Biomarker composition for predicting curative effect of kidney cancer immunotherapy and application of biomarker composition

Abstract

The invention discloses a biomarker composition for predicting the curative effect of kidney cancer immunotherapy and application of the biomarker composition. The biomarker combination comprises at least one of the following genes: MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6. The biomarker combination comprises at least one of MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6. The invention also provides application of a reagent for detecting the combination in preparation of a prediction or diagnosis product, a kit containing the corresponding detection reagent, and an immunotherapy drug resistance risk scoring model based on the combination. The model calculates a risk index through a specific equation. In addition, the invention also relates to an electronic device and a storage medium for realizing the model. The marker combination and the model can effectively distinguish the response of a kidney cancer patient to immunotherapy, provide a basis for an individualized treatment strategy, and have the advantages of high prediction accuracy and great clinical application potential.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of biological medicine. More specifically, it relates to a marker combination for predicting the efficacy of immunotherapy for renal cancer. BACKGROUND

[0002] Renal cell carcinoma (RCC) is a group of tumor diseases affecting the renal parenchyma, accounting for more than 80% of all renal malignancies. In recent years, its incidence has continued to rise, especially in high-income areas, and has become a major disease burden worldwide. About 70% of patients are diagnosed accidentally, and as many as 50% of patients are in the metastatic disease stage at the time of diagnosis. Local stage renal cancer is mainly treated with radical surgery or local treatment, while advanced renal cancer is mainly treated with drugs. Systemic drug treatment for advanced renal cancer has experienced a development process from cytokines to targeted drugs, and then to targeted combination immunotherapy. In recent years, immunotherapy has become increasingly important in the recommended position of authoritative renal cancer guidelines at home and abroad, and the treatment of advanced renal cancer has entered the era of drug combination therapy. Although immunotherapy combined with targeted therapy can significantly improve patient survival and quality of life, the choice of different drug combination modes is still a difficult and hot spot of current research.

[0003] Immune checkpoint blockade (ICB) therapy and its combination regimens have significantly improved the survival rate of patients with renal cell carcinoma, and are a major advance in the field of RCC treatment. However, there are significant differences in patients' responses to ICB, and more than 50% of renal cancer patients are resistant to immunotherapy, which highlights the urgency of developing reliable biomarkers to predict treatment efficacy. Traditional biomarkers, such as PD-L1 and tumor mutation burden, have some predictive value in other cancer types, but their effectiveness in RCC is uneven, and there is currently no standardized and validated biomarker that can be used to predict the efficacy of immunotherapy for renal cancer. Therefore, there is an urgent need to develop effective biomarkers in clinical practice to accurately identify potential beneficiaries of immunotherapy and guide individualized treatment decisions.

[0004] Renal cell carcinoma (RCC) is one of the most immunologically and vascularly invasive cancer types. Previous analyses of RCC genomic features were mostly based on bulk RNA sequencing or mass cytometry, which have limited resolution and are difficult to fully reveal the heterogeneity of the tumor microenvironment (TME). US20130101998A1 discloses detecting the level of biomarker S100A6 in a kidney sample to determine whether a subject has a kidney disease, i.e. acute renal failure or acute tubular necrosis. WO2010132676A1 discloses a method for assessing the risk of progression of chronic kidney disease (CKD), which comprises detecting biomarkers such as AFM, AXL, B2M, etc. to assess the risk of progression of CKD. Yoel Z Betancor et al. A three-gene expression score for predicting clinical benefit to anti-PD-1 blockade in advanced renal cell carcinoma discloses a three-gene expression score integrating HMGA1, NUP62 and ARHGAP42 transcripts to diagnose the clinical response of patients with advanced clear cell renal cell carcinoma (RCC) to nivolumab (an anti-PD-1 antibody) treatment. It can be seen that the prior art is still very limited in the study of biomarkers of renal cell carcinoma, and only scattered studies of markers have not been fully modeled. In addition, most transcriptome studies are based on single-region tumor biopsies, which fail to fully reflect the spatial variation of immune infiltration in patients. Therefore, a more complete and high-resolution description of the immune microenvironment of RCC is needed to accurately predict patient response and provide the basis for personalized treatment strategies. SUMMARY

[0005] The inventors dissected the immune microenvironment of RCC patients before and after ICB treatment in an unbiased and comprehensive manner using paired single-cell RNA sequencing. To quantify the variation of immune infiltration in patients, the inventors performed multi-region and multi-tissue sampling on each patient to obtain biological samples. A plurality of drug-resistant cell subpopulations related to the treatment efficacy of renal cancer were screened. Subsequently, in combination with transcriptome and immunotherapy efficacy information, a marker combination composed of 6 drug resistance related genes such as MARCO was identified. The present application constructs a comprehensive prediction model for the immunotherapy efficacy of renal cancer containing MARCO, AUP1, ARHGAP42, NUP62, ADCY5, S100A6 genes, and achieves good efficacy prediction performance. The biological sample is selected from one of tumor tissue biopsy sample, peripheral blood, plasma, serum, and urine.

[0006] The application provides a biomarker combination for predicting the efficacy of immunotherapy for kidney cancer, characterized in that the combination comprises at least one of MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6 genes.

[0007] Further, the kidney cancer is selected from renal clear cell carcinoma, renal papillary cell carcinoma or renal chromophobe carcinoma.

[0008] In a specific embodiment, the application further provides an application of a detection reagent in the preparation of a product for predicting or diagnosing the efficacy of immunotherapy for kidney cancer, characterized in that the detection reagent is used to detect the expression level of the aforementioned biomarker combination, and the detection reagent comprises at least one of a substance for detecting MARCO, a substance for detecting AUP1, a substance for detecting ARHGAP42, a substance for detecting NUP62, a substance for detecting ADCY5, and a substance for detecting S100A6.

[0009] The detection reagent is used to detect the gene sequence, mRNA expression level, protein expression level, or fragments or metabolites of the biomarker combination, and can be used to detect the methylation state of the gene (epigenetic level), detect specific protein cleavage fragments, or metabolize small molecules related to the function of these genes. Specifically, the detection reagent comprises primers or probes that specifically amplify the gene, or antibodies that specifically bind to the protein encoded by the gene, or gene chips, CRISPR-Cas system related detection elements (such as crRNA, Cas enzyme), aptamers (Aptamer) or standards for mass spectrometry detection.

[0010] In a specific embodiment, the application further provides a kit for predicting or diagnosing the efficacy of immunotherapy for kidney cancer, and the kit comprises reagents for detecting the expression level of MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6 genes. The reagents comprise primer pairs capable of specifically amplifying MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6 genes, such as the primer pairs shown in SEQ ID NO: 1-12.

[0011] The kit can also optionally comprise reagents for RNA or protein extraction, reverse transcription reagents, PCR amplification reagents, antibody diluents, chromogenic substrates, signal amplification systems, and information cards or electronic access credentials containing instructions for use of the risk score model and / or cutoff values.

[0012] In a more specific embodiment, the present application further provides a prediction model and / or detection method for identifying whether immunotherapy is effective for a kidney cancer patient, which is obtained by detecting the expression level of the aforementioned biomarker combination through transcriptome detection or RT-qPCR detection, and further, whether the immunotherapy is effective for the kidney cancer patient is determined by judging whether the immunotherapy drug resistance is successful. By detecting the expression level of the marker combination through transcriptome detection or RT-qPCR detection, the kidney cancer patients with effective and ineffective immunotherapy can be distinguished by using the risk score model, i.e., the present application provides a method for identifying the effectiveness of immunotherapy for kidney cancer patients, which is beneficial to provide the basis for personalized treatment strategies for kidney cancer patients and match the correct treatment plan.

[0013] The prediction model and / or detection method can be an immunotherapy drug resistance risk score model and / or detection method, and the model and / or detection method uses the following equation to calculate the risk index, i.e., risk-score = (0.0025 x MARCO expression level) + (0.488 x AUP1 expression level) - (0.265 x ARHGAP42 expression level) + (0.838 x NUP62 expression level) - (0.030 x ADCY5 expression level) + (0.195 x S100A6 expression level).

[0014] The calculated risk-score is compared with the cutoff value 0.89, if the risk-score > 0.89, it indicates that the patient is resistant to immunotherapy; if the risk-score ≤ 0.89, it indicates that the patient is effective for immunotherapy, wherein the model is constructed based on RNA-seq data; the calculated risk-score is compared with the cutoff value 0.5, if the model score > 0.5, it indicates that the immunotherapy is drug resistant, and if the model score <= 0.5, it indicates that the immunotherapy is effective, wherein the model is constructed based on RT-qPCR data. The model and / or detection method is based on the single cell transcriptome atlas of kidney cancer immunotherapy, mines the core factors related to early prediction of immunotherapy drug resistance, and screens the candidate molecules through high-throughput RNA sequencing (bulk RNA-seq) combined with single cell transcriptome technology (scRNA-seq). Based on LASSO regression, a marker combination model with dynamic response characteristics is constructed, the weight is optimized and its cross-queue universality is verified.

[0015] Further, the detection can be performed on the gene sequence, mRNA, protein or active form of the biomarker. Suitable sample types include, but are not limited to, tumor tissue biopsy samples, peripheral blood, plasma, serum, urine and other biological samples. Based on the prediction model constructed by the marker combination, not only can the effective and drug-resistant immunotherapy be distinguished, but also the progression-free survival (PFS) or overall survival (OS) of the patient can be predicted. The corresponding kit can be configured to include a complete set of components from sample processing to result interpretation.

[0016] In one specific embodiment, the present application provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the aforementioned prediction model when executing the program.

[0017] When the processor executes the program, it is used to implement the following steps: receiving the biomarker expression data input by the user; calling the stored risk score model for calculation; outputting the risk index and / or the corresponding efficacy prediction conclusion (such as 'immunotherapy effective' or 'immunotherapy resistant'); and optionally generating a visual report.

[0018] Advantages 1. The present application reveals the key cell subpopulations (such as TMEM176B+ tumor cells and MARCO+ TAM) related to renal cancer immunotherapy resistance and their characteristic genes by high-resolution single-cell transcriptome mapping, without bias, providing a solid data foundation for marker discovery.

[0019] 2. The marker combination is innovative and efficient: based on the above findings, the marker combination composed of MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6 screened by machine learning algorithms such as LASSO regression has a high synergistic prediction effect, and compared with single markers or traditional markers (such as PD-L1), it can more accurately reflect the heterogeneity of the tumor immune microenvironment.

[0020] 3. The model is strictly verified and has reliable performance: the risk score model (risk-score) and its cutoff value provided by the present application have been verified in an independent cohort, and exhibit high sensitivity and specificity for 1-year, 3-year and 5-year overall survival (OS), with stable and reliable prediction results, providing a quantitative basis for individualized treatment decisions.

[0021] 4 Form a complete technical solution, wide application prospect: The present application not only protects the marker combination itself, but also extends to its detection application, kit and electronic device, covering the whole process from sample detection to result analysis. The scheme can be applied to predict efficacy (effective / drug resistance) and long-term survival outcome (such as PFS, OS), and has great clinical transformation potential. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 . The atlas of RCC tumor cells. (A) UMAP plot of RCC tumor cells. (B) Dot plot showing the marker expression for defining tumor cell clusters. (C) UMAP plot comparison of tumor cells in the non-immunotherapy (NoICB), immunotherapy effective (ICB PR) and immunotherapy ineffective (ICB PD / SD) groups Figure 2 . Single-cell transcriptional atlas of myeloid cells in renal clear cell carcinoma immunotherapy cohort. (A) UMAP plot showing 13 myeloid cell subgroups. (B) Dot plot showing the marker expression for defining myeloid cell types. (C) Distribution of myeloid cells in various tissue types. (D) Bee swarm plot showing the abundance distribution of corrected log2 FC in PT and PN tissues in Nhood. (E) Bee swarm plot showing the abundance distribution of corrected log2 FC in ICB_PD / SD and ICB_PR in Nhood. (F) Box plot comparing the proportion of myeloid cells in tumor and normal tissues. (G) Box plot comparing the proportion of myeloid cells in the immunotherapy effective and ineffective groups Figure 3 . Hazard ratio (HR) and 95% confidence interval (CI) in Cox regression analysis Figure 4 . Screening of the best set of prognostic markers based on LASSO regression analysis (A) Process of selecting the best value of parameter λ in Lasso regression model. (B) LASSO coefficients of 25 markers in the modeling process.

[0023] Figure 5 . ROC curve analysis shows the performance of the risk score composed of 6 drug resistance related marker combinations in predicting 1-year (A), 3-year (B) and 5-year (C) overall survival (OS) in the training set and test set of RCC.

[0024] Figure 6 . Analysis of renal cancer immunotherapy efficacy in the training set and test set according to the best cutoff value.

[0025] Figure 7 . ROC curve analysis shows the performance of the classic immunotherapy marker PDL1 in predicting 1-year, 3-year and 5-year overall survival (OS) in the training set and test set of RCC.

[0026] Figure 8 The expression of molecules in the marker combination of the application in the immunotherapy effective (ICB PR) and immunotherapy resistant (ICB PD / SD) kidney cancer tissues of kidney cancer patients; wherein, from top to bottom, from left to right, are the expression of MARCO, AUP1, ARHGAP42, NUP62, ADCY5, S100A6; PR in the figure represents the immunotherapy effective kidney cancer tissue sample, and PD / SD represents the immunotherapy ineffective kidney cancer tissue sample. DETAILED DESCRIPTION

[0027] Example 1: Construction of kidney cancer immune microenvironment single cell map and identification of drug-resistant cell subpopulation In order to reveal the specific tumor microenvironment (TME) of renal cell carcinoma, the inventors obtained 18 surgical resection and biopsy tumor samples of 18 RCC patients, 6 tumor para-cancer normal tissue samples, and 2 peripheral blood mononuclear cell (PBMC) samples. Among the 18 RCC patients, 9 patients did not receive immunotherapy (No_ICB), 5 patients received immunotherapy and were effective (ICB PR), and 4 patients received immunotherapy and had poor effect (ICB PD / SD). Therefore, a total of 26 samples were obtained, and the samples were sequenced using the Droplet 10X platform to obtain single cell transcriptome data.

[0028] After strict quality control and screening, 103345 single cell transcriptomes were obtained. In order to determine the main group and subpopulation composition of the TME, unsupervised clustering was performed according to the lineage-specific classical markers defined in the literature, and 15 main cell subpopulations were determined, including tumor cells, lymphocytes (B / plasma cells, T cells and natural killer cells), myeloid cells (dendritic cells, monocytes, tumor-associated macrophages (TAM) and mast cells) and stromal cells (tumor-associated fibroblasts and endothelial cells).

[0029] In order to reveal the changes in the composition of the immune microenvironment of RCC patients with effective and ineffective immunotherapy, the tumor-immune-stromal cell composition of different types of samples was further compared. First, tumor cell clustering analysis revealed that TMEM176B+ tumor cells were a potential marker for ICB resistance (see Figure 1A-C). As a cation channel, TMEM176B might be a new immune regulatory target by activating inflammasome release, which might interfere with the CD8+ T cell-mediated tumor growth inhibition. Previous studies mainly focused on the role of TMEM176B in TAMs, while this study further confirmed the potential role of tumor cell-derived TMEM176B in ICB resistance (see Figure 1 A-C).

[0030] Meanwhile, the inventors performed unsupervised clustering analysis on 28340 myeloid cells to further determine the functional subpopulations of major myeloid cell lineages (see Figure 2 A). Three DC subsets were identified based on high expression of HLA-DRs and low expression of CD14: plasmacytoid DC (pDC: LILRA4 / GPR183), cDC2 (CD1C / FCER1A), and cDC1 (IDO1 / LAMP3) (B). Figure 2 B). The remaining cell subsets were identified as monocytes / macrophages according to their high expression of CD14, CD68, CD163, and MRC1. Among them, a subset of macrophages expressing the resident-like markers F13A1, MRC1, and FOLR2 was noted as tissue-resident macrophages (TRMs; C1). A subset of cells expressing TREM2, MARCO, and SPP1 was labeled as tumor-associated macrophages (TAMs) (C3 and C4) (B). Figure 2 B).

[0031] Cellular tissue distribution analysis suggested that macrophages were enriched in kidney cancer tissues, LST1+non-classical monocytes were enriched in normal kidney, and S100A8+classical monocytes were enriched in peripheral blood (C). Figure 2 C-D). Further analysis combined with immunotherapy efficacy information found that MARCO+TAMs were not only enriched in kidney cancer tissues (PT vs. PN: p = 0.02), but also significantly enriched in patients with immunotherapy resistance (PD vs. PR: p = 0.01). In addition, we found that the proliferation ability of macrophages was also significantly enhanced in patients with immunotherapy resistance (p = 0.04) (C). Figure 2 C-G).

[0032] Example 2: Screening of drug resistance-related characteristic genes The experimental results of Example 1 suggest that the significantly accumulated TMEM176B+tumor cells and MARCO+TAMs in the renal cancer tissues can be the core cell subpopulation mediating ICB resistance, forming a niche with strong pro-cancer activity and immunosuppressive potential, which highly expresses 76 pro-cancer and / or immunosuppressive molecules such as TMEM176B and MARCO, and is highly related to immunotherapy resistance. In order to establish an effective predictive model for the efficacy of immunotherapy for renal cancer, a model based on the transcriptome sequencing was developed and verified based on the markers of TMEM176B+tumor cells and MARCO+TAMs, for improving the predictive efficiency of the efficacy of immunotherapy for renal cancer. In order to clarify the transcriptome expression profile of renal cancer patients with immunotherapy resistance, 181 tumor tissue samples (samples are surgical resection or fine needle biopsy samples) of renal cancer patients receiving immunotherapy were collected for high-throughput transcriptome sequencing. The adapter and primer redundant sequences of the double-end transcriptome data were removed using Cutadapt software to obtain clean data. STAR version 2.7 software was used to construct a reference genome index, and STAR version 2.7 software was used to align the double-end clean transcriptome data to the reference genome. RSEM version 1.3 software was used to count the number of reads aligned to each gene and to perform gene quantification. Then, the FPKM (Fragments Per Kilobase Million) of each gene was calculated according to the gene length and the number of reads aligned to the gene, and the final expression matrix was obtained. In order to identify the differentially expressed resistance-related genes significantly related to efficacy, Cox proportional hazards regression analysis was performed on 76 resistance-related characteristic genes by the survival analysis package of R language, and 25 characteristic genes significantly related to the prognosis of immunotherapy were screened (P value <0.05 and hazard ratio (HR) ≠ 1 as the significant standard) Figure 3 ).

[0033] Example 3: Construction and verification of predictive biomarker combination 181 patients with renal cell carcinoma were randomly divided into a training set (n=127) and a test set (n=54) in a 7:3 ratio. Modeling was performed on the training set. To screen the optimal set of predictive factors for immunotherapy efficacy, LASSO and Cox hazard regression algorithms were used to further refine the optimal biomarker combination for the immunotherapy efficacy prediction model on the training set. 10-fold cross-validation was used to evaluate model performance and screen variables, identifying the variable combination required to form the optimal model (i.e., λ=lambda.min), determining that the model reached its optimum with 6 variables. After cross-validation, the LASSO-COX regression algorithm determined the optimal predictive model consisting of 6 variables: MARCO, AUP1, ARHGAP42, NUP62, ADCY5, and S100A6. The risk-score for immunotherapy resistance in this model is calculated as follows: risk-score = (0.0025 × MARCO expression level) + (0.488 × AUP1 expression level) - (0.265 × ARHGAP42 expression level) + (0.838 × NUP62 expression level) - (0.030 × ADCY5 expression level) + (0.195 × S100A6 expression level). The coefficients and selection process for each variable are as follows... Figure 4 As shown, the model exhibits good predictive efficacy for immunotherapy, demonstrating high sensitivity and specificity (AUC>0.90) for 1-year, 3-year, and 5-year overall survival (OS), and the prediction results are stable and reliable. Figure 5 (See Table 1). Furthermore, the optimal cutoff value was determined to be 0.89 using the maximum Youden index method. Patients were divided into high-risk and low-risk groups based on the optimal cutoff value of 0.89 for Kaplan-Meier survival analysis. The results are shown in the figure. Patients in the high-risk group had significantly shorter progression-free survival (PFS) and overall survival (OS) than those in the low-risk group (Log-rank test, p<0.001), indicating that the model can effectively differentiate patient prognoses. Figure 6 Furthermore, the model's performance significantly outperformed that of the classic immunotherapy biomarker PDL1. The inventors used PDL1 as a classic comparative biomarker, and the results showed that the average AUC on the training set was <0.6, and the average AUC on the test set was <0.8. Figure 7 As can be seen, the AUC values ​​of PDL1 in the training set were close to or lower than 0.6 (0.506–0.553), which is close to random guessing. Although there was some improvement in the test set (0.687–0.815), it was still significantly lower than the 6-gene immune prediction model constructed in this invention (AUC>0.90), indicating that the ability of PDL1 alone as a predictive biomarker for the efficacy of renal cell carcinoma immunotherapy is limited and far inferior to the multi-gene combination model proposed in this invention.

[0034] Table 1 Model performance Example 4: Expression levels of RNAs in marker combinations in immunotherapy- effective (ICB PR) and immunotherapy-resistant (ICB PD / SD) kidney cancer tissues Figure 8 To further supplement and verify the main model, this embodiment demonstrates the application of the model on the RT-qPCR platform, and emphasizes the need to recalibrate the cutoff value for specific detection methods. This embodiment independently verifies the expression differences of 6 genes (MARCO, AUP1, ARHGAP42, NUP62, ADCY5, S100A6) in the immunotherapy effective and ineffective groups by RT-qPCR, confirming the reliability of these genes as biomarkers, which is not limited by the detection platform. RT-qPCR is faster, more economical, and more suitable for routine clinical detection.

[0035] 1. Experimental method 26 cases of immunotherapy effective (ICB PR) and 26 cases of immunotherapy resistant (ICB PD / SD) renal cancer tissue samples (from Peking University Cancer Hospital) were selected, TRIzol™ reagent was used to extract sample RNA, and ReverTra AceTM qPCR RT Kit (Toyobo) was used for reverse transcription of cDNA at 37°C and 98°C for 15 min and 5 min. SYBR Green Realtime PCR Master Mix (Toyobo) was used for RT-qPCR reaction to detect the expression of RNA in the marker combination. The sequences of the fluorescent quantitative PCR primers for detecting RNA molecules in the marker combination are shown in SEQ ID NO: 1-12 (Table 2).

[0036] Table 2 Fluorescent quantitative PCR primers 2. Experimental results The expression of the RNA molecules in the marker combination in the immunotherapy effective (ICB PR) and immunotherapy resistant (ICB PD / SD) renal cancer tissues of the renal cancer patients is shown in Figure 8 Figure 8 From top to bottom and left to right in the figure are the expression of MARCO, AUP1, ARHGAP42, NUP62, ADCY5, and S100A6, respectively. PR in the figure represents immunotherapy effective renal cancer tissue samples, and PD / SD represents immunotherapy ineffective renal cancer tissue samples. ​ ​It can be known that MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6 have different expressions in the renal cancer tissues of the renal cancer patients with effective immunotherapy (ICB PR) and immunotherapy resistance (ICB PD / SD). Among them, the expression amount of MARCO, AUP1, NUP62 and S100A6 in the renal cancer immunotherapy ineffective group is higher than that in the immunotherapy effective group, the expression amount of ARHGAP42 and ADCY5 in the renal cancer immunotherapy effective group is higher than that in the immunotherapy ineffective group, and the differences have statistical significance. The same risk score formula as in Example 3 is used to calculate the gene expression amount obtained by RT-qPCR, and the model score is obtained. Since the expression amount value range detected by the RT-qPCR platform is different from the RNA-seq data, the optimal cutoff value determined by the maximum Youden index method is 0.5; the model score > 0.5 indicates immunotherapy resistance, and the model score <= 0.5 indicates effective immunotherapy. In subsequent application, the RT-qPCR detection method should be used to detect the cutoff value.

[0037] In addition, the inventors' internal validation analysis shows that the prediction model constructed based on the six-gene marker combination of the application has a significantly better prediction performance (AUC > 0.90) than the simplified models constructed using any single gene (Single Marker), random double-gene combination (Double-Marker Pair) or three to five gene combination (Triple-Marker Panel). This confirms that the six genes MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6 have an irreplaceable synergistic effect in predicting renal cancer immunotherapy resistance, and the comprehensive score model composed of the six genes can capture more comprehensive tumor microenvironment information, thereby achieving prediction accuracy and robustness far beyond a small number of gene combinations.

Claims

1. A combination of biomarkers for predicting the efficacy of immunotherapy for renal cell carcinoma, characterized in that, The combination contains at least one of the genes MARCO, AUP1, ARHGAP42, NUP62, ADCY5, and S100A6.

2. The biomarker combination according to claim 1, wherein the renal cell carcinoma is selected from clear cell renal carcinoma, papillary cell renal carcinoma, or chromophobe renal carcinoma.

3. The use of a detection reagent in the preparation of products for predicting or diagnosing the efficacy of immunotherapy for renal cell carcinoma, characterized in that, The detection reagent is used to detect the expression level of the combination of biomarkers according to any one of claims 1-2, wherein the detection reagent includes at least one of the following: a substance for detecting MARCO, a substance for detecting AUP1, a substance for detecting ARHGAP42, a substance for detecting NUP62, a substance for detecting ADCY5, and a substance for detecting S100A6.

4. According to claim 3, the detection reagent is used to detect the gene sequence, mRNA expression level, protein expression level, or fragments or metabolites of the biomarker combination, and can be used to detect the methylation status of genes, detect specific protein splice fragments, or small metabolic molecules related to the function of these genes; specifically, the detection reagent includes primers or probes that specifically amplify the gene, or antibodies that specifically bind to the protein encoded by the gene, or gene chips, detection elements related to the CRISPR-Cas system, aptamers, or standards for mass spectrometry detection; the detection elements related to the CRISPR-Cas system can be crRNA, Cas enzyme, etc.

5. A kit for predicting or diagnosing the efficacy of immunotherapy for renal cell carcinoma, characterized in that, The kit contains reagents for detecting the expression levels of the MARCO, AUP1, ARHGAP42, NUP62, ADCY5, and S100A6 genes.

6. The kit according to claim 5, wherein the reagent comprises primer pairs capable of specifically amplifying the MARCO, AUP1, ARHGAP42, NUP62, ADCY5 and S100A6 genes.

7. A predictive model for identifying the effectiveness of immunotherapy in patients with renal cell carcinoma, obtained by detecting the expression levels of the combination of biomarkers described in claim 1 through transcriptome analysis or RT-qPCR, wherein the effectiveness of immunotherapy in patients with renal cell carcinoma is determined by assessing whether immunotherapy resistance has been successfully established.

8. The prediction model according to claim 7, wherein the prediction model comprises an immunotherapy resistance risk scoring model, the model calculates the risk index using the following equation, namely risk-score = (0.0025 × MARCO expression level) + (0.488 × AUP1 expression level) - (0.265 × ARHGAP42 expression level) + (0.838 × NUP62 expression level) - (0.030 × ADCY5 expression level) + (0.195 × S100A6 expression level).

9. In the prediction model according to claim 7 or 8, the calculated risk-score is compared with a cutoff value of 0.

89. If the risk-score > 0.89, it indicates that the patient is resistant to immunotherapy; if the risk-score ≤ 0.89, it indicates that the patient is effective in immunotherapy, wherein the model is constructed based on transcriptome (RNA-seq) detection data; the calculated risk-score is compared with a cutoff value of 0.

5. A model score > 0.5 indicates resistance to immunotherapy, while a model score <= 0.5 indicates effectiveness in immunotherapy, wherein the model is constructed based on RT-qPCR data.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the prediction model according to any one of claims 7-9.

Images