Marker for predicting curative effect of tumor immunotherapy as well as application and detection method of marker

Abstract

The invention discloses a marker for predicting the curative effect of tumor immunotherapy as well as application and a detection method of the marker, and belongs to the technical field of biomedicine. The marker is glycerophosphorylcholine (GPC) in peripheral blood, and under the condition that the GPC level of peripheral plasma is low, it is judged that ICIs immunotherapy is poor in curative effect response. Peripheral blood samples are simple and noninvasive to obtain, longitudinal curative effect monitoring can be realized, the method is independent of tissue biopsy and is not influenced by tumor heterogeneity, the cost is reduced, and application is facilitated; meanwhile, peripheral blood not only objectively reflects the systematic immune state of a host, but also is closely related to tumor immunity as an important ring in a tumor immune cycle, and in the aspect of accuracy, the efficiency of a prediction model constructed by GPC reaches 0.886 and is remarkably higher than that of PDL1.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of biomedical technology, and in particular to a marker for predicting the efficacy of tumor immunotherapy, and an application and detection method thereof. BACKGROUND

[0002] In the past decade, immune checkpoint inhibitors (ICIs) have greatly improved the survival of tumor patients. At present, ICIs immunotherapy has completely changed the treatment mode of various solid tumors. Among them, PD-1 / PD-L1 antibodies have shown significant efficacy in lung cancer, esophageal cancer, head and neck squamous cell carcinoma, renal cancer, liver cancer, gastric cancer and other solid tumors. FDA has approved PD-1 / PD-L1 antibodies for multi-line clinical indications in different solid tumors. ICIs immunotherapy has achieved remarkable success in clinical practice of tumors, however, its efficacy varies significantly among different patients. The overall effective rate of existing reported solid tumor ICIs treatment is only 20%-30%. Whether a patient responds to ICIs treatment is affected by many factors, including tumor and treatment-related factors, tumor-unrelated host-level factors, and environmental factors, such as tumor molecular mutation spectrum and metabolic remodeling, immune and vascular remodeling characteristics of the microenvironment, intestinal microorganisms, and host genetic immune characteristics. In clinical diagnosis and treatment, ICIs immunotherapy needs efficient and convenient biomarkers to identify the responding population in time and accurately, so as to avoid unnecessary economic burden, the risk of irAEs and disease delay for non-responding population. FDA has approved three markers for clinical use, including tumor PD-L1 expression, tumor mutation burden (TMB) and microsatellite instability (MSI), however, there are still deficiencies in the accuracy and practicability of the three markers in identifying ICIs responding patients. For example, some patients with very low or negative PD-L1 expression are still effective for ICIs treatment, while some patients with PD-L1 expression higher than 50% do not respond to ICIs, and some studies have reported that induction of PD-L1 overexpression in non-small cell lung cancer may reduce the efficacy of PD-1 antibody; and TMB and MSI also cannot fully identify effective patients.

[0003] In summary, the existing clinical application of ICI efficacy markers, mainly tumor tissue molecular markers such as PD-L1 level detected by immunohistochemical staining, TMB, MSI, etc., are extremely dependent on biopsy tissue acquisition and pathologist's subjective judgment, are greatly affected by tumor heterogeneity, are controversial in tumor or immune cell judgment, antibody model difference, cutoff value selection, sequencing depth, sequencing panel, high cost, long detection period, etc., and are greatly limited in clinical application. At present, the precise prediction of ICI efficacy is still an important problem to be solved in clinical practice, and a large number of studies are focusing on the screening of precise molecular markers for immunotherapy. Literature reports a variety of molecular markers such as tumor POLE / POLD1 gene mutation, microenvironment immunity and host immune characteristics, tumor and host metabolic characteristics, peripheral blood secreted proteins and other circulating characteristics, but the markers under research have not been clinically verified. In summary, there is still a lack of precise and efficient markers to identify patients effectively treated by ICI in clinical practice. SUMMARY

[0004] In order to solve the above problems, the present application provides a marker for predicting the efficacy of tumor immunotherapy and its application.

[0005] In order to achieve the above-mentioned purpose, the present application is realized by the following technical scheme: A marker for predicting the efficacy of tumor immunotherapy, wherein the marker is glycerophosphatidylcholine (GPC) in a detection sample, and the detection sample includes peripheral blood.

[0006] The present application also discloses the application of the marker glycerophosphatidylcholine (GPC) in preparing a product for predicting the efficacy of tumor immunotherapy, enhancing the efficacy of tumor immunotherapy, or regulating tumor immune microenvironment.

[0007] Further, the product detection reagent or detection kit or diagnostic device for predicting the efficacy of tumor immunotherapy is used for detecting the expression level of the marker glycerophosphatidylcholine in a sample.

[0008] The present application also discloses the application of a detection kit for predicting the efficacy of tumor immunotherapy, which is not for diagnostic purposes, comprising (1) an analysis module, which is used to determine the expression level of a marker in a sample to be tested of a subject, and (2) an evaluation module, which is used to determine the efficacy of tumor immunotherapy of the subject according to the expression level of the marker determined in (1); In (1), the marker is glycerophosphatidylcholine in peripheral blood.

[0009] Further, the subject is a tumor patient using ICI immunotherapy. The tumor patient includes a solid tumor patient.

[0010] Further, the sample to be tested is peripheral blood.

[0011] Further, the judgment in the evaluation module is that the tumor immunotherapy efficacy is positively correlated with the glycerophosphocholine level in peripheral blood.

[0012] The present application discloses a marker for predicting tumor immunotherapy efficacy and application thereof, which has the beneficial effects of: (1) The GPC molecular marker in the present application is highly expressed in the plasma of patients with good response to ICI immunotherapy, and is a predictive marker for good efficacy of ICI immunotherapy. GPC may be involved in the regulation of tumor immune microenvironment and the potential marker for prognosis evaluation. When the GPC level in peripheral blood plasma is high, it is determined that the ICI immunotherapy efficacy is good; when the GPC level in peripheral blood plasma is low, it is determined that the ICI immunotherapy efficacy is poor.

[0013] (2) Peripheral blood blood samples are simple and non-invasive, can realize longitudinal efficacy monitoring, are not dependent on tissue biopsy, are not affected by tumor heterogeneity, have reduced cost and are easy to apply; at the same time, peripheral blood not only objectively reflects the systemic immune status of the host, but also is closely related to tumor immunity as an important part of the tumor immune cycle. In terms of accuracy, the efficiency of the prediction model constructed by GPC reaches 0.886, which is significantly higher than that of PDL1. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0015] Figure 1 is a flow chart of the establishment of a clinical cohort and the inclusion of patients; Figure 2 is a schematic diagram of the results of metabolomics analysis; Figure 3 is a schematic diagram of the results of differential analysis and survival analysis; Figure 4 is an ICI immunotherapy efficacy prediction model based on GPC; Figure 5are schematic diagrams of animal model establishment and detection results; wherein, A is a schematic diagram of animal model establishment process; B is a tumor growth curve diagram; C is a tumor live imaging diagram; D is a tumor volume columnar statistical diagram; E is a tumor gross picture; F is a survival curve diagram; in the diagram, Control (solvent control group), GPC (GPC treatment group), anti-PD-1 (anti-PD-1 treatment group), GPC & anti-PD-1 (GPC and anti-PD-1 combined treatment group); Figure 6 Results of detection of the effect of GPC on T cell function in vitro, wherein, A is the maximum safe dose of GPC stimulating Jurkat cells for 24 hours; B is the maximum safe dose of GPC stimulating Jurkat cells for 48 hours; C is the RNA synthesis of lymphocyte IFNγ after stimulating Jurkat cells with safe dose of GPC; D is the protein secretion result after stimulating Jurkat cells with safe dose of GPC. DETAILED DESCRIPTION

[0016] In order for those skilled in the art to better understand the present application, the technical solutions of the present application will be described clearly and completely in conjunction with the embodiments below. Obviously, the described embodiments are only some of the embodiments of the present application, but not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0017] A marker for predicting the efficacy of tumor immunotherapy is glycerophosphatidylcholine (Glycerophosphatidylcholine, GPC) in peripheral blood.

[0018] The marker of the present application can be used to prepare products for predicting the efficacy of tumor immunotherapy, such as kits. It can also be used to prepare products for enhancing the efficacy of tumor immunotherapy; and it can also be used to prepare products for detecting the regulation of tumor immune microenvironment.

[0019] The application first compares the plasma GPC levels of patients with good and poor response to ICI immunotherapy by a method verified by plasma non-targeted metabolomics detection (HPLC-MS / MS) and enzyme-linked immunoassay (ELISA), and finds that the GPC level of tumor patients with good response to ICI immunotherapy is significantly higher than that of patients with poor response. The application obtains the positive correlation between the immunotherapy effect and the GPC level through the survival analysis of the high and low expression of GPC after immunotherapy, and confirms that GPC is an independent influencing factor of poor prognosis of patients with immunotherapy through single factor and multi-factor analysis. In addition, the application uses various machine learning algorithms, including random forest (RF), support vector machine (SVM) and extreme gradient boosting (XGboost), combined with clinical electronic medical record information indicators, to construct a tumor immunotherapy efficacy prediction model, and through variable importance analysis, it is shown that GPC plays a key role in predicting the efficacy of tumor immunotherapy. Finally, the application enhances the efficacy of immunotherapy in a mouse tumor model by increasing the GPC level, and also shows that GPC may play an important regulatory role in the tumor immune microenvironment.

[0020] Example 1

[0021] Research method I. Establishment of clinical cohort (as shown in Figure 1 ) (1) Patient screening This study has been applied for and approved by the ethics committee of West China Hospital [Ethics Approval No: 2019 Audit (1045) No.], and the research includes solid tumor patients who visited Sichuan University West China Hospital for the first time from May 1, 2020 to May 1, 2021, and used ICI immunotherapy for the first time.

[0022] Inclusion criteria: Patients included need to meet the following criteria: ① Patients with clinically diagnosed solid tumors; ② No previous use of any type of ICI monotherapy or combination therapy; ③ Have communicated with doctors and patients, and are expected to undergo initial ICI immunotherapy in the near future.

[0023] Exclusion criteria: Excluded were patients with melanoma, reproductive system tumors, and colorectal cancer.

[0024] (2) Patient follow-up First, 1-month follow-up, i.e. follow-up of treatment plan within 1 month after sample collection, excluding patients who ultimately refuse ICI immunotherapy due to various reasons and samples.

[0025] Second, 6 months of periodic follow-up, record the patient each cycle of efficacy evaluation results (RECIST 1.1), collect patient electronic medical record information, and according to the following exclusion criteria screening patients.

[0026] Exclusion criteria: ① Exclusion of patients who do not comply with the rules of the study follow-up period to receive ICIs treatment or loss of follow-up; ② Exclusion of patients with unclear clinical outcome, efficacy and adverse reaction grouping is not accurate; ③ Exclusion of patients with efficacy evaluation for super progress; ④ Exclusion of patients with incomplete medical records.

[0027] Study endpoints: According to previous studies, patients with disease progression within 6 months were defined as ICIs treatment non-response group, and patients with stable disease / partial remission / complete remission for more than 6 months were defined as ICIs treatment response group.

[0028] This study finally included 155 patients, found 57 patients in the cohort, and 98 patients in the validation cohort.

[0029] II. Metabolomics analysis (mass spectrometry detection, data analysis, marker screening and verification, such as Figure 2 ) (1) Plasma sample pretreatment and extraction of metabolite molecules Use analytical reagents for sample pretreatment: ① Take out the plasma sample stored in the-80℃ refrigerator, dissolve on ice, take 100 μL plasma in 1.5 mL centrifuge tube for extraction of metabolite molecules, add standardized L-2-chlorophenylalanine as internal standard metabolite (final concentration 0.3 mg / mL; 10 μL methanol preparation), shake 10 sec and mix well; ② Organic solvent protein precipitation: all reagents were pre-cooled at-20℃, protein precipitant (methanol: acetonitrile 200 μL: 100 μL) was added, shaken for 1 min; 0℃ water bath, ultrasonic 10 min, -20℃ freezing 30 min; 13000 rpm, 4℃, centrifugation 10 min, the supernatant was small molecule metabolites, the precipitate was large molecule protein and cell debris, etc., take 200 μL supernatant, -20℃ volatilization dry, get metabolite molecules; ③ Resolubilization and re-extraction: 300 μL methanol-water (1:4) was used to resolubilize the metabolite molecules obtained in step ②, vortexed for 30 s, ultrasonic for 3 min, -20℃ frozen for 2 hours, 13000 rpm, 4℃, centrifuged for 10 min, 150 μL supernatant, filtered through a 0.22 μm needle hole filter, obtained metabolite extract, -80℃ stored.

[0030] (4) Preparation of quality control samples (QCs): Take an equal volume of metabolite extract from all samples, mix and prepare to form QC samples. The total volume of the QC sample is the same as the sample to be tested.

[0031] (2) High performance liquid chromatography tandem mass spectrometry (HPLC-MS / MS) scanning detection Take out the frozen sample metabolite extract and QC sample, dissolve on ice, and perform HPLC-MS / MS full scan detection. The Dionex Ultimate 3000 RS UHPLC ultra-high performance liquid chromatography tandem Thermo Q-EXACTIVE plus (QE plus) high resolution mass spectrometer is used in this study. The parameters of liquid chromatography and mass spectrometry are as follows. In the entire analysis process, in order to ensure the stability and reliability of the detection results, a QC sample is detected every 10 samples.

[0032] ① The chromatographic conditions are as follows: Chromatographic column: ACQUITY UPLC HSS T3 (100 mm x 2.1 mm, 1.8 um) Waters; Column temperature: 45 °C; Mobile phase: A-water (containing 0.1% formic acid), B-acetonitrile (containing 0.1% formic acid); Flow rate: 0.35 mL / min; Injection volume: 2 μL.

[0033]

[0034] ② The mass spectrometry conditions are as follows: Ion source: ESI electrospqray ionization, electrospqray ionization, Thermo Fisher; Sample mass spectrum signal acquisition: positive ion scanning mode and negative ion scanning mode.

[0035]

[0036] (3) HPLC-MS / MS raw data processing The system software MarkerView 1.2.1 is used for peak detection, peak alignment, data scaling and standardization of the mass spectrometry data. The mass-to-charge ratio, retention time and corresponding peak intensity of the metabolites are exported.

[0037] ① Base peak chromatogram (abbreviated as base peak chart) After the sample is separated by chromatography, each molecule flows out and enters the mass spectrometer. After ionization and fragmentation in the ion source, the fragments are separated in the mass analyzer according to the mass-to-charge ratio, and then the data are collected in the ion detector. Each scan at different times can obtain an ion spectrum, and the ion spectrum with the strongest signal at each time point is the base peak spectrum, which is usually used for qualitative analysis of metabolites. The vertical coordinate represents the ion intensity, and the height and area of the peak represent the signal intensity of the ion, which is related to the content of the ion and is used for quantitative analysis of metabolites.

[0038] ②Data quality control Because the high-performance liquid chromatography tandem mass spectrometry (HPLC-MS / MS) detection instrument is complex and precise, and the sample size of the same batch in this study exceeds 100, as the detection time is prolonged, the instrument is affected by various factors such as operating temperature and air humidity, which may cause systematic errors and reduce the reliability of the data results. Therefore, certain measures were taken for quality control during the detection and analysis process, including QC sample quality control and internal standard metabolite quality control.

[0039] QC sample quality control refers to regularly inserting QC samples between the samples to be tested. The QC samples before being put on the machine are used to balance the system, and the QC samples at intervals are used to evaluate the stability of the instrument. Internal standard metabolite quality control refers to adding a known stable internal reference metabolite to the sample to be tested to evaluate whether the detection results are accurate and stable. The uniformity of plasma metabolite extraction is reflected by the signal of the internal standard metabolite. The internal reference metabolite used in this experiment is L-2-chlorophenylalanine, 0.3 mg / mL. For QC sample quality control, the following four methods were used to evaluate the quality control results: 1) RSD screening of QC samples RSD (Relative standard deviation) is the ratio of the standard deviation of the QC sample detection results to the average value, which represents the dispersion degree of the QC sample data distribution and is used to measure the stability of the mass spectrometry platform at different time points. The smaller the RSD value, the lower the data dispersion, the more concentrated the QC sample data, and the better the stability of the mass spectrometry. In this study, ion peaks with an RSD value greater than 0.4 were deleted.

[0040] 2) PCA analysis of QC samples PCA is a non-supervised data dimensionality reduction analysis method that does not consider sample grouping and only considers sample number or metabolite number for data dimensionality reduction. Orthogonal transformation obtains new vectors that reflect the original variable information to the greatest extent. Through principal component analysis of QC samples, the stability of the detection system is evaluated. After 7-fold cross-validation, the PCA model diagram is obtained. According to the aggregation degree and distance of the QC samples, the stability of the instrument is judged.

[0041] 3) Metabolite intensity distribution of QC samples Boxplot of the metabolite intensities of QC samples and samples to be tested, with the horizontal axis representing each sample and the vertical axis representing the log10 value of the metabolite mass intensity. According to the distribution of the metabolite intensities of the QC samples, the uniformity and stability of the experiment are determined.

[0042] 4) Cluster analysis of QC samples Hierarchical cluster analysis is performed on the metabolite types and contents of the QC and samples to be tested, and a heat map is drawn. According to the position and degree of aggregation of the QC samples in the cluster analysis, the uniformity, stability of the detection, and the relationship between the QC samples and other samples are intuitively displayed.

[0043] ③ Data preprocessing Progenesis QI v2.3 software is used for data processing and preliminary analysis, then the database is searched for qualitative analysis and identification of metabolites. The databases used in this study include the Luming metabolite database, the HMDB database, the Lipidmaps database, and the METLIN database.

[0044] Finally, the metabolites are sorted according to the accuracy score of the qualitative results, with a full score of 60 points. A score below 36 points indicates low accuracy of the qualitative results, and the identification results below 36 points are deleted. Finally, the positive and negative ion data are combined to obtain the standardized data matrix of the metabolite content of each sample, which is used for subsequent analysis.

[0045] (4) Statistical analysis of sample-metabolite data ① Multivariate statistical analysis First, unsupervised PCA analysis is performed to observe the overall distribution of the samples, then PLS-DA and OPLS-DA are performed to observe the metabolic differences between groups and screen for differential metabolites.

[0046] 1) PCA Metabolomics data contains a large amount of variable information. Through PCA analysis, a large number of related variables are linearly transformed to obtain 2-3 decomposition variables that can maximally reflect the influence of all variables on the characteristics of the data. The data is reduced in dimension to form the main components, and finally the original data is described by the main components to reflect the distribution of the samples in two-dimensional or three-dimensional space. The score values t1 and t2 projected by the two principal components PC1 and PC2 are used as the horizontal and vertical coordinates to draw a two-dimensional coordinate graph. Each point on the coordinate graph represents a sample, and the distance between two points represents the difference between the two samples calculated based on the first principal component PC1 and the second principal component PC2.

[0047] 2) PLS-DA PLS-DA refers to partial least squares regression analysis combined with discriminant analysis. During the process of dimensionality reduction analysis, sample grouping information is added to facilitate the selection of characteristic variables that distinguish each group to the greatest extent, calculate the principal components related to the grouping information, and draw two-dimensional and three-dimensional coordinate graphs of the projection scores of each sample on the two to three principal components, better reflecting the sample clustering characteristics.

[0048] 3) OPLS-DA OPLS-DA refers to an extension based on PLS-DA analysis. Before analysis, orthogonal transformation is performed to correct and filter out irrelevant information from the grouping, remove noise variables that affect the model effect, and better reflect the variables most related to the differences between groups. The sample grouping effect is better. Since a large amount of noise is removed, OPLS-DA is more accurate than PLS-DA model, so we finally use the variable weight value (Variable important in projection, VIP) generated by OPLS-DA to screen differential metabolites.

[0049] ② Screening of differential metabolic molecules First, perform univariate statistical analysis, including statistical description and statistical inference. Use the P value and FC value of the t test to compare the differences between metabolites in two groups. Then, combine the results of the previous multivariate statistical analysis and univariate analysis to screen differential metabolites between different groups. The screening criteria for this study are as follows: VIP>1, P<0.05, FC>1.5, and a volcano plot is drawn. Then draw a cluster analysis heat map of the differential metabolic molecules in different samples. Next, to further understand the relationship between the two groups when the metabolic molecules have significant differences between the groups, use the Pearson method for correlation analysis and draw a bubble chart. Finally, perform metabolic pathway enrichment analysis (KEGG) to understand the metabolic pathways in which the differential metabolic molecules are involved.

[0050] Example 2 The detection method for predicting the efficacy of tumor immunotherapy comprises: (1) An analysis module for determining the expression level of a marker in a test sample of a subject; specifically, enzyme-linked immunoassay detection (ELISA): In this study, Jianglai Biological Human Glycerophosphocholine (GPC) ELISA Test Kit was used to detect the GPC level in patient plasma (peripheral blood). The specific steps are as follows: - Rewarming: ELISA kit, test sample room temperature balance 30-60 min; - Centrifugation: centrifuge the test sample, room temperature, 1000g, 15min; - The antibody-binding reaction of the test substance: 50 μL of the corresponding solution was added to the blank wells, standard wells, and sample wells on the antibody-coated well plate, respectively; - Detection of antibody-binding reaction: 100 μL of horseradish peroxidase (HRP)-labeled secondary antibody was added per well, and incubation was performed in a 37°C water bath for 60 min, during which the plate was sealed to prevent evaporation; - Washing of the plate: the reaction solution was spun dry, 350 μL of washing solution was added per well, and the plate was tapped to spin dry, which was repeated for 3-5 times; - Substrate reaction: 50 μL of substrate A and 50 μL of substrate B were added, and incubation was performed in a 37°C water bath in the dark for 15 min; - Reaction termination: 50 μL of termination solution was added, and the reaction was performed for 2-3 min; - Signal detection: the OD value (450 nm) was detected within 15 min after the reaction was terminated; (2) The evaluation module is used to determine the marker expression level to judge the tumor immunotherapy efficacy of the subject; specifically, the detection OD value obtained in step (1) is subjected to data analysis: The standardized OD value = OD value (standard / sample) - OD value (blank) is calculated, a standard curve is drawn, the concentration of the test substance of each sample is calculated, the standardized OD value (sample well) / average OD value (control group) is calculated, a column chart is drawn, and t-test statistical analysis is performed.

[0051] As shown in Figure 2 , based on the metabolite content matrix of each sample, differential metabolites were screened according to the grouping of patient efficacy. PLS-DA analysis ( Figure 2 -A) and OPLS-DA analysis ( Figure 2 -B) can effectively distinguish between ICIs efficacy responder patients and non-responder patients, and the distinguishing effect of OPLS-DA analysis is better; the differential markers are preliminarily screened according to the VIP value greater than 1 and the P value less than 0.05 calculated by OPLS-DA analysis, as shown in Figure 2 -C, and then the differential metabolites are further screened by combining the FC value greater than 1.5 ( Figure 2 -D), molecular pathway enrichment is performed, the most significant pathway enriched by the differential metabolites is the Choline Metabolism in Cancer tumor choline metabolism pathway (P<0.05), and the most significant molecule is glycerophosphatidylcholine (GPC) (as shown in Figure 2 -E, Figure 2 -F).

[0052] The present application utilizes the Survival and survminer packages in R version 4.0.2 to perform Kaplan-Meier (KM) survival analysis and multivariate cox regression analysis. First, the optimal cut-off value is determined using the "surv_cutpoint" function and the "surv_categorize" function. Then the survival curve is drawn and the log-rank P value is calculated to assess the difference between the two groups. Multivariate cox regression analysis is used to describe the effect of multiple variables on survival rate.

[0053] The levels of plasma GPC in ICIs immunotherapy responders and non-responders in the cohort were detected, and the results are shown in FIGS. 1A and 1B. Figure 3 As shown in FIG. 1A, the plasma GPC level in ICIs immunotherapy responders in the cohort was significantly higher than that in non-responders (P < 0.01) (FIG. 1B), and the GPC level was significantly correlated with survival (log-rank P = 0.012) (FIG. 1C), and the multivariate cox regression analysis found that cancer patients with lower GPC levels in the cohort had significantly reduced survival after receiving ICIs immunotherapy [GPC low vs. high, hazard ratio (HR): 4.14, 95% CI, 1.19-14.37, p = 0.0250]. Figure 3 Figure 3 As shown in FIG. 1D, the plasma GPC level in ICIs immunotherapy responders in the validation cohort was significantly higher than that in non-responders (P < 0.0001) (FIG. 1E), and the survival was significantly correlated (log-rank P = 0.035) (FIG. 1F). Multivariate COX regression analysis found that cancer patients with lower GPC levels had significantly reduced survival after receiving ICIs immunotherapy [GPC low vs. high, HR: 2.93, 95% CI, 1.25-6.85, P = 0.0133]. Figure 3 Figure 3

[0054] ​​​Immune therapy efficacy prediction model construction: Laboratory indices and candidate metabolites (including GPC) with P < 0.05 in univariate analysis were selected as model variables. To prevent overfitting, the Pearson correlation coefficient (R2) was used to assess the multicollinearity between variables, and variables with significant correlation were deleted according to the standard of R2>0.5 combined with clinical experience. Then, to ensure the clinical applicability of the model, variables with missing values more than 30% were excluded, while for other variables, random forest algorithm was used to fill in the missing values. The data was standardized using Z-score method. All samples were divided into training set and test set in the ratio of 7:3. Support vector machine (SVM), random forest (RF) and extreme gradient boosting (XGBoost) were used to build prediction models and evaluate the importance of variables. The best hyperparameters of these modeling methods were selected by 3-fold cross-validation. Receiver operating characteristic curve (ROC) and confusion matrix were drawn to intuitively show the prediction performance of the model. Accuracy, sensitivity, specificity and AUC were used to evaluate the prediction ability of the model.

[0055] Three machine learning methods were applied to build clinical prediction models, as shown in Figure 4 The AUC of the SVM model was the highest, which was 0.886 (95% CI: 0.786-0.985), followed by the RF model (AUC = 0.855) and the XGBoost model (AUC = 0.769) Figure 4 -A), and the sensitivity of the SVM model was 90.00%, the specificity was 84.38%, and the accuracy was 86.54 Figure 4 -B). Figure 4 -C and Figure 4 -D show the ranking of variable importance calculated by RF algorithm and XGBoost algorithm respectively, in which the importance of GPC is much higher than that of other variables Figure 4 -C and Figure 4 -D), indicating that GPC level can be used to predict the efficacy of tumor ICIs immunotherapy.

[0056] Example 3 Animal and cell experiments (1) Cell lines and CCK8 detection Human acute T lymphoblastic leukemia cell line Jurkat cells were purchased from CellCook Biotechnology Co., Ltd. (Guangzhou, China) and cultured in RPMI1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS) at 37°C, 5% CO2. The cells used in this study were identified by short tandem repeat (STR) analysis. The maximum GPC dose that had no obvious toxicity on Jurkat cell viability was analyzed by Cell Counting Kit-8 (CCK8, Biosharp, China) according to the manufacturer's instructions (Aladdin, China).

[0057] (2) Cytokine expression detection The cultured U937, THP-1 and Jurkat cells were grouped and given corresponding stimulation as follows: a. Control group (Control): only cell culture medium; b. GPC group (GPC): GPC cell culture medium; c. Cell activator stimulation group (Cell activator): cell activator culture medium; d. GPC plus cell activator stimulation group (GPC+cell activator): GPC and cell activator culture medium; Cells and culture supernatants were collected, and the former was extracted for RNA qPCR detection, and the latter was subjected to ELISA detection.

[0058] As shown in Figure 6 , the effect of GPC on T cell function was studied by in vitro experiments. Through CCK8 detection, the maximum safe dose of GPC for Jurkat cells (stimulated for 24 hours and 48 hours, respectively) Figure 6 -A, Figure 6 -B), after stimulating Jurkat cells with safe dose of GPC, we found that GPC pretreatment activated Jurkat lymphocytes increased the RNA synthesis (P<0.01, Figure 6 -C) and protein secretion (P<0.05, Figure 6 -D) of IFNγ, which might be the intrinsic mechanism for the improved efficacy of ICIs during GPC pretreatment.

[0059] (3) C57 / BL6 mouse subcutaneous tumor xenograft model Male C57BL / 6 mice (Ensville, Chengdu, China), 6-8 weeks old, weighing 20 ± 2 g, were raised in the laboratory animal room of the School of Basic Medical Sciences and Forensic Medicine of Sichuan University, with a constant temperature of 20-26°C, a constant humidity of 55-65%, a 12-hour alternating day and night, 5 mice per cage, and SPF grade maintenance feed and sawdust bedding. The C57BL / 6 mouse subcutaneous tumor model was constructed (as described in Example 1) Figure 5 -A: I, the mice were divided into 4 groups, 5 in each group, 2 of which were given GPC sterilized water (0.06 mg / mL), and the other 2 used ordinary sterilized water, which was changed regularly (every 2-3 days), for at least 3 days; II, LLC cells in the best active state were cultured, expanded to a density of about 80%, and the suspension cells were collected and the adherent cells were digested to a single cell state, mixed evenly, counted 1.5×106 cells per mouse, 200 μL of fresh culture medium per mouse, and resuspended the cells; III, the right side of the mouse was shaved, disinfected with alcohol cotton ball, and insulin injection needle was used to suck 200 μL of cell suspension, after air was discharged, it was injected subcutaneously on the right side of the mouse, and observation was made to see if there was a subcutaneous bump and no leakage, which indicated that the injection was successful, and anti-PD1 drug treatment was given after 3 days; IV, the corresponding treatment was given according to the following grouping: a. Control group (Control): ordinary sterilized water, intraperitoneal injection of PBS; b. GPC group (GPC): GPC sterilized water, intraperitoneal injection of PBS; c. anti-PD1 treatment group (anti-PD1): ordinary sterilized water, intraperitoneal injection of anti-PD1, 200 μg per mouse per time, once every 3 days, a total of 5 times; d. GPC plus anti-PD1 treatment group (GPC and anti-PD1): GPC sterilized water, intraperitoneal injection of anti-PD1, 200 μg per mouse per time, once every 3 days, a total of 5 times; After injection of cells, the tumor size was measured every three days, the longest diameter (L) and the vertical diameter (W) were recorded, the tumor volume V = L x W2 x 0.52; the body weight of the mice was measured and recorded; all mice were sacrificed after 15 days.

[0060] (5) Small animal living body imaging (Living Image) After subcutaneous injection of tumor cells, small animal living body imaging detection was performed every 3-5 days (as described in Example 1) Figure 5 -B and Figure 5- C), respectively, 4 groups of mice were labeled, insulin injection needle intraperitoneal injection of D-luciferin (15 mg / mL) 250 μL, using isoflurane gas anesthesia machine induced anesthesia success, adjust to the level of maintenance anesthesia, select luciferase mode, exposure to collect images, anesthetic model is MATRX (VMR), small animal live imaging instrument model is IVIS spectrum.

[0061] (6) Histopathology H&E staining After the mice were anesthetized and sacrificed, the complete tumor tissue was stripped, and the heart, liver, kidney, spleen, and lung were dissected, photographed, and fixed, embedded, and stained with hematoxylin-eosin (H&E): - Tissue fixation: the tissue was immersed in 4% paraformaldehyde (PFA) for fixation at room temperature on a shaker overnight, for more than 24 hours; - Dehydration and embedding: dehydrated with ethanol and xylene with increasing concentration; then immersed in wax, placed in melted paraffin, each time for 1 h, a total of 3 times; -20 cooled and solidified, trimmed the wax block; - Sectioning: using a pathological tissue sectioning machine, cut into 3-5 μm tissue slices, flatten the tissue on warm water at 40°C, then paste on a glass slide, and bake the slide at 60°C; - De-waxing and hydration: hydrated with xylene and ethanol with decreasing concentration, and slowly washed with clean tap water for at least 2 times; - Staining: stained with hematoxylin for 5-20 minutes, and slowly washed with clean tap water for at least 3 times; counterstained with ammonia, and washed; stained with eosin for 5-15 minutes, and washed; - Dehydration, transparency, and mounting: dehydrated with ethanol and xylene with increasing concentration; mounted with neutral gum; - Image acquisition: automatic scanning was performed using a Pannoramic pathological section scanner, and the images were opened and viewed using SlideViewer, and tissue images were collected at different magnifications.

[0062] (7) Immunohistochemical staining (IHC) - Section de-waxing and hydration: hydrated with xylene and ethanol with decreasing concentration, and slowly washed with clean tap water for at least 2 times; - Antigen retrieval: immersed in citric acid antigen retrieval solution, heated to boiling in a microwave oven, incubated for 5-8 min, and naturally cooled to prevent drying; washed with PBS on a shaker for 5 min each time, a total of 3 times; - Blocking endogenous peroxidase: H2O2 solution at room temperature, avoid light for 25 min, PBS shaker washing, 5 min / time, a total of 3 times; - Punching and blocking: the tissue range was circled using an immunohistochemical pen, and BSA was blocked at room temperature for 30 min; - Primary antibody (anti-CD3): Prepare the primary antibody with PBS at a ratio of 1:100. Add water to the bottom of the dark box, place the slide, add the primary antibody, and incubate overnight at 4°C. - Secondary antibody: Wash with PBS on a shaker for 5 min each time, for a total of 5 times, then add HRP-labeled secondary antibody and incubate at room temperature for 60 min. -Staining: Wash with PBS on a shaker for 5 min each time, for a total of 5 times. Add DAB for staining. After a positive result appears under the microscope, rinse slowly with clean tap water. - Staining nuclei: Stain nuclei with hematoxylin for about 3 minutes, then rinse slowly with clean tap water; 1% hydrochloric acid alcohol for 5-10 seconds, rinse; ammonia solution to turn blue, then rinse. - Dehydration and sealing: Dehydration with ethanol and xylene in increasing concentration gradients, sealing with neutral resin, and fixing the edges with clear nail polish; - Image acquisition: Use the Pannoramic pathology slide scanner to automatically scan slides, open and view the images using SlideViewer, and acquire the images.

[0063] (8) Serum biochemical tests Before mice were anesthetized and euthanized, blood was collected from their eyeballs. After blood coagulation at room temperature, the blood was centrifuged at 2000 rpm for 10 min, and the serum was collected in another EP tube and stored at -80℃ for later testing. This study detected 12 biochemical indicators, including: ALB, ALP, ALT, AST, D-Bil, T-Bil, LDH, CK, CK-MB, CREA, UA, and UREA. The fully automated biochemical analyzer was used for testing: Pre-start checks and instrument self-test procedures were completed, the software was successfully accessed, and detection parameters were set. Reagent information was entered sequentially into the actual parameters, reagents were loaded, and the system waited for automatic scanning. Sample information was entered, and quality control and calibration parameters were set according to the instructions. Samples, quality control samples, and standards were placed sequentially for calibration, quality control, and sample testing. Under the premise of good quality control results and successful calibration, the data results were analyzed and exported.

[0064] like Figure 5 As shown, after GPC pretreatment, Anti-PD1 significantly improved the efficacy of tumor-bearing mice, delayed tumor growth (P<0.01, Fig. 5-B and Fig. 5-C), reduced tumor size (P<0.01, Fig. 5-D and Fig. 5-E), and prolonged mouse survival time (P<0.01, Fig. 5-F).

[0065] Statistical methods: R language and GraphPad Prism 9 statistical software were used for statistical analysis of data. Count data was expressed as frequency and percentage, and measurement data was expressed as mean ± SD. Kaplan-Meier method was used to draw survival curve, and single factor and multivariate Cox regression analysis was used to determine the independent factors affecting the prognosis of gastric cancer patients; and the hazard ratio (HR) and the corresponding 95% confidence interval (CI) were calculated.

[0066] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily mutually exclusive of one another. It is expressly understood that the embodiments described herein are merely examples from a multitude of possible embodiments which can be claimed.

[0067] Finally, it should be noted that: the embodiments disclosed by the embodiments of the present application are only the preferred embodiments of the present application, and are only used to illustrate the technical solutions of the present application, not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand; it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A marker for predicting the efficacy of tumor immunotherapy, characterized by: The marker is glycerophosphocholine in the detection sample, which is peripheral blood.

2. Use of the marker according to claim 1 in the preparation of a product for predicting the efficacy of tumor immunotherapy or enhancing the efficacy of tumor immunotherapy or modulating tumor immune microenvironment.

3. Use of the marker according to claim 2 for the manufacture of a product for predicting the efficacy of an immunotherapy of a tumor or for enhancing the efficacy of an immunotherapy of a tumor or for modulating the tumor immune microenvironment, characterized in that: The product for predicting the efficacy of tumor immunotherapy comprises a detection reagent or a detection kit or a diagnostic device, and the product is used for detecting the expression level of the marker glycerophosphocholine in a sample.

4. A method of detecting the efficacy of an immunotherapy of a tumor, characterized in that: Non-diagnostic purposes, including (1) an analysis module for determining the expression level of the marker in the test sample of the subject, and; (2) an evaluation module for determining the efficacy of tumor immunotherapy of the subject according to the expression level of the marker determined in (1); Wherein, the marker in (1) is glycerophosphocholine in peripheral blood.

5. The method of claim 4, wherein the method is for predicting the efficacy of an immunotherapy for a tumor. The subject is a tumor patient using ICIs immunotherapy.

6. The method of claim 4, wherein the method is for predicting the efficacy of an immunotherapy for a tumor. The test sample is peripheral blood.

7. The method of claim 4-6, wherein the method is for predicting the efficacy of immunotherapy of a tumor. The determination in the evaluation module is that the efficacy of tumor immunotherapy is positively correlated with the level of glycerophosphocholine in peripheral blood.

Images