A set of biomarkers for predicting the benefit of immunotherapy of nasopharyngeal carcinoma and application thereof

Abstract

The application provides a set of biomarkers for predicting the benefit of nasopharyngeal carcinoma immunotherapy and application thereof, transcriptome data analysis of non-metastatic initial treatment of locally advanced nasopharyngeal carcinoma patients obtains a three-classification model using 318 genes as features based on nearest shrinkage centroid classification method, in the model, the event-free survival rate of patients of type 1 is significantly improved after receiving anti-PD-1 immunotherapy compared with the standard treatment group, the model can reflect the biological heterogeneity of nasopharyngeal carcinoma patients, can more accurately predict the immunotherapy benefit of nasopharyngeal carcinoma patients, and better guide clinical medication. The method for predicting the benefit of nasopharyngeal carcinoma immunotherapy can detect paraffin tissue samples, and does not depend on large sample quantity clustering, can type patients at a single sample level, has high feasibility and is easy to popularize, and has important significance for nasopharyngeal carcinoma immunotherapy.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, and in particular to a group of biomarkers for predicting the benefit of immunotherapy for nasopharyngeal carcinoma and their applications. Background Technology

[0002] Nasopharyngeal carcinoma (NPC) is an aggressive malignant tumor of the head and neck. In recent years, with the rise and application of immunotherapy, especially anti-PD-(L)1 drugs, immunotherapy combined with radiotherapy and chemotherapy has gradually become the main treatment modality for NPC. Currently, immunotherapy has become the first-line treatment for recurrent and metastatic NPC, and large-scale phase 3 randomized controlled trials have demonstrated that immunotherapy combined with radiotherapy and chemotherapy can significantly improve event-free survival in patients with newly diagnosed locally advanced NPC. However, approximately 15% of NPC patients do not benefit from immunotherapy.

[0003] Currently, the prediction of immunotherapy benefit is mainly based on tumor tissue PD-L1 protein expression levels, tumor mutation burden (TMB), and microsatellite instability high (MSI-H). However, the low mutation rate in nasopharyngeal carcinoma limits the application of indicators such as TMB and MSI-H. On the other hand, recent subgroup analyses of clinical trials have found that patients regardless of PD-L1 expression levels can benefit from immunotherapy. Therefore, several immunotherapy efficacy prediction indicators commonly used in other cancer types cannot accurately predict the immunotherapy benefit in nasopharyngeal carcinoma patients. In the field of nasopharyngeal carcinoma, some previous studies have reported on immunophenotyping to predict the short-term efficacy of immunotherapy. However, the immunophenotyping cohorts used in these studies were small, and due to phenomena such as pseudoprogression in immunotherapy, short-term efficacy may not necessarily translate into long-term survival benefit. Therefore, effectively screening patients with nasopharyngeal carcinoma phenotypes who can benefit from immunotherapy in the long term, and guiding clinical medication, is of great significance.

[0004] Predicting and identifying nasopharyngeal carcinoma patients who can benefit from immunotherapy and providing them with effective treatment can significantly improve the prognosis of these patients, while also having significant socioeconomic benefits. Summary of the Invention

[0005] The present invention aims to at least solve one of the aforementioned technical problems existing in the prior art. Therefore, the object of the present invention is to provide a set of biomarkers for predicting the benefit of immunotherapy for nasopharyngeal carcinoma and their applications.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect, the present invention provides a biomarker for predicting the benefit of immunotherapy for nasopharyngeal carcinoma, said biomarker comprising the following genes: BAD, CD99, CFLAR, ARF5, SLC25A5, RPS20, UQCRC1, POLR3B, CAPN1, CD74, SAMD4A, FHL1, BIRC3, STRAP, LCP2, HEBP2, PRSS8, KCNH2, RPL18, GPC1, CLEC2D, TGFBR3, ATP1B3, ENO1, SLC25A3, FGFR1, ITGB5, PLOD1, GSTP1, SF3B2, RPLPO, FXYD3, PHGDH, PTGS1. HSP90AB1, NDUFB7, GADD45B, 1A, JAK3, RPL18A, EIF3B, RBM28, FKTN, DVL1, LZTS2, CCDC6, RPL28, RPL19, SYNGR2, LGALS3BP, YWHAE, FBXW7, MDK, CD81, ATP5F1B, PTGES3, KRT18, LTBR, ​​G APDH, USP5, TPI1, FBXL4, BICRAL, CUL9, CDH6, DBN1, AMOTL2, PPM1G, RPS15, STAT1, IGFBP2, ST6GALNAC5, SLC2A1, EBNA1BP2, CDC20, PRDX1, STAG1, PPL, EPCAM, TMEM131L, PSPC1, OBSL1, H2BC11, MYRF, MT2A, PAX8, SNRPB, PRDX5, FGFRL1, ATF4, CDKN1C, TOMM40, BST2, ACTN4, LSM4, ASS1, AOC3, TRIM5, TRIM22 , H3-3B, RBM38, TBC1D8B, WDR74, FKBP11, CD63, GDF11, COX5B, ANKRD36, HUS1, RPS6, RPLP1, DIAPH3, NLRC5, RPS2, ARHGDIA, EIF4A3, ANAPC11, RPL13A, RP L11, CCN1, RPS8, CERS2, RHOB, RPS27A, SNRNP200, CTDSP1, RPL32, TBC1D32, OGT, RPL10, AOPEP, STOM, GSN, RPL7A, RPS3, ATM, FAU, ALDOA, LYPD1, NDUFC2,SCLT1、ASAP2、SLC25A4、ZFP36L2、ADAMTS1、RHOC、CFAP70、EEF1A1、UBE2L6、 EIF4A2、SLC34A2、WASF2、H2BC5、NRG2、TAGLN2、ADAMTS4、CDC42SE2、PSMB4、C TBP1, ADAR, LY6E, TONSL, PPP1R16A, MGAT4B, RPL8, SRSF2, IGSFF8, FBLN2, NMNAT3, TKT, WDR82, CTSB, ZNF503, DIPK1B, METTL3, PSMC3, GPT2, SPINT1, CCT2 B2M, PPIB, PHB1, GPX4, EEF2, PLIN4, OTUB1, FTH1, SLC3A2, RPSA, H1-4, SEPTIN2, TAP1, LRATD2, SLC25A6, RAB3B, NPR1, SCN9A, PLEKHA2, FASN, TMED10, K RT8、KIF5B、RPS9、RPS21、BSG、GNB2、CFL1、OR10K1、VEGFB、SULT1B1、EIF1、R NF213、RPL15、YIF1A、PSMD2、COX8A、METRNL、SHMT1、ARIH2、JUN、PNPLA2、CD1 51、GRINA、TUFM、CYC1、CALR、PUF60、H2BC4、SCRIB、C4orf50、RFX7、UNC5C、S HMT2、EPHB3、RPS17、CRIP2、COL18A1、ACTG1、H2AC20、KCNK12、H2AC21、TACST D2, H1-5, HSF1, ARL15, HGS, P4HB, SLC52A2, BCAP31, NPIBB4, NDUFA13, KRT5, BCAM, PTMA, TUBB4B, SBK1, H2AX, CLDN4, TUBB, XRCC6, GM2A, KPNA5, PCBP2, C DC42SE1、PSAP、RPL37A、CFD、GPAA1、PLCG2、RPL12、C1orf122、SFI1、RPL10A 、SELENOT、H2BC18、CHIC1、HLA-DRA、HSPA1B、HLA-C、HLA-E、HLA-A、IGKC、IGH A1、ATF6B、DDAH2、PHB2、CEBPD、RPL41、RPS18、HLA-B、CCDC200、DDOST、EEF1 G、TMEM178B、MRPL12、TSNAX-DISC1、H4C15、H4C14、NBPF26、PLD5P1、H2AC18、XNDC1N-ZNF705EP-ALG1L9P.

[0007] A second aspect of the present invention provides a method for establishing a model to predict the benefit of nasopharyngeal carcinoma immunotherapy, the method comprising the following steps: (1) Collect gene expression data from nasopharyngeal carcinoma patients and screen for metabolism-related genes; (2) Based on the expression profile of metabolism-related genes obtained in step (1), perform unsupervised cluster analysis to identify metabolic subtypes; (3) Using the genes obtained in step (1) as the initial feature set, and based on the metabolic subtypes in step (2), the Boruta algorithm is used to evaluate and screen the features in the initial feature set to obtain the gene set. (4) Using the gene set obtained in step (3) as input, the model for predicting the benefit of nasopharyngeal carcinoma immunotherapy is trained based on the centroid classification method.

[0008] In some embodiments of the present invention, the unsupervised clustering analysis classifies patients into three types.

[0009] In some embodiments of the present invention, the unsupervised clustering analysis includes using at least one of the following: a centroid partitioning algorithm, hierarchical clustering, and K-means clustering.

[0010] In some embodiments of the invention, the patient includes a patient receiving at least one of chemotherapy and immunotherapy.

[0011] In some embodiments of the invention, the patient includes a non-metastatic, treatment-naïve patient with locally advanced nasopharyngeal carcinoma.

[0012] In some embodiments of the invention, the patient includes a patient receiving gemcitabine and cisplatin induction chemotherapy combined with concurrent cisplatin chemoradiotherapy.

[0013] In some embodiments of the present invention, the patients include those receiving gemcitabine and cisplatin induction chemotherapy combined with concurrent cisplatin chemoradiotherapy and anti-PD-1 immunotherapy.

[0014] In some embodiments of the present invention, the centroid-based classification method includes at least one of the nearest centroid classification method and the nearest contracting centroid classification method.

[0015] In some embodiments of the present invention, the Boruta algorithm evaluates the importance of feature iterations based on a random forest model.

[0016] In some embodiments of the present invention, the classification includes dividing patients into three types: type 1 patients have relatively rich immune infiltration compared to other types, type 2 patients have relatively high matrix scores compared to other types, and type 3 patients have relatively high tumor purity compared to other types.

[0017] In some embodiments of the present invention, the gene set consists of the markers described above.

[0018] In some embodiments of the present invention, the immunotherapy includes treatment with immune checkpoint inhibitors.

[0019] In some embodiments of the present invention, the immune checkpoint inhibitor targets include at least one of PD-1, PD-L1, CTLA-4, BTLA, TIM-3, LAG-3, TIGIT, LAIR1, 2B4 and CD160.

[0020] A third aspect of the present invention provides the application of the biomarkers described above in constructing models for predicting the benefits of immunotherapy for nasopharyngeal carcinoma.

[0021] A fourth aspect of the present invention provides the application of the predictive model for the benefit of nasopharyngeal carcinoma immunotherapy established by the method described above in the guidance of nasopharyngeal carcinoma medication.

[0022] A fifth aspect of the invention provides the use of a substance for detecting the biomarkers described above in the preparation of a product for predicting the degree of benefit from immunotherapy for nasopharyngeal carcinoma.

[0023] A sixth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs the function of establishing a predictive model for the benefit of nasopharyngeal carcinoma immunotherapy by the establishment method described above.

[0024] The beneficial effects of this invention are: This invention provides a set of biomarkers for predicting the benefit of immunotherapy in nasopharyngeal carcinoma (NPC) and their applications. Transcriptome data analysis of non-metastatic, treatment-naïve locally advanced NPC patients yielded a three-class classification model using 318 gene features and the recent contraction centroid classification method. In this model, type 1 patients showed significantly improved event-free survival after receiving anti-PD-1 immunotherapy compared to the standard treatment group; type 2 patients had the best prognosis and could be exempt from anti-PD-1 immunotherapy; and type 3 patients had the worst prognosis, with no significant benefit in event-free survival after receiving anti-PD-1 immunotherapy compared to the standard treatment group, requiring combination therapy to improve efficacy. This model reflects the biological heterogeneity of NPC patients, can more accurately predict the benefit of immunotherapy, and better guide clinical medication. The method for predicting the benefit of immunotherapy in NPC can utilize paraffin-embedded tissue samples and does not rely on large-sample clustering; it can classify patients at the single-sample level, making it a highly feasible and easily applicable method.

[0025] Meanwhile, this invention has developed a web-based tool for predicting molecular subtypes of nasopharyngeal carcinoma. It can predict subtypes by uploading transcriptome sequencing gene expression matrix data of nasopharyngeal carcinoma samples, which is an easy method to use in clinical practice. Attached Figure Description

[0026] Figure 1 To identify the results of unsupervised consensus clustering of patients in the cohort. A is a consensus clustering heatmap; B is an immune infiltration feature map of tumor purity; C is an immune infiltration feature map of immune score; D is an immune infiltration feature map of matrix score; E is a metabolic pathway feature map.

[0027] Figure 2 This is a score map showing the importance of each feature to the model calculated iteratively by the Boruta algorithm.

[0028] Figure 3 The classification performance of a classifier based on the nearest centroid classification method was tested on an internal validation set. A is a confusion matrix showing the accuracy of the classifier; B is the area under the subject response curve of the multi-class model.

[0029] Figure 4 These are subtype feature maps after applying a classifier based on the nearest centroid shrinkage classification method to an external validation set. A is the metabolic pathway feature map; B is the immune infiltration feature map.

[0030] Figure 5 The chart shows event-free survival curves for patients with different immune subtypes. A is the cohort plot from the CONTINUUM trial; B is the cohort plot from the DIPPER trial.

[0031] Figure 6A graph showing the efficacy of 318 genes included in the molecular subtyping and classification model of nasopharyngeal carcinoma in predicting the benefit of immunotherapy.

[0032] Figure 7 A web-based tool for molecular subtyping and classification models of nasopharyngeal carcinoma. Detailed Implementation

[0033] The present invention will be further described in detail below through specific embodiments. Unless otherwise specified, the raw materials, reagents, or apparatus used in the embodiments and comparative examples are all available from conventional commercial sources or can be obtained by existing technical methods. Unless otherwise specified, the test or experimental methods are conventional methods in the art.

[0034] Example 1 This embodiment provides the establishment of a molecular subtyping and classification model for nasopharyngeal carcinoma. The specific experimental steps are as follows.

[0035] The case screening is as follows: This embodiment uses baseline paraffin-embedded tissue samples from 187 non-metastatic, treatment-naïve patients with locally advanced nasopharyngeal carcinoma collected in a phase 3 randomized controlled clinical trial (CONTINUUM trial) as a biomarker discovery and internal validation cohort. These patients were randomly assigned to two groups: the control group received standard treatment: gemcitabine plus cisplatin induction chemotherapy combined with cisplatin concurrent chemoradiotherapy (n=88), and the experimental group received standard treatment plus full-course anti-PD-1 immunotherapy (n=99). RNA was extracted from the paraffin-embedded tissue samples for whole-transcriptome sequencing.

[0036] This embodiment uses baseline paraffin-embedded tissue samples from 220 patients with non-metastatic, treatment-naïve locally advanced nasopharyngeal carcinoma collected in another phase 3 randomized controlled clinical trial (DIPPER trial) as an external independent validation cohort. These patients all received standard gemcitabine plus cisplatin induction chemotherapy combined with concurrent cisplatin chemoradiotherapy, and were subsequently randomized into two groups: a control group (n=104) receiving observation, and an experimental group receiving adjuvant anti-PD-1 immunotherapy (n=116). The paraffin-embedded tissue samples were used to extract RNA and perform whole transcriptome sequencing.

[0037] Establishment of molecular subtyping and classification models for nasopharyngeal carcinoma: This embodiment uses the CONTINUUM experimental cohort as the discovery cohort, collecting 1681 metabolism-related genes from 84 metabolism-related pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG; www.kegg.jp) database. Genes expressed in at least 70% of the samples were screened, and the bimodal index of each gene was calculated using the BimodalIndex package in R (Wang J, Wen S, Symmans WF, Pusztai L, Coombes KR. The bimodality index: A criterion for discovering and ranking bimodal signatures from cancer gene expression profiling data. Cancer Informatics, 2009 Aug 5; 7:199–216). Furthermore, the top 70% of genes with the highest bimodal index were selected for genotyping. This embodiment uses the Partitioning Around Medoids (PAM) algorithm for unsupervised clustering (R language ConsensusClusterPlus package, ConsensusClusterPlus function: maxK=10, reps=500, pItem=0.7, pFeature=1, clusterAlg='pam', distance='euclidean', finalLinkage='ward.D2'), and the results are as follows. Figure 1 As shown in Figure A, nasopharyngeal carcinoma patients can be automatically classified into three types, each with unique metabolic and immune characteristics. Regarding the immune microenvironment, type 1 has abundant immune infiltration (…). Figure 1 (B), type 2 matrix score is high ( Figure 1 (C), high purity of type 3 tumors ( Figure 1 In terms of metabolic pathways, type 1 is enriched in relatively few metabolic pathways, type 2 is enriched in unsaturated fatty acid metabolic pathways such as arachidonic acid metabolism, and type 3 is enriched in multiple metabolic pathways such as amino acid, fatty acid, and energy metabolism. Figure 1 (E).

[0038] To construct a model capable of classifying individual samples and further simplify and optimize the model, this embodiment randomly divides the discovery queue into a training set (75%, n=140) and an internal validation set (25%, n=47) for subtype classification model development. In the training set, the metabolic subtypes of the three patient types mentioned above are used as classification targets. This embodiment employs the Boruta algorithm for feature selection (R language Boruta package Boruta function: maxRuns = 200). The Boruta algorithm iteratively evaluates the importance of all expressed genes based on a random forest classification model. The random forest classification model is trained using the training set data, using the default parameters of the Boruta package. By comparing random noise features with the original features, it selects the best combination from the feature subset, filtering out the features most important for prediction and identifying and removing redundant features. Figure 2 The 318 genes automatically selected by the Boruta algorithm, and their importance to the model, are shown in Table 1 below: Table 1. Importance of the 318 genes selected based on the Boruta algorithm to the model contribution.

[0039] Then, the 318 most important genes were included, and a three-class classification model (nasopharyngeal carcinoma subtype classification model) was trained on the training set using the nearest centroid classification method (R language pamr package: the pamr.train function is used to train the model, the pamr.cv function is used to select the value that minimizes the classification error as the threshold parameter of the pamr.predict function, and the pamr.predict function is used to predict the sample subtype). The performance of the classification model was verified on the internal validation set, and the classification model's prediction results were in good agreement with the actual results. Figure 3 In the middle A group, the average area under the response curve was 0.92 ( ). Figure 3 (B)

[0040] External verification of the marker: Furthermore, this embodiment also verified whether the classification model could reproduce the molecular subtyping of nasopharyngeal carcinoma in an external independent cohort (DIPPER cohort). The results showed that the classification model could also divide the DIPPER cohort into three types, all of which had similar metabolic and immune characteristics to the discovery cohort. Figure 4 (A and B in the middle).

[0041] Survival analyses of patients with types 1, 2, and 3 in the DIPPER cohort and CONTINUUM trial yielded the following results: Figure 5 As shown, in the discovery and validation cohort, patients with type 1 disease who received anti-PD-1 immunotherapy had significantly improved event-free survival compared to the standard treatment group; while patients with type 2 or 3 disease did not show statistically significant differences in event-free survival regardless of whether they received anti-PD-1 immunotherapy.

[0042] To further compare the efficacy of this classification model with that of the 318 genes included individually in predicting the benefit of nasopharyngeal carcinoma immunotherapy, this embodiment calculated the p-value of the interaction between biomarker grouping factors and treatment grouping factors. Specifically, after including the biomarker grouping factor, treatment grouping factor, and their interaction term (biomarker:treatment) into a Cox regression model, the p-value of the interaction term was tested. If the p-value of the interaction term was significant, it indicated that there was an interaction between the biomarker and the treatment effect, that is, the biomarker affects the treatment efficacy. For example, if the biomarker is positive (the biomarker test result is higher than the normal range), the patient can benefit from the treatment; while if the biomarker is negative (the biomarker test result is lower than the normal range), the patient cannot benefit from the treatment. Therefore, the biomarker can be used as a predictive biomarker for treatment efficacy, rather than a simple prognostic biomarker (regardless of whether the biomarker is positive or negative, the patient can benefit from the treatment).

[0043] The results are as follows Figure 6 As shown, this classification model has the highest predictive value for the efficacy of immunotherapy, outperforming the effects of single genes included in the model.

[0044] In summary, the nasopharyngeal carcinoma subtype classification model of the present invention can reflect the biological heterogeneity of nasopharyngeal carcinoma tissues in different patients and their different responses to immunotherapy. It can accurately predict the benefits of immunotherapy for nasopharyngeal carcinoma patients and better guide clinical medication.

[0045] Example 2 This embodiment provides a computer-readable storage medium storing computer-executable instructions for executing the nasopharyngeal carcinoma subtype classification model obtained in Embodiment 1.

[0046] Example 3 This embodiment is based on the nasopharyngeal carcinoma subtype classification model obtained in Embodiment 1, and uses R Shiny to create the following... Figure 7The web-based tool shown has a built-in nasopharyngeal carcinoma subtype classification model trained in Example 1. Users can upload gene expression matrices obtained from transcriptome sequencing of paraffin-embedded tissue RNA samples from nasopharyngeal carcinoma patients in CSV format. After detecting that the matrix contains the 318 genes selected in Example 1, the web-based tool will assign a predicted subtype to each patient based on the nasopharyngeal carcinoma subtype classification model obtained in Example 1, using the same threshold parameter. Users can guide stratified treatment based on the subtype classification results.

[0047] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A biomarker for predicting the benefit of anti-PD-1 immunotherapy in nasopharyngeal carcinoma, characterized in that, The biomarkers are combinations of the following genes: BAD, CD99, CFLAR, ARF5, SLC25A5, RPS20, UQCRC1, POLR3B, CAPN1, CD74, SAMD4A, FHL1, BIRC3, STRAP, LCP2, HEBP2, PRSS8, KCNH2, RPL18, GPC1, CLEC2D, TGFBR3, ATP1B3, ENO1, SLC25A3, FGFR1, ITGB5, PLOD1, GSTP1, SF3B2, RPLP0, FXYD3, PHGDH, PTGS1, HSP90AB1, NDUFB7, GADD45B, XBP1. RPL3, PLTP, PSMA7, SAMHD1, MYOM1, PGRMC1, MMP15, LAPTM4B, EEF1D, FCGRT, OAZ1, TLE5, ILVBL, RPS16, RPS19, PPP2R1A, JAK3, RPL18A, EIF3B, RBM28, FK TN, DVL1, LZTS2, CCDC6, RPL28, RPL19, SYNGR2, LGALS3BP, YWHAE, FBXW7, MDK, CD81, ATP5F1B, PTGES3, KRT18, LTBR, ​​GAPDH, USP5, TPI1, FBXL4, BICRAL, C UL9, CDH6, DBN1, AMOTL2, PPM1G, RPS15, STAT1, IGFBP2, ST6GALNAC5, SLC2A1, EBNA1BP2, CDC20, PRDX1, STAG1, PPL, EPCAM, TMEM131L, PSPC1, OBSL1, H2 BC11, MYRF, MT2A, PAX8, SNRPB, PRDX5, FGFRL1, ATF4, CDKN1C, TOMM40, BST2, ACTN4, LSM4, ASS1, AOC3, TRIM5, TRIM22, H3-3B, RBM38, TBC1D8B, WDR74, FK BP11, CD63, GDF11, COX5B, ANKRD36, HUS1, RPS6, RPLP1, DIAPH3, NLRC5, RPS2, ARHGDIA, EIF4A3, ANAPC11, RPL13A, RPL11, CCN1, RPS8, CERS2, RHOB, RPS 27A, SNRNP200, CTDSP1, RPL32, TBC1D32, OGT, RPL10, AOPEP, STOM, GSN, RPL7A, RPS3, ATM, FAU, ALDOA, LYPD1, NDUFC2, SCLT1, ASAP2, SLC25A4, ZFP36L2,ADAMTS1、RHOC、CFAP70、EEF1A1、UBE2L6、EIF4A2、SLC34A2、WASF2、H2BC5、N RG2、TAGLN2、ADAMTS4、CDC42SE2、PSMB4、CTBP1、ADAR、LY6E、TONSL、PPP1R1 6A, MGAT4B, RPL8, SRSF2, IGSSF8, FBLN2, NMNAT3, TKT, WDR82, CTSB, ZNF503, DIPK1B, METTL3, PSMC3, GPT2, SPINT1, CCT2, B2M, PPIB, PHB1, GPX4, EEF2, PL IN4, OTUB1, FTH1, SLC3A2, RPSA, H1-4, SEPTIN2, TAP1, LRATD2, SLC25A6, RAB3B, NPR1, SCN9A, PLEKHA2, FASN, TMED10, KRT8, KIF5B, RPS9, RPS21, BSG, G NB2、CFL1、OR10K1、VEGFB、SULT1B1、EIF1、RNF213、RPL15、YIF1A、PSMD2、CO X8A,METRNL,SHMT1,ARIH2,JUN,PNPLA2,CD151,GRINA,TUFM,CYC1,CALR,PU F60、H2BC4、SCRIB、C4orf50、RFX7、UNC5C、SHMT2、EPHB3、RPS17、CRIP2、COL 18A1、ACTG1、H2AC20、KCNK12、H2AC21、TACSTD2、H1-5、HSF1、ARL15、HGS、P4 HB、SLC52A2、BCAP31、NPIPB4、NDUFA13、KRT5、BCAM、PTMA、TUBB4B、SBK1、H2 AX、CLDN4、TUBB、XRCC6、GM2A、KPNA5、PCBP2、CDC42SE1、PSAP、RPL37A、CFD、G PAA1、PLCG2、RPL12、C1orf122、SFI1、RPL10A、SELENOT、H2BC18、CHIC1、HLA -DRA、HSPA1B、HLA-C、HLA-E、HLA-A、IGKC、IGHA1、ATF6B、DDAH2、PHB2、CEBPD 、RPL41、RPS18、HLA-B、CCDC200、DDOST、EEF1G、TMEM178B、MRPL12、TSNAX-D ISC1、H4C15、H4C14、NBPF26、PLD5P1、H2AC18、XNDC1N-ZNF705EP-ALG1L9P。、 2. The use of the biomarker described in claim 1 in constructing a model for predicting the benefit of anti-PD-1 immunotherapy in nasopharyngeal carcinoma.

3. The use of the substance that detects the biomarker of claim 1 in the preparation of a product for predicting the benefit of anti-PD-1 immunotherapy for nasopharyngeal carcinoma.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the function of the model for predicting the benefit of anti-PD-1 immunotherapy for nasopharyngeal carcinoma as described in claim 2.

Images