Application of combined genome in preparation of reagent, chip or kit for renal clear cell carcinoma prognosis and / or immunotherapy curative effect evaluation

Abstract

The invention provides an application of a combined genome in preparation of a reagent, a chip or a kit for prognosis and / or immunotherapy curative effect evaluation of renal clear cell carcinoma. The combined genome contains a combination of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14 and TNFSF4. The five genes are used as prognosis marker genes of the renal clear cell carcinoma, the five genes are combined for prognosis and immunotherapy sensitivity prediction, the credibility is higher, the combined prediction performance of the five genes can show a remarkable prognosis difference in a clinical research queue, and patients with different immunotherapy curative effects can be distinguished.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical biological detection, and relates to the use of a combined genome composed of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 as a biomarker. Specifically, it is an application of this combined genome in the preparation of reagents, detection chips, or kits for the prognosis of clear cell renal cell carcinoma and the evaluation of the efficacy of immunotherapy. Background Art

[0002] Clear cell renal cell carcinoma (ccRCC) is the most common subtype of renal cell carcinoma. More than 400,000 people are diagnosed with ccRCC worldwide every year, and it causes approximately 175,000 deaths annually. Especially for advanced and metastatic ccRCC, the prognosis is even worse, and the 5-year survival rate after diagnosis is only 12%. Among them, the main reasons for the poor prognosis are the high intrinsic drug resistance of ccRCC tumors to conventional chemotherapy and radiotherapy and the drug resistance of currently applied targeted therapies. To combat treatment resistance, different targeted drugs are usually used to treat patients differently. However, in most cases, some patients with local lesions still relapse or progress to metastatic diseases after surgery. And in the treatment of metastatic ccRCC, the number of targeted pathways is limited, leaving almost no treatment options for non-responsive patients, resulting in a very poor prognosis. Therefore, the prognosis of ccRCC has significant heterogeneity, which forces the diagnosis and treatment of ccRCC to focus on personalized treatment options.

[0003] The main problems and solutions currently faced in the diagnosis and treatment of clear cell renal cell carcinoma include:

[0004] 1) The existing staging system cannot accurately guide the prognosis judgment of clear cell renal cell carcinoma

[0005] The improvement of the efficacy of ccRCC depends on accurate prognosis prediction. The current renal cancer staging is mainly based on the tumor staging system of the American Joint Committee on Cancer, namely the AJCC TNM staging system. However, considering that the occurrence and development of tumors is a multi-stage, multi-gene and multi-factor model, and in the era of the increasing prosperity of molecular pathology and postoperative adjuvant therapy, the AJCC TNM staging system constructed only from clinical pathological data gradually shows its limitation in accurately predicting the prognosis of patients in clinical applications. In recent years, the pattern of renal cancer prognosis prediction models is changing: based on the research of cancer cell biology and genomics, it has become possible to screen patients with specific prognoses and different treatment responses using drug sensitivity genes as the standard. Therefore, nowadays, the prognosis prediction models of many tumors intentionally combine clinical pathological factors with tumor-specific molecular markers in order to jointly evaluate various factors of the tumor itself in a single scoring system and guide clinical decisions. However, people have gradually realized that the malignant characteristics of tumors not only depend on the intrinsic characteristics of tumor cells, such as tumor driver genes, but also on the surrounding microenvironment in which tumor cells survive, that is, external characteristics, such as immune cells, inflammatory cytokines, and stromal cells recruited and activated in the tumor microenvironment. The current prognostic prediction systems for clear cell renal cell carcinoma have not paid attention to the heterogeneity of the immune microenvironment. Therefore, finding more precise molecular markers that can comprehensively reflect the heterogeneity of ccRCC is a very important issue. At the same time, if such molecular markers can reflect the changes in both the intrinsic malignancy of the tumor and the external immune microenvironment of the tumor, it will have the significance of determining prognostic classification and will also provide a new basis for the prognostic evaluation of ccRCC.

[0006] 2) Lack of an evaluation system for the efficacy of immunotherapy for clear cell renal cell carcinoma

[0007] In recent years, based on the fact that immune checkpoint inhibitors (ICIs) have made great breakthroughs in the field of tumor immunotherapy and their application value has been verified in clinical research, patients with ccRCC have obtained new options and hopes. In the phase III clinical study CHECKMATE-025 (NCT01668784), 821 patients with advanced metastatic renal cell carcinoma who had received previous treatment were enrolled to compare the treatment effects of immunotherapy and everolimus. The results showed that the median overall survival (OS) of the immunotherapy group vs. everolimus was 25 months vs. 19.6 months, the objective response rate (ORR) was 21.5% vs. 3.9%, and the duration of tumor response (DOR) was 23.0 months vs. 13.7 months. Thus, the situation of renal cancer immunotherapy has been opened.

[0008] Based on the benefits of PD-1 monoclonal antibody in the second-line treatment of advanced renal cell carcinoma, immunotherapy has been advanced, and its efficacy in the first-line treatment of advanced renal cell carcinoma has been affirmed. The objective response rate of patients with clear cell carcinoma is 36.4%. However, while immunotherapy has achieved success in the study of renal cell carcinoma, the inventors found that ccRCC is less sensitive to immunotherapy than other tumors, and the current immunotherapy sensitivity markers are not applicable in renal cell carcinoma. Moreover, the true efficacy is completely opposite to the current prediction and discrimination criteria. These research findings confirm that renal cell carcinoma has its unique immune status. Therefore, it is very necessary to find more sensitive new biomarkers for clear cell renal cell carcinoma and construct a prediction model for the efficacy of immunotherapy.

[0009] At the same time, for patients receiving immunotherapy, clinical observations have found that about 5% of patients experience "hyperprogression" of tumors (that is, after using immunotherapy drugs, the tumor progression is accelerated instead). At present, there have been some research results on the prediction of related biomarkers for tumor "hyperprogression", but the conclusions are not yet conclusive.

[0010] Therefore, the key to realizing the individualized treatment of ccRCC, ensuring that patients obtain the greatest benefit with the least toxicity, and improving the long-term survival rate of the overall ccRCC population is whether it is possible to predict the efficacy of ccRCC immunotherapy and screen out sensitive and tolerant groups. Summary of the Invention

[0011] The object of the present invention is to provide a prognostic and immunotherapy efficacy evaluation marker for clear cell renal cell carcinoma. Further, the object of the present invention is to provide a new use of a combined genome composed of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4, that is, its application in a prognostic evaluation kit for clear cell renal cell carcinoma.

[0012] The inventors found that the co-stimulatory factor family is the key for immune checkpoint inhibitors to exert their effects. The success of molecular ICIs stems from an in-depth understanding of the immune-active and immunosuppressive conditions in the tumor microenvironment (TME). In the TME, the role of T cells is to distinguish cancer cells from normal cells and initiate an attack on the "foreign" - cancer cells. Before the attack, naive T cells need to be activated by two signals. The recognition of a specific antigen by the T cell receptor (TCR) is the first signal, i.e., the specific signal, while the second signal comes from a non-specific co-stimulatory signal. Because naive T cells cannot further exert their killing effects without being activated by co-stimulatory signals, cancer cells prevent the transmission and recognition of these signals by altering the co-stimulatory molecular signals and expressions in the TME. Therefore, ICIs can prevent tumor cells from transmitting false information to T cells, thereby selectively repairing the immune defects induced by tumor cells in the TME. There are two categories of the co-stimulatory factor family, one is the B7-CD28 (PD-L1 / PD-1, CD86 / CTLA4) family, and the other is the tumor necrosis factor (TNF) family.

[0013] Based on the biological understanding of the co-stimulatory factor family and the mechanism of action of ICIs, the inventors speculated that the co-stimulatory factor family might be a breakthrough in the study of ICIs sensitivity markers for clear cell renal cell carcinoma. In previous studies, the inventors constructed an immune prognosis model (IPM) for clear cell renal cell carcinoma (ccRCC) by integrative analysis of the co-stimulatory factor family genes. Pre-experiments found that IPM had excellent prognostic prediction performance. At the same time, IPM classification could reflect the heterogeneity of the abundance of immune infiltrating cells in the tumor microenvironment, confirming that the prognosis of clear cell renal cell carcinoma was closely related to the immunosuppressive state, and IPM classification was related to the sensitivity of ccRCC to immune checkpoint inhibitors (ICIs). Thus, a type of ccRCC that was insensitive to immunotherapy and had extremely poor prognosis was classified - refractory clear cell renal cell carcinoma (IPM high risk classification). Further research found that if judged by the current traditional immune therapy sensitivity indicators (tumor mutation burden TMB, cytolytic activity, immune infiltration score), the IPM high risk classification should belong to "hot tumors" and should be sensitive to immune checkpoint inhibitor immunotherapy and have a good prognosis. This new discovery inspired the inventors to study the mechanism of the immunosuppressive state formation in ccRCC. Through the study of IPM classification, it was further found that the reason why refractory ccRCC could not benefit from immunotherapy was the dysfunction of T cells, that is, the true nature of the "hot tumor" appearance reflected by the increase in immune cell infiltration was the dysfunction of CD8+ T cells, the dysfunction of CD4+ T cells, and the increase in the infiltration of immunosuppressive T cells (iTreg, nTreg, CAF). At the same time, it was also found that the microsatellite instability of the IPM high risk classification was low. The immunotherapy sensitivity prediction of the IPM classification constructed by us was verified in the renal cancer immunotherapy dataset published in the recent Nature Medicine journal, and the results confirmed the previous findings. Therefore, the present invention will provide a new translational medicine basis for the patient selection of clear cell renal cell carcinoma immunotherapy, enhance the economic and social benefits of clear cell renal cell carcinoma treatment, and achieve the purpose of translational medicine. Through complex bioinformatics research, the inventors performed big data analysis on the RNA sequencing data, clinicopathological and long-term follow-up information of 533 cases of clear cell renal cell carcinoma (ccRCC), and thus discovered and screened out 5 independent prognostic genes for ccRCC, namely HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14 and TNFSF4.Among them, HHLA2 (Human endogenous retrovirus-H long terminal repeat-associating protein 2) belongs to the members of the B7 family and is a novel target site in tumor immunotherapy; while TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 belong to the tumor necrosis factor receptor superfamily and have attracted attention in tumor immunotherapy.

[0014] In the first step, 533 cases of ccRCC were randomly divided into a training group and a validation group at a ratio of 7:3.

[0015] In the second step, univariate COX analysis, Lasso regression analysis, and multivariate COX analysis were used in the training group to explore the correlation between the expression of co-stimulatory factor family genes and the long-term prognosis of ccRCC. Finally, 5 independent prognostic genes of ccRCC were screened out, namely HHLA2 (HR = 0.68, 95%CI 0.57 - 0.80), TNFRSF10A (HR = 0.84, 95%CI 0.72 - 0.98), TNFRSF1A (HR = 1.24, 95%CI 1.03 - 1.50), TNFSF14 (HR = 1.38, 95%CI 1.15 - 1.66), TNFSF4 (HR = 1.37, 95%CI 1.14 - 1.65). Furthermore, a COX proportional hazards model was constructed through these 5 independent prognostic genes to obtain an immune molecular prognostic scoring model. Using -0.14 as the threshold, ccRCC was divided into a high-risk group and a low-risk group according to the death risk, that is, the IPM classification.

[0016] In the third step, Kaplan-Meier survival curve analysis and ROC curve analysis were used in the validation group to verify the accuracy of IPM in prognostic grouping of ccRCC. It was found that IPM could significantly distinguish high-risk and low-risk patients in the validation group (HR = 3.95, 95% CI 2.77 - 5.64, P < 0.001). At the same time, compared with the AJCC stage (AUC = 0.67), the predictive performance of IPM was better (AUC = 0.74).

[0017] In the fourth step, because the IPM classification can reflect the differences in the prognosis of ccRCC, the project team further explored the potential mechanism. By screening the correlation between the IPM score and the expression levels of 22,300 human genes' RNAs, 336 genes related to the IPM score were screened out (|Rho| > 0.5, P < 0.05). Bioinformatics analysis of the 336 related genes found that biological processes such as immunity, inflammation, and T cell homeostasis were highly enriched, indicating that the IPM score was closely related to immune activity and T cell homeostasis.

[0018] Step 5. Based on the findings regarding the relationship between IPM scores and immune activity and T cell homeostasis, it was hypothesized that IPM classification might reflect differences in immunotherapy sensitivity, and that such differences might be caused by changes in T cell homeostasis. Therefore, in-depth exploration was conducted. By combining two bioinformatics algorithms, ESTIMATE and ImmuCellAI, which can predict the composition and abundance of tumor microenvironment cells from RNA sequencing data, the differences in immune infiltration abundance between the two IPM classifications were evaluated. It was found that immune cell infiltration was more abundant in the high-risk IPM group with poor prognosis (P<0.05), indicating that the microenvironment presented a "hot tumor" state with abundant immune cells. Based on the TIDE algorithm, which can predict the response to immune checkpoint inhibitor therapy from RNA sequencing data, the differences in responsiveness between the two IPM classifications were speculated. It was found that the high-risk IPM group with poor prognosis and a "hot tumor" state was less sensitive to immunotherapy (P<0.001). Meanwhile, using the sequencing data of a prospective cohort of renal cancer patients treated with nivolumab, it was verified that the high-risk IPM group was refractory ccRCC with poor prognosis and insensitivity to immunotherapy. However, judged by the current traditional immunotherapy sensitivity indicators (tumor mutation burden TMB, cytolytic activity), the high-risk IPM group had high TMB and strong cytotoxic activity, and such renal cancers should have been sensitive to immune checkpoint inhibitors and had a better prognosis.

[0019] Step 6. The contradiction between the research findings and the existing understanding inspired the inventors to study the formation mechanism of the immune state of ccRCC. By further refining the cell abundance and type through IPM classification, it was found that the reason why the high-risk IPM group of refractory ccRCC could not benefit from immunotherapy was the dysfunction of T cells. The true nature of the "hot tumor" illusion was actually the dysfunction of CD8+ T cells and CD4+ T cells, and the increased infiltration of immunosuppressive T cells (iTreg, nTreg, CAF), that is, the impairment of T cell homeostasis. At the same time, it was also found that the microsatellite instability score of the high-risk IPM group was lower (P<0.001).

[0020] The present invention relates to the first discovery of using the above five genes as prognostic marker genes for clear cell renal cell carcinoma. Combining them for prognostic and immunotherapy sensitivity prediction has higher credibility, and the combined prediction performance of the 5 genes can show significant prognostic differences in clinical research cohorts, and can distinguish patients with different immunotherapy efficacies.

[0021] In the first aspect of the present invention, there is provided the use of a combined genome in the preparation of a reagent, chip or kit for evaluating the prognosis and / or immunotherapy efficacy of clear cell renal cell carcinoma, characterized in that the combined genome contains a combination of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14 and TNFSF4.

[0022] In the second aspect of the present invention, the combined genome contains a combination of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4.

[0023] In the third aspect of the present invention, the reagent for prognosis and immunotherapy efficacy evaluation is a reagent combination for detecting the relative expression levels of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 in a biological sample, and the kit for prognosis and immunotherapy efficacy evaluation contains a reagent combination for detecting the relative expression levels of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 in a biological sample.

[0024] In the fourth aspect of the present invention, the reagent combination is a reagent combination for detecting the relative mRNA expression levels of the above genes, and the reagent combination contains PCR primers with detection specificity for the genes.

[0025] In the fifth aspect of the present invention, the PCR primers with detection specificity are shown as SEQ ID NO:1 to SEQ ID NO:10:

[0026]

[0027]

[0028] In the sixth aspect of the present invention, the prognostic risk value of clear cell renal cell carcinoma is calculated according to the following formula: Risk value = HHLA2 expression level × (-0.3866) + TNFRSF10A expression level × (-0.1716) + TNFRSF1A expression level × (0.2182) + TNFSF14 expression level × (0.3248).

[0029] In the seventh aspect of the present invention, the biological sample is a tumor specimen section surgically resected from a patient with clear cell renal cell carcinoma.

[0030] In another aspect of the present invention, a reagent, chip, or kit for prognosis and / or immunotherapy efficacy evaluation of clear cell renal cell carcinoma, the reagent, chip, or kit contains a reagent for detecting the relative expression levels of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 in a biological sample.

[0031] In another aspect of the present invention, the reagent, chip, or kit includes a reverse transcription system, a primer system, and an amplification system, and the primer system includes the PCR primers shown as SEQ ID NO:1 to SEQ ID NO:10:

[0032]

[0033] On the other hand of the present invention, the prognostic risk value of clear cell renal cell carcinoma is calculated according to the following formula: Risk value = HHLA2 expression level × (-0.3866) + TNFRSF10A expression level × (-0.1716) + TNFRSF1A expression level × (0.2182) + TNFSF14 expression level × (0.3248).

[0034] The innovation of the present invention lies in:

[0035] 1. The immune molecular prognostic model IPM of clear cell renal cell carcinoma was first constructed by the inventors, ensuring the originality at the source of this project.

[0036] 2. Existing staging systems such as AJCC staging only elaborate from the perspective of traditional clinical pathology, far from being able to explain the biological heterogeneity of clear cell renal cell carcinoma. Although the MSKCC scoring system and IMDC scoring system for renal cancer incorporate molecular test indicators, simple molecular typing can only reflect the biological characteristics of tumors at the gene level of tumor cells. Because the malignant characteristics of tumors depend not only on the internal characteristics of tumor cells, such as tumor driver genes, but also on the surrounding microenvironment in which tumor cells survive, that is, external characteristics, such as immune cells, inflammatory cytokines, and stromal cells recruited and activated in the tumor microenvironment. However, the IPM typing constructed in the present invention not only reflects the infiltration changes of immune cells in the immune microenvironment, the prognostic differences of clear cell renal cell carcinoma, but also may reflect the sensitivity of immune checkpoint inhibitors. Therefore, IPM will more precisely reflect the heterogeneity of the malignant characteristics of clear cell renal cell carcinoma, provide a new basis for the individualized diagnosis and treatment of clear cell renal cell carcinoma, and at the same time ensure the innovation of the results of the present invention in clinical transformation, which is also the feature of the present invention.

[0037] 3. In the research, the present inventors found the IPM-high subtype with poor prognosis and insensitivity to immune checkpoint inhibitors. Further research found that if judged by the current traditional immune therapy sensitivity indicators (tumor mutation burden TMB, cytolytic activity, immune infiltration score), the IPM-high subtype should belong to "hot tumors", be sensitive to immune checkpoint inhibitor immunotherapy, and have a better prognosis. This new discovery inspired the inventors to study the formation mechanism of the immune suppression state in ccRCC. Through the study of IPM typing, it was further found that the reason why refractory ccRCC cannot benefit from immunotherapy is the dysfunction of T cells, that is, the true nature of the "hot tumor" appearance reflected by the increased infiltration of immune cells is the dysfunction of CD8+ T cells, the dysfunction of CD4+ T cells, and the increased infiltration of immunosuppressive T cells (iTreg, nTreg, CAF). At the same time, it was also found that the microsatellite instability of IPM high risk is low. The immune therapy sensitivity prediction performance of IPM typing was verified in the renal cancer immunotherapy dataset published in the recent Nature Medicine journal, and the results were consistent with the previous bioinformatics findings. The present inventors not only discovered the role of the current immune therapy sensitivity markers and CD8+ T cells in predicting the efficacy of renal cancer immunotherapy, but also for the first time in the present invention explained the possible mechanism of the insensitivity of clear cell renal cell carcinoma seemingly "hot tumors" to immunotherapy, reflecting the innovation of the present invention.

[0038] The present invention has achieved beneficial guarantees and effects:

[0039] The gene combination of the present invention is derived from the molecular markers of immune co-stimulatory factors in clear cell renal cell carcinoma. The discovery of this gene combination model provides a brand-new strategy for predicting the long-term survival status after surgery and immune therapy sensitivity markers in clear cell renal cell carcinoma, plays an important role in judging the prognosis of patients with clear cell renal cell carcinoma, can evaluate the risk level of tumor progression or death after surgery in patients, helps guide clinicians to implement individualized precision treatment strategies, improve the survival rate of patients after surgery, and also has important guiding significance for the follow-up monitoring and sequential treatment management of patients with clear cell renal cell carcinoma after surgery.

[0040] Technically speaking, the five-gene detection is essentially a quantitative PCR detection of tissue samples, with the characteristics of simple operation, high detection sensitivity, good specificity, high repeatability, etc. Nowadays, it has been increasingly applied to clinical testing technologies. This technology has been proven to be a highly sensitive and accurate detection method in modern experimental diagnostics, and the test technology has been very mature. And we use the standard curve quantification method in this technology, which can accurately quantify specific nucleic acid molecules in various samples. Brief Description of the Drawings

[0041] Figure 1 Figures 4 are schematic diagrams of molecular screening and validation of the immune molecular prognostic prediction model (IPM) for clear cell renal cell carcinoma. Among them,

[0042] Figure 1 are immune co-stimulatory molecules potentially related to prognosis screened by univariate COX analysis;

[0043] Figure 2A 、 Figure 2B are 13 immune co-stimulatory molecules obtained by further narrowing the range through LASSO Cox regression analysis;

[0044] Figure 3 are 5 immune co-stimulatory molecules finally screened by multivariate COX;

[0045] Figures 4A - 4F are the prognostic performance graphs of the 5 molecules and the constructed IPM model.

[0046] Figures 5 - 13 are multi-group validations of the predictive performance of the immune molecular prognostic prediction model (IPM). Among them,

[0047] Figures 5 - 7 respectively show the survival differences of the model in distinguishing patients in the training set, test set, and overall set;

[0048] Figures 8 - 10 respectively show the performance of the model in distinguishing the survival differences of patients in the training set, test set, and overall set;

[0049] Figures 11 - 13 respectively show the distributions of the model scores in patients in the training set, test set, and overall set.

[0050] Figures 14 - 15 shows the intratumoral immunological activity of the immune molecular prognostic prediction model (IPM). Among them,

[0051] Figure 14 are genes screened related to the IPM model;

[0052] Figure 15 are the biological pathways enriched by genes related to the IPM model.

[0053] Figures 16 - 27 shows that the immune molecular prognostic prediction model (IPM) has the potential to predict immunotherapy sensitivity. Among them,

[0054] Figure 16 is the abundance of microenvironment cells;

[0055] Figure 17 、 Figure 18 are the abundances of immune infiltrating cells;

[0056] Figure 19 is the tumor purity;

[0057] Figure 20 is the infiltration abundance of different types of immune cells;

[0058] Figure 21 and Figure 22 there are differences in the predicted immunotherapy response in the IPM classification;

[0059] Figure 23 and Figure 24 there are differences in the response of patients to immunotherapy in the real cohort in the IPM classification;

[0060] Figure 25 there are differences in the tumor mutational burden in the IPM classification;

[0061] Figure 26 there are differences in the cytotoxic effect in the IPM classification;

[0062] Figure 27 The analysis reveals other microenvironment differences in the IPM classification. Detailed implementation manners

[0063] The following details the implementation of the present invention in conjunction with the drawings and embodiments of the present invention. The following embodiments are implemented on the premise of the technical solution of the present invention, and the detailed implementation manners and specific operation processes are given, but the protection scope of the present invention is not limited to the following embodiments.

[0064] The reagents and raw materials used in the present invention are all commercially available or can be prepared according to the methods in the literature. The experimental methods without specific conditions noted in the following embodiments are usually carried out according to the conventional conditions in the laboratory or according to the conditions recommended by the manufacturer.

[0065] In the training set of patients, immune co-stimulatory molecules related to prognosis are screened. Figure 1 Univariate COX analysis is used to preliminarily screen immune co-stimulatory molecules that may be related to prognosis. Figure 2A and Figure 2B LASSO Cox regression analysis is used to further find hub genes related to prognosis and narrow the research scope, indicating that a total of 13 immune co-stimulatory molecules are important genes. Figure 3 Multivariate COX analysis is used to further screen and find that 5 immune co-stimulatory molecules are independent predictors of prognosis. The risk value (IPM risk score) of the prognosis model composed of these 5 gene combinations = expression level of HHLA2 × (-0.3866) + expression level of TNFRSF10A × (-0.1716) + expression level of TNFRSF1A × (0.2182) + expression level of TNFSF14 × (0.3248). Figures 4A - 4F, the Kaplan–Meier method showed an association between the transcriptional expression of five independent prognostic genes (including HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4) in ccRCC patients and prognosis. The prognostic model risk value (IPM risk score) composed of the five-gene combination was clustered with a threshold of -0.14 into high-risk and low-risk patients, that is, IPM classification, and a significant association with prognosis was found, which could significantly distinguish the two groups of patients.

[0066] Figures 5 - 13 For predicting the prognostic prediction performance of the combined prognostic model: Figure 5 The differences in patient survival between the two training set groups were compared, and there was a significant association between the prognostic model risk value (IPM risk score) composed of the five-gene combination and prognosis; Figure 6 , Figure 7 The obtained results were verified using the test set and the overall set respectively, and there was a significant association between the prognostic model risk value (IPM risk score) composed of the five-gene combination and prognosis; Figures 8 - 10 , in the training set, test set, and overall ROC analyses, the area under the curve (AUC) values were 0.75, 0.74, and 0.75 respectively, indicating that the combined gene prognostic model of the present invention shows stable and good predictive ability for patient survival by stratifying low-risk and high-risk. In addition, the features we constructed showed the best AUC value among other conventional clinicopathological features in the TCGA cases, which also reflected its excellent predictive ability. Figures 11 - 13 Respectively showed the distribution of the prognostic model risk value composed of the five-gene combination, the co-stimulatory molecule expression, and the survival status of the training, test, and overall experiments.

[0067] Figures 14 - 15 IPM classification can reflect the intratumoral immunological activity heterogeneity of ccRCC. Figure 14 Through correlation analysis of 205,300 genes, genes related to the prognostic model risk value of the immune co-stimulatory factor score were screened (|rhoR|>0.5, P<0.05), and 330 genes meeting the conditions were screened. Figure 15 DAVID GO functional annotation analysis was performed on the above 330 genes related to the prognostic model risk value of the immune co-stimulatory factor score, and it was found that the enriched biological processes were in pathways such as inflammation and immune response, indicating that IPM classification can reflect the intratumoral immunological activity heterogeneity of ccRCC.

[0068] Figures 16 - 27 It is shown that IPM classification has the potential to predict immunotherapy sensitivity. Figure 16, The ESTIMATE score indicated that the IPM classification reflected the heterogeneity of the microenvironment cells among individual ccRCC tumors. The tumor microenvironment of the IPM-high classification had a richer cell infiltration than that of the IPM-low classification (P<0.001); Figure 17 、 Figure 18 , The Immune score indicated that the IPM classification reflected the heterogeneity of the abundance of immune infiltrating cells among individual ccRCC tumors. The immune cell infiltration in the tumor microenvironment of the IPM-high classification was richer than that of the IPM-low classification (P<0.001); Figure 19 , The tumor purity analysis indicated that the IPM classification reflected the heterogeneity of the tumor parenchymal cell purity among individual ccRCC tumors. The tumor purity of the IPM-high classification was lower than that of the IPM-low classification (P<0.001); Figure 20 Regarding the differences in the infiltration of immune cell types, it was found that CD8+ T cells, CD4+ T cells, macrophages, and dendritic cells increased in the IPM-high classification. At the same time, exhausted T cells, immunosuppressive cells (nTreg, iTreg, MAIT), etc. also increased; Figure 21 、 Figure 22 , The TIDE analysis combined with existing immunotherapy data to predict the immunotherapy response of each patient. The rank sum test found that the IPM classification had differences in immunotherapy sensitivity. The IPM-high classification with a poor prognosis was insensitive to immunotherapy (P<0.01), AUC = 0.63. Figure 23 、 Figure 24 , In the prospective cohort data of the immunotherapy of nivolumab in renal clear cell carcinoma, verification was carried out. The IPM risk score of each patient was calculated, combined with their true immunotherapy response. The rank sum test found that the IPM classification had differences in immunotherapy sensitivity. The IPM-high classification with a poor prognosis was insensitive to immunotherapy (P = 0.071), AUC = 0.56. Figure 25 , There were differences in TMB between the two IPM classifications, but the TMB of the IPM-high classification insensitive to immunotherapy was higher, subverting the previous impression that the higher the TMB, the more sensitive the immunotherapy; Figure 25 , There were differences in the cytotoxicity (CYT) between the two IPM classifications, but the cytotoxicity (CYT) of the IPM-high classification insensitive to immunotherapy was greater, subverting the previous impression that the greater the cytotoxicity (CYT), the more sensitive the immunotherapy; Figure 27 , The TIDE analysis found that there were differences in the T cell functional activities between the two IPM classifications. The T cell activity of the IPM-high classification insensitive to immunotherapy was inhibited, and there were more CAF cells and increased expression of CD274 (i.e., PD-L1).

[0069] Example 1:

[0070] RNA sequencing data of 533 patients with clear cell renal cell carcinoma (ccRCC) were collected, and combined with clinicopathological and long-term follow-up information for big data analysis of bioinformatics.

[0071] Step 1: Randomly divide 533 cases of ccRCC into a training group and a validation group at a ratio of 7:3.

[0072] Step 2: In the training group, univariate COX analysis, Lasso regression analysis, and multivariate COX analysis were used to explore the correlation between the expression of co-stimulatory factor family genes and the long-term prognosis of ccRCC. Finally, 5 independent prognostic genes for ccRCC were screened out, namely HHLA2 (HR = 0.68, 95% CI 0.57 - 0.80), TNFRSF10A (HR = 0.84, 95% CI 0.72 - 0.98), TNFRSF1A (HR = 1.24, 95% CI 1.03 - 1.50), TNFSF14 (HR = 1.38, 95% CI 1.15 - 1.66), TNFSF4 (HR = 1.37, 95% CI 1.14 - 1.65). Then, a COX proportional hazards model was constructed through these 5 independent prognostic genes to obtain an immune molecular prognostic scoring model. Using -0.14 as the threshold, ccRCC was divided into a high-risk group and a low-risk group according to the risk of death, that is, IPM classification ( Figure 1 ~Figure 4).

[0073] Step 3: In the validation group, Kaplan-Meier survival curve analysis and ROC curve analysis were used to verify the accuracy of IPM in prognostic grouping of ccRCC. It was found that IPM could significantly distinguish high-risk and low-risk patients in the validation group (HR = 3.95, 95% CI 2.77 - 5.64, P < 0.001). At the same time, compared with AJCC stage (AUC = 0.67), the predictive performance of IPM was better (AUC = 0.74) ( Figures 5 - 13 ).

[0074] Step 4: Since the IPM classification can reflect the differences in the prognosis of ccRCC, the project team further explored the potential mechanism. By screening the correlation between the IPM score and the expression levels of 22,300 human genes, 336 genes related to the IPM score were screened out (|Rho| > 0.5, P < 0.05). Bioinformatics enrichment analysis of the 336 related genes found that biological processes such as immunity, inflammation, and T cell homeostasis were highly enriched, indicating that the IPM score was closely related to immune activity and T cell homeostasis. Figures 14 - 15 )

[0075] Step 5. Based on the discovery of the relationship between IPM score and immune activity and T cell homeostasis, it was speculated that IPM classification might reflect differences in immunotherapy sensitivity, and this difference might be caused by changes in T cell homeostasis. Therefore, in-depth exploration was carried out. By combining two bioinformatics algorithms, ESTIMATE and ImmuCellAI, which can predict the composition and abundance of tumor microenvironment cells from RNA sequencing data, the differences in immune infiltration abundance between the two IPM classifications were evaluated. It was found that immune cell infiltration was more abundant in the high-risk group of IPM with poor prognosis (P<0.05), that is, the microenvironment showed a "hot tumor" state with abundant immune cells ( Figures 16 - 20 ). Based on the TIDE algorithm, which can predict the responsiveness to immune checkpoint inhibitors from RNA sequencing data, the differences in responsiveness between the two IPM classifications were speculated. It was found that the high-risk group of IPM with poor prognosis and "hot tumor" state was less sensitive to immunotherapy (P<0.001). At the same time, using the sequencing data of a prospective cohort of renal cancer patients treated with nivolumab, it was verified that the high-risk group of IPM was refractory ccRCC with poor prognosis and insensitive to immunotherapy ( Figures 21 - 24 ). However, judged by the current traditional immunotherapy sensitivity indicators (tumor mutation burden TMB, cytolytic activity), the high-risk group of IPM had high TMB and strong cytotoxic activity, and this type of renal cancer should be sensitive to immune checkpoint inhibitors and have a good prognosis ( Figure 25 、 Figure 26 ).

[0076] Step 6. The contradiction between the research findings and the existing understanding inspired the inventors to study the formation mechanism of the immune state of ccRCC. By further refining the cell abundance and type through IPM classification, it was found that the reason why the high-risk group of IPM in refractory ccRCC could not benefit from immunotherapy was the dysfunction of T cells. The true face of the "hot tumor" illusion was actually the dysfunction of CD8+ T cells and CD4+ T cells, and the increased infiltration of immunosuppressive T cells (iTreg, nTreg, CAF), that is, the impairment of T cell homeostasis. At the same time, it was also found that the microsatellite instability score of the high-risk group of IPM was lower (P<0.001) ( Figure 27 ).

[0077] Example 2:

[0078] Process and result analysis of quantitative detection of 5 genes in patient samples

[0079] Step 1. Real-time quantitative PCR (RT-qPCR) analysis

[0080] Total RNA was isolated from harvested tumor tissue cells using Trizol (Invitrogen, Carlsbad, CA), and it was reverse-transcribed into cDNA using the PrimeScript RT kit (Termo Fisher, USA). The primers were diluted and mixed in RNase-free dH2O using the SYBR Green qPCR method (Takara Biotechnology Co.). The primer sequences used are shown in Table 1 below. GAPDH RNA expression was measured for normalization. The specific cycling conditions for the operation of mRNA and GAPDH were determined according to the protocol of the Green qPCR premix (Applied Biosystems), and the relative expression level of the target mRNA was calculated by 2-ΔΔCt.

[0081] Table 1 Summary of primer sequences for five genes in qRT-PCR

[0082]

[0083] Step 2: Normalization analysis

[0084] The relative expression levels of the mRNAs of the above 5 genes were normalized by z-score.

[0085] Step 3: Calculate the IPM risk value (IPM risk score, prognostic model risk value)

[0086] The normalized expression levels of the above 5 genes were substituted into the IPM risk value formula: IPM risk value = HHLA2 expression level × (-0.3866) + TNFRSF10A expression level × (-0.1716) + TNFRSF1A expression level × (0.2182) + TNFSF14 expression level × (0.3248). The IPM risk value of the sample was calculated. If the IPM risk value < -0.14, the patient was a low-risk patient for prognosis, while if the IPM risk value > -0.14, the patient was a high-risk patient for prognosis and could not benefit from ICI immunotherapy.

[0087] Step 4: If the IPM risk value < -0.14, the patient was identified as having a low-risk prognosis and a good surgical prognosis. If the IPM risk value > -0.14, the patient was identified as having a high-risk prognosis. High-risk patients had a poor surgical prognosis and were insensitive to ICI immunotherapy at the same time. And it could be judged that the tumor samples of patients with an IPM risk value > -0.14 showed a false "hot tumor" state in the tumor microenvironment, and the tumor mutation burden (TMB) and cytolytic activity could not accurately infer the prognosis of the IPM high-risk group.

[0088] Example 3:

[0089] Before nivolumab immunotherapy, biopsy specimens were collected from 250 patients with clear cell renal cell carcinoma who were intended to receive nivolumab. The expression levels of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 were detected in the biopsy specimens. After Z-SCORE normalization of the expression levels of the above five genes, according to the IPM risk value = HHLA2 expression level × (-0.3866) + TNFRSF10A expression level × (-0.1716) + TNFRSF1A expression level × (0.2182) + TNFSF14 expression level × (0.3248), the IPM risk value of each patient was calculated. Patients with an IPM risk value < -0.14 were predicted to be the ICI immunotherapy benefit group, and patients with an IPM risk value > -0.14 were predicted to be the ICI immunotherapy insensitive group. Combining with the actual immunotherapy efficacy of the patients, a confusion matrix was constructed, and the statistics indicated that the IPM classification had differences in immunotherapy sensitivity. The IPM high-risk classification was insensitive to immunotherapy (P = 0.041), and the AUC = 0.68.

[0090] This example confirmed that the prognostic kit composed of these 5 genes could effectively evaluate the prognostic efficacy of immunotherapy for clear cell renal cell carcinoma.

[0091] Table 2 Aliases of the 5 genes HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 of the present invention.

[0092]

[0093]

[0094]

[0095] The above is the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. Use of a combined genome in the preparation of a reagent, chip or kit for evaluating the prognosis and / or efficacy of immunotherapy for clear cell renal cell carcinoma, characterized in that The combined genome contains a combination of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4.

2. The use of the combined genome according to claim 1 in the preparation of a reagent, chip or kit for evaluating the prognosis and efficacy of immunotherapy for clear cell renal cell carcinoma, characterized in that The combined genome contains a combination of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4.

3. The use according to claim 2, characterized in that The prognostic and immunotherapy efficacy evaluation reagent is a reagent combination for detecting the relative expression levels of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 in a biological sample, and the prognostic and immunotherapy efficacy evaluation kit contains a reagent combination for detecting the relative expression levels of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 in a biological sample.

4. The use according to claim 3, characterized in that The reagent combination is a reagent combination for detecting the relative mRNA expression levels of the above genes, and the reagent combination contains PCR primers specific for detecting the genes.

5. The use according to claim 4, characterized in that The PCR primers with detection specificity are shown as SEQ ID NO:1 to SEQ ID NO:10: HHLA2-F SEQ ID NO:1 AGTGCATGTAGAACCGAGCC HHLA2-R SEQ ID NO:2 CTCACCAGGAGGGACACAAC TNFRSF10A-F SEQ ID NO:3 ACGAGATTCTGAGCAACGCA TNFRSF10A-R SEQ ID NO:4 CAGCACCATTTGCTGGAACC TNFRSF1A-F SEQ ID NO:5 CTGGAGCTGTTGGTGGGAAT TNFRSF1A-R SEQ ID NO:6 TTGTGGCACTTGGTACAGCA TNFSF14-F SEQ ID NO:7 GGGTCTCTTGCTGTTGCTGA TNFSF14-R SEQ ID NO:8 CTGGGTTGACCTCGTGAGAC TNFSF4-F SEQ ID NO:9 CTCCTTGATGGTGGCCTCTC TNFSF4-R SEQ ID NO:10 ATCAGTTCTCCGCCATTCACA GAPDH-F SEQ ID NO:11 GTCTTCTCCACCATGGAGAAGG GAPDH-R SEQ ID NO:12 CATGCCAGTGAGCTTCCCGTTCA.

6. The use according to claim 4, characterized in that: Among them, The prognostic risk value of clear cell renal cell carcinoma is calculated according to the following formula: Risk value = HHLA2 expression level × (-0.3866) + TNFRSF10A expression level × (-0.1716) + TNFRSF1A expression level × (0.2182) + TNFSF14 expression level × (0.3248).

7. The use according to any one of claims 1 to 6, characterized in that The biological sample is a section of a tumor specimen surgically resected from a patient with clear cell renal cell carcinoma.

8. A reagent, chip or kit for evaluating the prognosis and / or efficacy of immunotherapy for clear cell renal cell carcinoma, the reagent, chip or kit comprising a reagent for detecting the relative expression levels of HHLA2, TNFRSF10A, TNFRSF1A, TNFSF14, and TNFSF4 in a biological sample.

9. The reagent, chip or kit according to claim 8, characterized in that The reagent, chip or kit includes a reverse transcription system, a primer system and an amplification system, and the primer system includes the PCR primers shown in SEQ ID NO:1 to SEQ ID NO:10: HHLA2-F SEQ ID NO:1AGTGCATGTAGAACCGAGCC HHLA2-R SEQ ID NO:2CTCACCAGGAGGGACACAAC TNFRSF10A-F SEQ ID NO:3ACGAGATTCTGAGCAACGCA TNFRSF10A-R SEQ ID NO:4CAGCACCATTTGCTGGAACC TNFRSF1A-F SEQ ID NO:5CTGGAGCTGTTGGTGGGAAT TNFRSF1A-R SEQ ID NO:6TTGTGGCACTTGGTACAGCA TNFSF14-F SEQ ID NO:7GGGTCTCTTGCTGTTGCTGA TNFSF14-R SEQ ID NO:8CTGGGTTGACCTCGTGAGAC TNFSF4-F SEQ ID NO:9CTCCTTGATGGTGGCCTCTC TNFSF4-R SEQ ID NO:10ATCAGTTCTCCGCCATTCACA GAPDH-F SEQ ID NO:11GTCTTCTCCACCATGGAGAAGG GAPDH-R SEQ ID NO:12CATGCCAGTGAGCTTCCCGTTCA。 10. The reagent, chip or kit according to claim 8, wherein: The prognostic risk value of clear cell renal carcinoma is calculated according to the following formula: Risk value = HHLA2 expression level × (-0.3866) + TNFRSF10A expression level × (-0.1716) + TNFRSF1A expression level × (0.2182) + TNFSF14 expression level × (0.3248).

Images