oncol, 2014,15(12):1389-96.

[0094] [5]SASAKO M,INOUE M,LIN J T,et al.Gastric Cancer Working Group report[J].Jpn J Clin Oncol,2010,40Suppl 1:i28-37.

[0095] [6]MCLEAN M H,EL-OMAR E M.Genetics of gastric cancer[J].Nat RevGastroenterol Hepatol,2014,11(11):664-74.

[0096] [7]WADHWA R,SONG S,LEE J S,et al.Gastric cancer-molecular andclinical dimensions[J].Nat Rev Clin Oncol,2013,10(11):643-55.

[0097] [8]HANAHAN D,WEINBERG R A.Hallmarks of cancer:the next generation[J].Cell,2011,144(5):646-74.

[0098] [9]ENGBLOM C,PFIRSCHKE C,PITTET M J.The role of myeloid cells incancer therapies[J].Nat Rev Cancer,2016,16(7):447-62.

[0099]

[10] SCHMID M C,VARNER J A.Myeloid cells in the tumormicroenvironment:modulation of tumor angiogenesis and tumor inflammation[J].JOncol,2010,2010:201026.

[0100]

[11] NAKAMURAK,SMYTH M J.Myeloid immunosuppression and immunecheckpoints in the tumor microenvironment[J].Cell Mol Immunol,2020,17(1):1-12.

[0101]

[12] YANG J,ZHANG L,YU C,et al.Monocyte and macrophagedifferentiation:circulation inflammatory monocyte as biomarker forinflammatory diseases[J].Biomark Res,2014,2(1):1.

[0102]

[13] DENARDO D G,RUFFELL B.Macrophages as regulators of tumourimmunity and immunotherapy[J].Nat Rev Immunol,2019,19(6):369-82.

[0103]

[14] VOGEL D Y,GLIM J E,STAVENUITER AW,et al.Human macrophagepolarization in vitro:maturation and activation methods compared[J].Immunobiology,2014,219(9):695-703.

[0104]

[15] GINHOUX F,SCHULTZE J L,MURRAY P J,et al.New insights into themultidimensional concept of macrophage ontogeny,activation and function[J].Nat Immunol,2016,17(1):34-40.

[0105]

[16] MELLMAN I.Dendritic cells:master regulators of the immuneresponse[J].Cancer Immunol Res,2013,1(3):145-9.

[0106]

[17] BINNEWIES M,MUJAL AM,POLLACK J L,et al.Unleashing Type-2DendriticCells to Drive Protective Antitumor CD4(+)T Cell Immunity[J].Cell,2019,177(3):556-71 e16.

[0107]

[18] ZHANG Q W,LIU L,GONG C Y,et al.Prognostic significance of tumor-associated macrophages in solid tumor:a meta-analysis of the literature[J].PLoS One,2012,7(12):e50946.

[0108]

[19] GENTLES AJ,NEWMAN AM,LIU C L,et al.The prognostic landscape ofgenes and infiltrating immune cells across human cancers[J].Nat Med,2015,21(8):938-45.

[0109]

[20] GROUP G,PAOLETTI X,OBAK,et al.Benefit of adjuvant chemotherapyfor resectable gastric cancer:a meta-analysis[J].JAMA,2010,303(17):1729-37.

[0110]

[21] BANG Y J,KIM Y W,YANG H K,et al.Adjuvant capecitabine andoxaliplatin for gastric cancer after D2 gastrectomy(CLASSIC):a phase 3 open-label,randomised controlled trial[J].Lancet,2012,379(9813):315-21.

[0111]

[22] NOH S H,PARK S R,YANG H K,et al.Adjuvant capecitabine plusoxaliplatin for gastric cancer after D2 gastrectomy(CLASSIC):5-year follow-upof an open-label,randomised phase 3 trial[J].Lancet Oncol,2014,15(12):1389-96.

[0112]

[23] SASAKO M,INOUE M,LIN J T,et al.Gastric Cancer Working Groupreport[J].Jpn J Clin Oncol,2010,40 Suppl 1:i28-37.

[0113]

[24] MCLEAN M H,EL-OMAR E M.Genetics of gastric cancer[J].Nat RevGastroenterol Hepatol,2014,11(11):664-74.

[0114]

[25] WADHWAR,SONG S,LEE J S,et al.Gastric cancer-molecular andclinical dimensions[J].Nat Rev Clin Oncol,2013,10(11):643-55.

[0115]

[26] HANAHAN D,WEINBERG R A.Hallmarks of cancer:the next generation[J].Cell,2011,144(5):646-74.

[0116]

[27] ENGBLOM C,PFIRSCHKE C,PITTET M J.The role of myeloid cells incancer therapies[J].Nat Rev Cancer,2016,16(7):447-62.

[0117]

[28] SCHMID M C,VARNER J A.Myeloid cells in the tumormicroenvironment:modulation of tumor angiogenesis and tumor inflammation[J].JOncol,2010,2010:201026.

[0118]

[29] NAKAMURAK,SMYTH M J.Myeloid immunosuppression and immunecheckpoints in the tumor microenvironment[J].Cell Mol Immunol,2020,17(1):1-12.

[0119]

[30] YANG J,ZHANG L,YU C,et al.Monocyte and macrophagedifferentiation:circulation inflammatory monocyte as biomarker forinflammatory diseases[J].Biomark Res,2014,2(1):1.

[0120]

[31] DENARDO D G,RUFFELL B.Macrophages as regulators of tumourimmunity and immunotherapy[J].Nat Rev Immunol,2019,19(6):369-82.

[0121]

[32] VOGEL D Y,GLIM J E,STAVENUITER AW,et al.Human macrophagepolarization in vitro:maturation and activation methods compared[J].Immunobiology,2014,219(9):695-703.

[0122]

[33] GINHOUX F,SCHULTZE J L,MURRAY P J,et al.New insights into themultidimensional concept of macrophage ontogeny,activation and function[J].Nat Immunol,2016,17(1):34-40.

[0123]

[34] MELLMAN I.Dendritic cells:master regulators of the immuneresponse[J].Cancer Immunol Res,2013,1(3):145-9.

[0124]

[35] BINNEWIES M,MUJAL AM,POLLACK J L,et al.Unleashing Type-2DendriticCells to Drive Protective Antitumor CD4(+)T Cell Immunity[J].Cell,2019,177(3):556-71e16.

[0125]

[36] ZHANG Q W,LIU L,GONG C Y,et al.Prognostic significance of tumor-associated macrophages in solid tumor:a meta-analysis of the literature[J].PLoS One,2012,7(12):e50946.

[0126]

[37] GENTLES AJ,NEWMAN AM,LIU C L,et al.The prognostic landscape ofgenes and infiltrating immune cells across human cancers[J].Nat Med,2015,21(8):938-45.

[0127] It should be understood that in the embodiments of the present invention, the term "and / or" is only a description of the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0128] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0129] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for predicting gastric cancer prognosis, characterized in that: The prognosis of gastric cancer patients is predicted by a trained gastric cancer prognostic model, wherein the gastric cancer prognostic model is constructed based on the myeloid response score.

2. The method according to claim 1, characterized in that: The process of constructing the gastric cancer prognosis model includes screening gastric cancer myeloid immune markers, including multiple myeloid cell markers and multiple immune checkpoints.

3. The method according to claim 2, characterized in that The outputs of the gastric cancer prognostic model include myeloid response score MRS, MRS=-2.17e-04*CD14_PT+8.77e-07*CD15_PT+7.23e-05*CD206_PT+4.64e-06*SIGLEC9_PT, Among them, CD14_PT, CD15_PT, and CD206_PT are myeloid cell markers taken from PT, and SIGLEC9_PT is an immune checkpoint taken from PT.

4. The method according to claim 3, characterized in that After obtaining MRS through the gastric cancer prognostic model, the optimal cutoff points for high-risk and low-risk groups were calculated, and Kaplan-Meier survival curves were drawn for survival analysis.

5. The method according to claim 3, characterized in that: The Lasso-Cox regression model was used to summarize and screen features to select the lambda value under the premise of the optimal model. The number of features with non-zero importance and the weight of each feature were determined by the selected lambda value to obtain the myeloid response score (MRS) of gastric cancer.

6. The method according to claim 2, characterized in that The screening of myeloid immune markers for gastric cancer was carried out by sorting out the literature, and 11 myeloid cell markers were screened out, including: CD11b, CD33, CD11c, CD14, CD16, CD68, CD86, CD163, CD206, CD15, CD66b, and 7 immune checkpoints, including: CD73, IDO, LAG3, PD-1, SIGLEC9, SIRPA, and TIM3.

7. The method according to claim 6, characterized in that Tissue microarrays of gastric cancer samples were prepared, and immunohistochemical staining was used to stain 11 myeloid cell markers CD11b, CD33, CD11c, CD14, CD16, CD68, CD86, CD163, CD206, CD15, CD66b, and 7 immune checkpoints CD73, IDO, LAG3, PD-1, SIGLEC9, SIRPA, and TIM3 in the tissue microarray.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor runs the computer program to implement the method according to any one of claims 1 to 5.

9. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

10. A computer program product, comprising a computer program, characterized in that The computer program is executed by a processor to implement the method according to any one of claims 1 to 5.

Images