MiRNA marker combination, kit, system and medium for evaluating NSCLC prognostic risk

Abstract

The invention discloses a miRNA marker combination, a kit, a system and a medium for evaluating the prognostic risk of NSCLC, and belongs to the technical field of cancer markers. The miRNA marker combination comprises at least one of the following components: hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280, and the miRNA marker combination comprises at least one of the following components: the hsa-miR-223-3p, the hsa-miR-324-5p, the hsa-miR-596, the hsa-miR-7-1-3p, the The miRNA marker combination, the kit, the method, the system or the medium provided by the invention have better effect on risk assessment of non-small cell lung cancer patients than NCCN guide and CSCO guide, and have very high clinical application value.

Description

Technical Field

[0001] The present invention relates to the technical field of cancer markers, and in particular to a miRNA marker combination, a kit, a system and a medium for evaluating the prognostic risk of non-small cell lung cancer (NSCLC). Background Art

[0002] Lung cancer is the leading cause of cancer death worldwide, accounting for about 19% of all malignant tumors. The main pathological types of lung cancer include non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). NSCLC accounts for 80-85% of the total number of lung cancers, ranking first in tumor morbidity and mortality worldwide. NSCLC mainly includes squamous cell cancer (SCC) and adenocarcinoma (AC). Currently, the treatment of NSCLC is mainly based on the clinical staging of lung cancer. For stages I, II, and IIIa, surgical resection is the main treatment. For patients with significant lymph node metastasis, chemotherapy or radiotherapy can be used before surgery.

[0003] Low-dose computed tomography (LDCT) is a screening tool currently used for the diagnosis of early non-small cell lung cancer (NSCLC). With the implementation of LDCT screening strategies, 1-3% of the population is diagnosed with lung cancer each year, of which stage I cancer accounts for 50-70% of these cases (M. Oudkerk, S. Liu, MA Heuvelmans, JE Walter, JK Field. Lung cancer LDCT screening and mortality reduction-evidence, pitfalls and future perspectives [J], Nat Rev Clin Oncol. 2021, 18 (3): 135-151). However, the 5-year survival rate of early-stage NSCLC is still not optimistic, exceeding 75% in stage Ia, 68% in stage Ib, and 49-57% in stage II (SX Yan, MM Qureshi, K. Suzuki, M. Dyer, MT Truong, V. Litle, KS Mak. Definitive treatment patterns and survival in stage II non-small cell lung cancer [J], Lung Cancer. 2018, 124: 135-142). Therefore, with the increase in the annual incidence of early lung cancer screening by LDCT, it is imperative to identify high-risk groups with poor prognosis as early as possible to provide timely intervention and improve survival rate. NCCN Clinical Practice Guidelines for Oncology (NCCN ) (NCCN Guidelines) believe that poorly differentiated tumors, vascular invasion, wedge resection, tumors > 4 cm, visceral pleural involvement, and unknown lymph node status (NX) are high-risk factors for poor prognosis. The Chinese Society of Clinical Oncology (CSCO) Non-Small Cell Lung Cancer Diagnosis and Treatment Guidelines (CSCO Guidelines) believe that staging is an important factor in predicting prognosis. However, the existing survival prediction standards only consider clinical factors. Relying solely on clinical factors to predict the prognosis of patients cannot enable patients to benefit from chemoradiotherapy; therefore, the survival prediction of patients' prognosis based on these standards is still not ideal, and even the survival rate of patients with early non-small cell lung cancer has not been significantly improved.

[0004] With the emergence of the concept of precision medicine and the use of liquid biopsy to assess tumor prognosis and treatment, exploring new prognostic biomarkers including microRNAs and developing new model algorithms can be used to try to identify people who benefit from postoperative chemoradiotherapy for early-stage non-small cell lung cancer. This is also an effective way to improve the 5-year survival rate of patients with early-stage lung cancer.

[0005] Cancer prognosis is a prediction of the possible outcome of an individual's current medical condition and is an important tool for improving patient diagnosis and treatment management. Accurate prognosis is crucial for choosing the right cancer treatment method and predicting survival rate. Currently, there is a lack of molecular markers with high accuracy and clinical application efficacy for cancer prognosis. Most types of cancer require frequent imaging follow-up to monitor the course of the disease, which not only imposes a certain burden on patients, but also may lag in the detection of tumor progression. Therefore, there is a huge clinical need for molecular markers that are independent of pathological staging and other clinical factors and can accurately predict patient prognosis.

[0006] MicroRNA (miRNA) is a type of non-coding small RNA with a length of 19-25 nucleotides. It mainly participates in the regulation of individual development, cell apoptosis, proliferation and differentiation and other life activities by completely or incompletely pairing with the 3'-UTR of the target gene, degrading the target gene mRNA or inhibiting its translation. In the occurrence and development of tumors, the function of miRNA is similar to that of oncogenes or tumor suppressor genes. The expression profile of miRNA has obvious tissue specificity and has specific expression patterns in different tumors. In short, due to the characteristics of stable high expression, identifiable biological signals and specific expression patterns, miRNA can be used as a molecular marker to predict the prognosis of lung cancer (SUUmu, H. Langseth, C. Bucher-Johannessen, B. Fromm, A. Keller, E. Meese, M. Lauritzen, M. Leithaug, R. Lyle, T B Rounge. A comprehensive profile of circulating RNAs in human serum [J]. RNA Biol. 2018, 15 (2): 242-250).

[0007] After searching, the existing patents and literatures related to miRNA related to the prognosis of lung cancer (especially non-small cell lung cancer) are as follows:

[0008] Patent Publication No. CN101638656A discloses a serum / plasma miRNA marker associated with the prognosis of non-small cell lung cancer and its application as a diagnostic reagent. The marker is a combination of miR-486, miR-30d, miR-1 and miR-499.

[0009] Patent publication number CN103602678A discloses a marker miRNA combination for non-small cell lung cancer prognosis detection reagent, which consists of miR-638 and miR-27a.

[0010] Patent publication number CN113151462A discloses a biomarker for diagnosing the prognosis of lung cancer and its application as a diagnostic reagent, which is composed of hsa-miR-21-5p, hsa-miR-141-5p and hsa-miR-490-3p.

[0011] The literature (J.Lv, J.An, YDZhang, ZXLi, GLZhao, J.Gao, WWHu, HMChen, AMLi, QSJiang. A three serum miRNA panel as diagnostic biomarkers ofradiotherapy-related metastasis in non-small cell lung cancer[J].OncolLett.2020,20(5):236) reported a study evaluating the prognostic role of miRNAs in the blood of patients with non-small cell lung cancer after radiotherapy (including all clinical stages). It was found that increased expression of miR-130a, miR-25 and miR-191 in the serum of patients with non-small cell lung cancer was associated with poor survival rate.

[0012] The literature (Z.Hu, X.Chen, Y.Zhao, T.Tian, ​​G.Jin, Y.Shu, Y.Chen, L.Xu, K.Zen, C.Zhang, H.Shen.Serum microRNA signatures identified in a genome-wide serum microRNA expression profiling predict survival of non-small-cell lung cancer [J]. J Clin Oncol. 2010, 28 (10): 1721-6) used genome-wide serum miRNA expression analysis to study the role of serum miRNA in predicting the prognosis of non-small cell lung cancer. The study observed the prognosis of patients with stage I to IIIa NSCLC and found that 4 of the 11 serum miRNAs (i.e., miR-486, miR-30d, miR-1, and miR-1) were associated with overall survival (OS).

[0013] The literature (Akanksha Khandelwal, Uttam Sharma, Tushar Singh Barwal, Rajeev Kumar Seam, et al. Circulating miR-320a Acts as a Tumor Suppressor and Prognostic Factor in Non-small Cell Lung Cancer[J]. Front Oncol. 2021, 23(11): 645475) reported that miR-320a was significantly downregulated in NSCLC patients, so miR-320a may serve as a prognostic marker for NSCLC patients.

[0014] However, the common shortcomings of the above prior arts are:

[0015] 1) The results of many reported molecular markers for the prognosis of early lung cancer are inconsistent and cannot be verified with each other. The presumed reason is the differences in the subtypes, pathological types and inclusion and exclusion criteria of the study samples: most studies include postoperative patients in stage III in the verification samples, which may cause a higher mortality rate and thus affect the statistical results; in addition, some studies did not restrict the staging of NSCLC, resulting in the inability to accurately and objectively assess the risk of recurrence in the prognosis of early NSCLC patients.

[0016] 2) Most literature studies have only examined the correlation between the expression of specific miRNA markers and the prognosis and survival rate of early-stage NSCLC patients, and have not stratified these patients according to high or low risk of poor prognosis. In other words, it is impossible to confirm that the screened miRNA marker combination is better than the current CSCO standard and NCCN standard for the prognosis stratification of early-stage NSCLC patients.

[0017] In summary, how to screen out a serum / plasma miRNA marker combination that can be used to accurately assess the prognostic risk of patients with non-small cell lung cancer, especially those with early non-small cell lung cancer, and apply it to the development of corresponding kits, systems and media is a technical problem that has not yet been solved by technicians in this field. Summary of the invention

[0018] In order to solve the above technical problems, the inventors screened a miRNA marker combination that can be used for prognosis risk assessment of non-small cell lung cancer based on serum samples of 280 patients with early non-small cell lung cancer by quantifying the expression level of miRNA and combining machine learning methods, thereby completing the present invention. The technical solution adopted by the present invention is as follows:

[0019] In a first aspect, the present invention provides a miRNA marker combination for assessing the prognostic risk of non-small cell lung cancer, wherein the miRNA marker combination includes at least one of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280, for example, one, two, three, four, five, six, seven or eight.

[0020] In some embodiments of the present invention, the marker combination includes hsa-miR-7-1-3p and at least one of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280, for example, one, two, three, four, five, six or seven.

[0021] In other embodiments of the present invention, the marker combination includes hsa-miR-1280 and at least one of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b and hsa-miR-1275, for example, one, two, three, four, five, six or seven.

[0022] In other embodiments of the present invention, the marker combination includes hsa-miR-223-3p and at least one of hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280, for example, one, two, three, four, five, six or seven.

[0023] In some further embodiments of the present invention, the miRNA marker combination includes hsa-miR-223-3p and hsa-miR-324-5p. In some specific embodiments of the present invention, the miRNA marker combination consists of hsa-miR-223-3p and hsa-miR-324-5p.

[0024] In some further embodiments of the present invention, the miRNA marker combination includes hsa-miR-223-3p, hsa-miR-324-5p and hsa-miR-596. In some specific embodiments of the present invention, the miRNA marker combination consists of hsa-miR-223-3p, hsa-miR-324-5p and hsa-miR-596.

[0025] In some further embodiments of the present invention, the miRNA marker combination includes hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596 and hsa-miR-7-1-3p. In some specific embodiments of the present invention, the miRNA marker combination consists of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596 and hsa-miR-7-1-3p.

[0026] In some further embodiments of the present invention, the miRNA marker combination includes hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p and hsa-miR-320a. In some specific embodiments of the present invention, the miRNA marker combination consists of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p and hsa-miR-320a.

[0027] In some further embodiments of the present invention, the miRNA marker combination includes hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a and hsa-miR-320b. In some specific embodiments of the present invention, the miRNA marker combination consists of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a and hsa-miR-320b.

[0028] In some further embodiments of the present invention, the miRNA marker combination includes hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b and hsa-miR-1275. In some specific embodiments of the present invention, the miRNA marker combination consists of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b and hsa-miR-1275.

[0029] In some further embodiments of the present invention, the miRNA marker combination includes hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280. In some specific embodiments of the present invention, the miRNA marker combination consists of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280.

[0030] The second aspect of the present invention provides use of any one of the miRNA marker combinations described in the first aspect of the present invention in preparing a kit for assessing the prognostic risk of non-small cell lung cancer.

[0031] In some embodiments of the present invention, in the kit, the miRNA marker combination serves as or is prepared as a standard or quality control product.

[0032] Furthermore, when used as or prepared as a standard, each miRNA marker in the miRNA marker combination can be prepared into a standard working solution with a gradient concentration, and detected separately to obtain a detection signal value, and a standard curve is prepared based on the concentration and the detection signal value, and then quantified according to the detection signal value of the corresponding miRNA in the sample to be tested.

[0033] In some embodiments of the present invention, reverse transcription fluorescence quantitative PCR (RT-qPCR) is used to detect the expression level of miRNA markers. Specifically, miRNA is first reverse transcribed into cDNA, and then qPCR is performed on each standard working solution using specific primers / probes to obtain different Ct values. A standard curve is prepared using the concentration and Ct value of each standard working solution, and quantification is performed based on the Ct value of the corresponding miRNA in the sample to be tested.

[0034] Furthermore, when used or prepared as a quality control product, the miRNA marker combination can be used to test the effectiveness of its detection reagent. For example, before testing the sample to be tested, the quality control product is first tested using the detection reagent of the miRNA marker combination. When the detection signal meets the requirements, it indicates that the detection reagent is effective and can be used to test the sample to be tested.

[0035] The third aspect of the present invention provides use of a detection reagent for any miRNA marker combination as described in the first aspect of the present invention in the preparation of a kit for assessing the prognostic risk of non-small cell lung cancer.

[0036] In the present invention, the detection reagent is used to detect the expression level of the miRNA marker combination, and the expression level of the miRNA marker combination refers to the combination of the expression levels of each miRNA marker therein, that is, a set of expression levels (series or arrays).

[0037] In some embodiments of the present invention, the detection reagent is a reagent based on the RT-qPCR method, including primers and / or probes that specifically bind to each miRNA marker in the miRNA marker combination.

[0038] Furthermore, a reverse transcription reagent is also included, and the reverse transcription reagent includes a specific primer for reverse transcription of each miRNA marker, or includes a universal primer for reverse transcription of all miRNAs in the sample, including but not limited to stem-loop primers, gene-specific primers with poly(T) tails or tail sequences, etc. The miRNA-specific reverse transcription primer can effectively reduce background noise, while the universal primer is suitable for studying multiple different miRNAs at the same time.

[0039] Those skilled in the art may also use reagents based on other methods for detecting miRNA expression levels.

[0040] For example, reagents based on the Northern blotting method, i.e., probes. When using Northern blotting detection, RNA samples are digested by restriction endonucleases, separated by agarose gel electrophoresis, transferred to nitrocellulose membranes or nylon membranes according to their positions in the gel after denaturation, and reacted with isotope- or other labeled specific probes after fixation. After washing the probes, miRNAs can be detected by autoradiography or other suitable techniques.

[0041] Another example is the detection reagent based on microarray, which is also a probe. Microarray technology is a rapid and high-throughput method for detecting miRNA. Microarray technology uses labeled probes to produce sample RNA by reverse transcription, and uses solid phase oligonucleotides with the same sequence as the target miRNA to detect these fluorescent groups or biotin-labeled cDNA. The labeled cDNA sample is added to each well, followed by a series of washes to remove free DNA. If the hybridized cDNA is biotinylated, the fluorophore can be labeled; if the cDNA has been labeled with a fluorophore, the fluorescence intensity of each well can be directly measured. The fluorescence intensity of each well can be used to determine the expression level of miRNA.

[0042] Another example is the detection reagents based on small RNA sequencing (small RNA-seq), including but not limited to RNA extraction reagents, library construction reagents such as 3' adapters, 5' adapters, reverse transcription primers, amplification primers, sequencing adapters, and index sequences. Small RNA sequencing based on the next-generation high-throughput sequencing technology can obtain millions of small RNA sequences at one time, which can quickly identify all known small RNAs of a certain tissue under a specific state and discover new small RNAs, providing a powerful tool for small RNA function research.

[0043] The fourth aspect of the present invention provides a kit for assessing the prognostic risk of non-small cell lung cancer, comprising any miRNA marker combination and / or its detection reagent as described in the first aspect of the present invention.

[0044] In some embodiments of the present invention, RNA extraction reagents, reverse transcription reagents, PCR amplification reagents and / or probes are also included.

[0045] In some embodiments of the present invention, the RNA extraction reagent is used to extract RNA samples including miRNA from blood samples, wherein the blood samples include serum or plasma samples. Further, an rRNA removal reagent is also included.

[0046] In some embodiments of the present invention, a detection reagent for an internal reference miRNA is also included.

[0047] In order to achieve accurate and reproducible miRNA quantification results, it is necessary to use a suitable endogenous reference miRNA (internal reference miRNA) to normalize the amount of the measured miRNA. Normalization can avoid inaccurate quantitative results and enable direct comparison between different experiments and different sample test results. The ideal internal reference miRNA for miRNA detection data normalization should meet the following requirements:

[0048] (1) All samples involved have stable expression levels;

[0049] (2) similar in size to the miRNA being tested;

[0050] (3) Expression level is similar to the tested miRNA

[0051] (4) Expression levels will not be regulated under the test conditions

[0052] (5) The primers for endogenous reference RNA and miRNA should have similar amplification efficiencies (close to 100%).

[0053] In some specific embodiments of the present invention, the internal reference miRNA is selected from at least one of miR-361-5p and miR-425-5p. When there is more than one internal reference miRNA, the average value of the expression levels of multiple internal reference miRNAs is used for normalization.

[0054] A fifth aspect of the present invention provides a system for assessing the prognostic risk of non-small cell lung cancer, comprising the following modules:

[0055] A data input module, for receiving the expression level of any one of the miRNA marker combinations described in the first aspect of the present invention in a blood sample of a subject, wherein the subject suffers from early non-small cell lung cancer;

[0056] A database storage module, used to store population data, wherein the population data includes the expression level of the miRNA marker combination of a plurality of non-small cell lung cancer patients with different prognostic risks;

[0057] Evaluation module: connected to the data input module and the database storage module respectively, for constructing a machine learning classifier using population data, and evaluating the prognostic risk of the subject using the machine learning classifier and the expression level of the miRNA marker combination of the subject.

[0058] In the present invention, the group data is also referred to as large sample data, and the plurality is at least 100, at least 200, at least 300, at least 500, at least 1000, at least 5000 or more. Since the group data is used to construct a machine learning classifier, the larger the number in the group data, the better, and the group data can be continuously updated as the clinical data continues to increase.

[0059] In the present invention, the prognostic risk being in different stratifications means having different prognostic risks. In some embodiments of the present invention, the survival time can be used to judge the level of the prognostic risk. In some specific embodiments of the present invention, it can be divided into multiple types (i.e., multiple layers) according to the prognosis. For example, a prognostic survival time of less than 5 years is a high prognostic risk, and a prognostic survival time of not less than 5 years is a low prognostic risk, thereby dividing the prognostic risk into two layers.

[0060] In some embodiments of the present invention, the machine learning classifier is trained using a supervised machine learning algorithm or an unsupervised machine learning algorithm. In some specific embodiments of the present invention, the machine learning algorithm is selected from linear regression, logistic regression (LOG), polynomial regression, stepwise regression, ridge regression, Lasso regression, elastic network (ElasticNet, EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbor (kNN), generalized linear model (GLM), naive Bayes (NB), neural network, random forest (RF), deep learning algorithm, linear discriminant analysis (LDA), decision tree learning (DTREE), adaptive boosting (ADB) and any combination thereof. In some preferred embodiments of the present invention, the machine learning algorithm is SVM. In some specific embodiments of the present invention, the machine learning algorithm is a Cox proportional-hazards model, which is a semi-parametric regression model that can simultaneously study the relationship between multiple risk factors and the occurrence and occurrence time of event outcomes, thereby overcoming the shortcomings of single factor restrictions in simple survival analysis. In some other specific embodiments of the present invention, the machine learning algorithm is exponential regression. In some other specific embodiments of the present invention, the machine learning algorithm is Weibull regression, and Weibull regression Weibull distribution is the most popular and widely used distribution in parametric survival analysis models.

[0061] Furthermore, the machine learning classifier is compared and verified. The verification includes internal verification (such as random division training, test set and verification set, K-fold cross validation, nested K-fold cross validation, Hold-out cross validation, leave-one-out cross validation, etc.) and external verification (such as external cohort verification). During verification, the evaluation method is usually the receiver operating characteristic curve (ROC) and its area under the curve (Area under the curve, AUC), or the net reclassification index (Net reclassification index, NRI), the integrated discrimination improvement index (Integrated discrimination improvement, IDI) and other indicators. Taking AUC as an example, the larger the AUC, the better the discrimination ability of the machine learning classifier, that is, the higher the accuracy used for evaluation. Generally speaking, AUC<0.6 is considered to have poor discrimination, 0.6~0.75 is considered to have a certain ability to distinguish, and >0.75 is considered to have good discrimination.

[0062] A sixth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the functions of the system as described in the fifth aspect of the present invention are implemented.

[0063] The present invention also provides a method for assessing the prognostic risk of non-small cell lung cancer, comprising the following steps:

[0064] S1, obtaining the expression levels of the above miRNA marker combination in pre-treatment samples of non-small cell lung cancer subjects;

[0065] S2, evaluating the prognostic risk of the subject using a machine learning classifier and the expression level of a miRNA marker combination, wherein the machine learning classifier is constructed using population data, and the population data includes the expression level of the miRNA marker combination of multiple non-small cell lung cancer patients with different prognostic risks.

[0066] The present invention also provides another method for assessing the prognostic risk of non-small cell lung cancer, comprising the following steps:

[0067] Obtaining the expression levels of the above miRNA marker combinations in pre-treatment samples of non-small cell lung cancer subjects;

[0068] If the expression level of one or more miRNA markers in the miRNA marker combination is different from the control or a preset threshold, it indicates that the subject has a high risk of prognosis.

[0069] The present invention also provides a method for treating patients with non-small cell lung cancer, comprising the following steps:

[0070] Obtaining the expression level of the above miRNA marker combination in the pre-treatment sample of the non-small cell lung cancer patient;

[0071] Assessing the patient's prognostic risk using a machine learning classifier and the expression level of a miRNA marker combination, wherein the machine learning classifier is constructed using population data, and the population data includes the expression level of the miRNA marker combination of a plurality of non-small cell lung cancer patients with different prognostic risks;

[0072] If the patient has a high risk of prognosis, cancer treatment is administered to the patient, wherein the cancer treatment is selected from one or more of chemotherapeutic agents, radiotherapy, targeted cancer therapeutic agents and immunotherapeutic agents, or one or more of chemotherapeutic agents, radiotherapy, targeted cancer therapeutic agents and immunotherapeutic agents are selected for cancer treatment before or after surgical treatment.

[0073] The present invention also provides another method for treating a patient with non-small cell lung cancer, comprising the following steps:

[0074] Obtaining the expression level of the above miRNA marker combination in the blood sample of the non-small cell lung cancer patient before treatment;

[0075] If the expression level of one or more miRNA markers in the miRNA marker combination is different from the control or a preset threshold, cancer treatment is administered to the patient, wherein the cancer treatment is selected from one or more of chemotherapeutic agents, radiotherapy, targeted cancer therapeutic agents and immunotherapeutic agents, or one or more of chemotherapeutic agents, radiotherapy, targeted cancer therapeutic agents and immunotherapeutic agents are selected for cancer treatment before or after surgical treatment.

[0076] In the present invention, the non-small cell lung cancer refers to early stage non-small cell lung cancer, preferably non-small cell lung cancer clinically diagnosed as stage I and stage II.

[0077] In some embodiments of the present invention, the non-small cell lung cancer includes lung squamous cell carcinoma and lung adenocarcinoma.

[0078] In the present invention, the prognosis refers to treatment or intervention for patients with non-small cell lung cancer, including but not limited to surgery, radiotherapy and / or chemotherapy. In some embodiments of the present invention, it is surgery.

[0079] Furthermore, the prognostic risk refers to the situation or possibility of recurrence, complication, death, etc. of the patient after treatment or intervention. As described herein, assessing the prognostic risk includes determining or assessing the overall survival rate of patients with early non-small cell lung cancer, and the overall survival rate is the 5-year survival rate.

[0080] In the present invention, the sample is a body fluid sample, such as a serum or plasma sample.

[0081] In the present invention, the marker may refer to a gene, protein or miRNA whose expression level or concentration in a sample is changed compared to a control. As described herein, a control refers to the expression level or concentration of a marker indicating a different result or associated with a different result compared to a result of interest. For example, a marker may be a miRNA whose expression level or concentration is changed (e.g., increased or decreased) compared to a control in a sample of a subject with a disease and / or result (e.g., suffering from non-small cell lung cancer and not surviving a predetermined survival period). A control may also be an average level or mean of miRNA expression or concentration in a group of control subject samples, and these control subjects do not have a disease and / or result (e.g., suffering from early non-small cell lung cancer and surviving a predetermined survival period; i.e., a negative control). It should be understood by those skilled in the art that it is not necessary to obtain a sample from a subject with or without a disease and / or a specified result for the comparison of expression levels in the control, and to test the sample while testing. In some embodiments, the control can be a control sample incorporated into the kit, or a threshold setting or cutoff representing a range of biomarker expression, wherein expression levels falling within the range will identify a subject as having a high or low prognostic risk (e.g., overall survival rate).

[0082] In the present invention, the evaluation is auxiliary, and in clinical application, it can also or needs to be further judged in combination with clinical pathological factors. The clinical pathological factors include, but are not limited to, pathological histological type, tumor differentiation degree, maximum tumor diameter (cm), whether the tumor invades nerves, whether it involves the visceral pleura, whether it invades blood vessels, whether there is a tumor thrombus, and whether the resection margin is positive. Preferably, the clinical pathological factors include the maximum tumor diameter ≥ 3 cm, whether it invades blood vessels, and whether the resection margin is positive.

[0083] Beneficial effects of the present invention

[0084] It is well known to those skilled in the art that not all miRNA markers that can be used for cancer detection are applicable to prognosis, which will lead to uncertainty in which miRNA markers the research team of the present invention should screen as 2-miRNA to 8-miRNA combinations:

[0085] For example, a study published in 2013 pointed out that the expression levels of hsa-296-5p, hsa-191-5p, hsa-145-5p, hsa-24-3p, and let-7f-5p in plasma samples can be used to distinguish NSCLC patients from healthy people, but have no correlation with the OS of NSCLC patients (Sanfiorenzo C, Ilie MI, Belaid A, Barlési F, Mouroux J, Marquette CH, Brest P, Hofman P. Two panels of plasma microRNAs as non-invasive biomarkers for prediction of recurrence in resectable NSCLC [J]. PLoS One. 2013, 8(1): e54596).

[0086] For example, a 2019 analysis of the application value of miRNA-155 as a biomarker in lung cancer showed that the AUC of miRNA-155 in the diagnosis of lung cancer was 0.87 (95% CI: 0.84-0.90), and its significantly increased expression level was not associated with poor OS, DFS, and PFS (Shao C, Yang F, Qin Z, Jing X, Shu Y, Shen H. The value of miR-155 as a biomarker for the diagnosis and prognosis of lung cancer: a systematic review with meta-analysis [J]. BMC Cancer. 2019, 19 (1): 1103).

[0087] For example, a study on early stage lung squamous cell carcinoma showed that the level of miR-324-3p in plasma can be used as an early detection marker for lung squamous cell carcinoma, but is not suitable for prognosis (X. Gao, Y. Wang, H. Zhao, F. Wei, X. Zhang, Y. Su, C. Wang, H. Li, X. Ren, Plasma miR-324-3p and miR-1285 as diagnostic and prognostic biomarkers for early stage lung squamous cell carcinoma[J]. Oncotarget. 2016, 7(37): 59664-59675).

[0088] Compared with the prior art, the beneficial effects of the present invention include at least:

[0089] ① The present invention innovatively discovered for the first time that hsa-miR-7-1-3p, hsa-miR-1280 and / or hsa-miR-223-3p can be used as markers for early NSCLC prognosis risk assessment, and the above markers are of great value in assessing the prognosis of early NSCLC patients.

[0090] ②The miRNA marker combination of the present invention (including at least one of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280) can be used to assess the prognostic risk of non-small cell lung cancer.

[0091] ③ The effect of risk assessment of patients with non-small cell lung cancer using the miRNA marker combination, kit, method, system or medium of the present invention is better than that of the NCCN guidelines and CSCO guidelines, and has a very high clinical application value.

[0092] ④ The performance of evaluating the prognostic risk of non-small cell lung cancer is better by combining the miRNA markers of the present invention with clinical pathological factors (CP model). BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 The overall survival curve (follow-up period of 8 years) of the population samples screened for miRNA markers for assessing the prognostic risk of early non-small cell lung cancer in Example 1 of the present invention is shown.

[0094] Figure 2 The ROC curves of the Cox models constructed using hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280 respectively in Example 2 of the present invention for evaluating the prognostic risk of non-small cell lung cancer are shown.

[0095] Figure 3The Kaplan-Meier survival curves of the high-risk prognostic group and the low-risk prognostic group after prognostic risk stratification using the Cox model constructed using hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280 in Example 2 of the present invention are shown.

[0096] Figure 4 The performance distribution of the Cox model constructed using different numbers of miRNAs in Example 3 of the present invention for evaluating the prognostic risk of non-small cell lung cancer is shown.

[0097] Figure 5 The ROC curve of the 3-miRNA model in Example 3 of the present invention for evaluating the prognostic risk of non-small cell lung cancer is shown.

[0098] Figure 6 The ROC curve of the 7-miRNA model in Example 3 of the present invention for evaluating the prognostic risk of non-small cell lung cancer is shown.

[0099] Figure 7 The Kaplan-Meier survival curves of the high-risk prognosis group and the low-risk prognosis group after the 3-miRNA model in Example 3 of the present invention was used for prognostic risk stratification are shown.

[0100] Figure 8 The Kaplan-Meier survival curves of the high-risk prognosis group and the low-risk prognosis group after the 7-miRNA model in Example 3 of the present invention was used for prognostic risk stratification are shown.

[0101] Fig. 9 The Kaplan-Meier survival curves of the high-risk prognosis group and the low-risk prognosis group after the 8-miRNA model in Example 3 of the present invention was used for prognostic risk stratification are shown.

[0102] Fig.10 The ROC curves of the 8-miRNA model in Example 4 of the present invention, the NCCN guidelines, and the CSCO guidelines for evaluating the prognostic risk of non-small cell lung cancer are shown.

[0103] Fig.11 The Kaplan-Meier survival curves of the high-risk prognosis group and the low-risk prognosis group after prognostic risk stratification using the 8-miRNA model in Example 4 of the present invention, the NCCN guidelines, and the CSCO guidelines are shown. DETAILED DESCRIPTION

[0104] Unless otherwise indicated, implied from the context, or customary in the prior art, all parts and percentages in this application are based on weight, and the tests and characterization methods used are all synchronized with the filing date of this application. Where applicable, the contents of any patent, patent application or disclosure involved in this application are fully incorporated herein by reference, and their equivalent patent families are also introduced as references, especially the definitions of relevant terms in the art disclosed in these documents. If the definition of a specific term disclosed in the prior art is inconsistent with any definition provided in this application, the definition of the term provided in this application shall prevail.

[0105] In order to make the technical problems, technical solutions and beneficial effects solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments.

[0106] The following examples are used to demonstrate preferred embodiments of the present invention. It will be appreciated by those skilled in the art that the techniques disclosed in the following examples represent techniques discovered by the inventors that can be used to implement the present invention and therefore can be considered as preferred embodiments of the present invention. However, it will be appreciated by those skilled in the art based on this specification that many modifications may be made to the specific embodiments disclosed herein and still achieve the same or similar results without departing from the spirit or scope of the present invention.

[0107] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, and the disclosure and materials cited therein are hereby incorporated by reference.

[0108] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many technical equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the claims.

[0109] The experimental methods in the following examples are conventional methods unless otherwise specified. The instruments and equipment used in the following examples are conventional laboratory instruments and equipment unless otherwise specified; the experimental materials used in the following examples are purchased from conventional biochemical reagent stores unless otherwise specified.

[0110] Example 1 Screening of miRNA markers for assessing the prognostic risk of early non-small cell lung cancer

[0111] (1) Research cohort and sample source

[0112] In a study evaluating prognostic and recurrence markers for non-small cell lung cancer, 280 patients with stage I non-small cell lung cancer were used for preoperative blood. According to the average follow-up results of 8 years, 84 (30.4%) patients died and 195 (69.6%) patients were still alive before the follow-up cutoff time (e.g. Figure 1 As shown). The 84 deaths were defined as the experimental group (high-risk prognosis group), and the 195 survivors were defined as the control group (low-risk prognosis group) for screening and validating miRNA markers or miRNA marker combinations for assessing the prognostic risk of early non-small cell lung cancer. The 280 blood samples and clinical and follow-up information in the cohort were all from Zhejiang Cancer Hospital, specifically, 280 NSCLC patients who underwent surgical treatment at Zhejiang Cancer Hospital from April 2008 to September 2019 (inclusion criteria: diagnosed with primary lung cancer, no other treatment before surgical treatment), and the blood samples used in the present invention were serum samples collected from the above-mentioned population before surgery.

[0113] (2) RT-qPCR

[0114] The total RNA sample including miRNA was extracted from the serum sample according to the operating requirements of the Qiagen miRNeasy Serum / Plasma Kit; The reverse transcription kit was used for reverse transcription-fluorescence PCR amplification according to the instructions, and the expression levels of 155 miRNAs (which can be stably detected in all samples) were obtained. The expression levels of each miRNA were further normalized according to the average expression level of the miRNA internal reference (miR-361-5p and miR-425-5p). The analysis found that the relative expression levels of 22 miRNAs were significantly different between the experimental group and the control group.

[0115] (3) Marker screening

[0116] In order to further screen miRNAs that can be used as prognostic risk assessment markers for early non-small cell lung cancer, the inventors used Sequential Floating Forward Selection (SFFS) and support vector machine (SVM) to further screen the above 22 miRNAs, specifically:

[0117] The candidate marker set X starts from an empty set. Each time, a subset x is selected from the unselected miRNAs to be added, so that the performance of the SVM model established after adding the subset x is optimal, that is, the area under the ROC curve (AUC) is maximized. Then, a subset z is selected from the selected miRNAs, so that the performance of the SVM model established after removing the subset z is optimal.

[0118] The final candidate marker set X includes 8 miRNA markers, namely hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280. Their names, accession numbers and sequence information in MiRBase (http: / / www.mirbase.org) are shown in Table 1.

[0119] Table 1 Information of 8 miRNA markers

[0120]

[0121]

[0122] Example 2 Performance of evaluation model based on single miRNA marker construction

[0123] For each individual miRNA marker in the candidate marker set X screened in Example 1, a proportional hazards model (Cox model for short) was constructed, and validation was performed using leave one out cross validation (LOOCV). For each Cox model, during validation, values ​​above the cutoff were considered high risk for prognosis, and values ​​below the cutoff were considered low risk for prognosis, thereby completing prognostic risk stratification (classification or categorization). The evaluation results are shown in Tables 2 and Figure 2 shown.

[0124] Table 2 Evaluation results of Cox models constructed with single miRNA markers

[0125]

[0126] Note: Positive means that the patient is assessed as having a high risk of prognosis; negative means that the patient is assessed as having a low risk of prognosis.

[0127] As shown in Table 2, the specificity of the Cox model constructed using hsa-miR-223-3p and hsa-miR-7-1-3p was relatively high, reaching about 0.9, while the sensitivity of the Cox model constructed using hsa-miR-1280 was very high, reaching 0.978.

[0128] According to the risk stratification results of each Cox model, Kaplan-Meier survival curves were drawn with survival time (years) as the horizontal axis and OS (ratio) as the vertical axis. The results are shown in Figure 3 shown.

[0129] From Table 2 and Figure 2-Figure 3 It can be seen that the Cox model constructed by the eight miRNA markers alone cannot effectively evaluate the prognostic risk of early non-small cell lung cancer. It is worth noting that when hsa-miR-223-3p and hsa-miR-7-1-3p were used alone to construct the Cox model, the difference in overall survival between the high-risk prognostic group and the low-risk prognostic group reached a significant level (p<0.05).

[0130] Example 3 Performance of evaluation model based on miRNA marker combination construction

[0131] Based on the results of Example 2, the inventors further constructed Cox models using miRNA marker combinations including at least 2 miRNA markers, and used LOOCV for verification. Similarly, for each Cox model, when verifying, those above the cutoff were high-risk prognoses, and those below the cutoff were low-risk prognoses.

[0132] The performance (AUC) distribution of the Cox model including different numbers of miRNAs is shown in Figure 2. Figure 4 As shown by Figure 4 It can be seen that, in general, the more miRNAs are included, the better the performance.

[0133] The performance results of some models are shown in Table 3 (only the results of models with AUC greater than 0.65 are retained).

[0134] Table 3 Evaluation results of the Cox model constructed by miRNA marker combination

[0135]

[0136]

[0137]

[0138]

[0139] Note: In the first column of the table, A represents hsa-miR-223-3p; B represents hsa-miR-324-5p; C represents hsa-miR-596; D represents hsa-miR-7-1-3p;

[0140] E represents hsa-miR-320a; F represents hsa-miR-320b; G represents hsa-miR-1275; H represents hsa-miR-1280.

[0141] from Figure 4As can be seen from Table 3, as the number of miRNA markers in the miRNA marker combination gradually increases to 8, the AUC of the constructed Cox model for assessing prognostic risk continues to increase. For different miRNA marker combinations containing the same number of miRNA markers, the AUC and ln(HR) of the constructed Cox model for assessing prognostic risk are also significantly different. The example performance of the Cox model constructed by different miRNA marker combinations including 2, 3, 4, 5, 6, 7 and 8 miRNA markers is as follows:

[0142] For different miRNA marker combinations including 2 miRNA markers, the Cox model constructed by hsa-miR-223-3p and hsa-miR-324-5p had an AUC of 0.687 and ln(HR) of 0.609 for assessing prognostic risk. This Cox model was recorded as the 2-miRNA model.

[0143] For different miRNA marker combinations including 3 miRNA markers, the AUC of the Cox model constructed by hsa-miR-223-3p, hsa-miR-324-5p and hsa-miR-596 for assessing prognostic risk reached 0.699 ( Figure 5 ), ln(HR) reached 0.667, and the Cox model was recorded as the 3-miRNA model.

[0144] For different miRNA marker combinations including 4 miRNA markers, the Cox model constructed by hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-320a and hsa-miR-320b for evaluating prognostic risk reached an AUC of 0.746 and ln(HR) of 0.741. This Cox model was recorded as the 4-miRNA model.

[0145] For different miRNA marker combinations including 5 miRNA markers, the Cox model constructed by hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-320a and hsa-miR-320b for evaluating prognostic risk reached an AUC of 0.753 and ln(HR) of 0.755. This Cox model was recorded as the 5-miRNA model.

[0146] For different miRNA marker combinations including 6 miRNA markers, the Cox model constructed by hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-320a, hsa-miR-320b and hsa-miR-1280 for evaluating prognostic risk reached an AUC of 0.758 and ln(HR) of 0.759. This Cox model was recorded as the 6-miRNA model.

[0147] For different miRNA marker combinations including 7 miRNA markers, the AUC of the Cox model constructed by hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280 for assessing prognostic risk reached 0.759 ( Figure 6 ), ln(HR) reached 0.808, and the Cox model was recorded as the 7-miRNA model.

[0148] For the miRNA marker combination including all 8 miRNA markers (hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280), the AUC of the constructed Cox model for assessing prognostic risk reached 0.766, and the Cox model was recorded as the 8-miRNA model.

[0149] The above-mentioned 2-miRNA model, 3-miRNA model, 4-miRNA model, 5-miRNA model, 6-miRNA model, 7-miRNA model and 8-miRNA model were used for prognostic risk stratification, and Kaplan-Meier survival analysis was performed according to the risk stratification results. The hazard ratio (HR) of each Cox model in the training set and the validation set was statistically analyzed. The results are shown in Table 4.

[0150] Table 4 Differences in hazard ratios of different Cox model risk stratifications in the training set and validation set

[0151]

[0152]

[0153] As shown in Table 4, when the number of miRNA markers in the miRNA marker combination used to construct the Cox model gradually increased to 8, the ln(HR) value continued to increase, and the 8-miRNA model had the best performance in evaluating the prognostic risk of early non-small cell lung cancer.

[0154] The Kaplan-Meier survival curve of the stratified results according to the prognostic risk assessment of the 3-miRNA model is shown in Figure 7 As shown, after prognostic risk stratification according to the 3-miRNA model, the Kaplan-Meier survival curve was clearly distinguished, with a significant difference (p=2.06E-04).

[0155] The Kaplan-Meier survival curve of the stratified results according to the prognostic risk assessment of the 7-miRNA model is shown in Figure 8 As shown, after prognostic risk stratification according to the 7-miRNA model, the Kaplan-Meier survival curve was clearly distinguished, with a significant difference (p=3.37E-06).

[0156] The Kaplan-Meier survival curve of the stratified results according to the prognostic risk assessment of the 8-miRNA model is shown in Fig. 9 As shown, after prognostic risk stratification according to the 8-miRNA model, the Kaplan-Meier survival curves were more clearly differentiated, and the difference was very significant, p = 1.88E-6.

[0157] In summary, the 8-miRNA model has the best performance in prognostic risk assessment. In the 8-miRNA model, the regression coefficients of each miRNA are shown in Table 5.

[0158] Table 5 Regression coefficients of each miRNA in the 8-miRNA model

[0159] miRNA marker name Regression coefficient p-value hsa-miR-223-3p -0.686 0.0003 hsa-miR-324-5p 0.814 0.0011 hsa-miR-596 0.368 0.0064 hsa-miR-7-1-3p -0.675 0.0079 hsa-miR-320a 1.430 <0.0001 hsa-miR-320b -1.147 0.0007 hsa-miR-1275 -0.440 0.0233 hsa-miR-1280 0.350 0.0463

[0160] In Table 5, it is important to note the signs of the regression coefficients; a positive sign (upregulation) indicates that the miRNA marker is a bad risk factor, and a negative sign (downregulation) indicates that the miRNA marker is a good risk factor.

[0161] Example 4 Further verification of the performance of the evaluation model constructed based on the miRNA marker combination

[0162] In order to further verify the performance of the 8-miRNA model in Example 3 for the prognostic risk assessment of patients with early non-small cell lung cancer, the inventors further performed prognostic risk stratification on 280 patients with early non-small cell lung cancer according to the NCCN guidelines, CSCO guidelines and CP model (Clinicopathological panel). The performance comparison results are shown in Table 6 and Fig.10 As shown. The CP model is obtained as follows:

[0163] Clinical data of the included patients were collected through the hospital medical record system, including pathological histological type, tumor differentiation degree, maximum tumor diameter (cm), whether the tumor invaded nerves, whether it involved the visceral pleura, whether it invaded blood vessels, whether there was tumor thrombus, whether there was positive resection margin, etc. Univariate Cox regression analysis was used to calculate the correlation between clinical pathological factors and OS, and independent risk factors related to prognosis were screened by forward stepwise selection. Among them, the maximum tumor diameter ≥ 3 cm, whether it invaded blood vessels, and whether it was positive resection margin were significantly correlated with OS, and the Cox model of clinical factors, namely the CP model, was established as a variable.

[0164] Furthermore, the inventors added the scores of the 8-mRNA model and the CP model to obtain the 8-miRNA+CP model.

[0165] Table 6 Comparison of the evaluation results of the 8-miRNA model with the evaluation performance based on the NCCN guidelines and CSCO guidelines

[0166]

[0167] Kaplan-Meier survival curves based on the 8-miRNA model evaluation results and the NCCN guidelines, CSCO guidelines and CP model evaluation results were drawn respectively, as shown in Fig.11 shown.

[0168] From Table 6 and Fig.11 It can be seen that the 8-miRNA model is more effective in stratifying the prognostic risk of 280 patients with early-stage non-small cell lung cancer than the evaluation results based on the CSCO guidelines, NCCN guidelines and CP model. The risk stratification effect of the 8-miRNA+CP model is better than that of the single model (i.e., 8-miRNA model or CP model).

[0169] Example 5 Performance evaluation of multiple models based on a single miRNA marker

[0170] For the 8 miRNA markers obtained in Example 2, the inventors traversed all combinations and calculated the average AUC of the Cox model constructed by each specific miRNA marker, as well as 2 miRNA marker combinations, 3 miRNA marker combinations, 4 miRNA marker combinations, 5 miRNA marker combinations, 6 miRNA marker combinations, 7 miRNA marker combinations and 8 miRNA marker combinations containing the miRNA marker, as shown in Table 7.

[0171] Table 7 Mean AUC values ​​of Cox models constructed using miRNA marker combinations containing specific miRNA markers

[0172]

[0173] Furthermore, the inventors evaluated the performance of the Cox model constructed using a combination of miRNA markers with hsa-miR-223-3p, hsa-miR-7-1-3p and hsa-miR-1280 as the core, including the mean AUC, mean sensitivity, mean specificity, mean positive predictive value (PPV) and mean negative predictive value (NPV), as shown in Tables 8 to 10, respectively.

[0174] Table 8 Performance of Cox model constructed with miRNA marker combination including hsa-miR-223-3p

[0175] miRNA marker panel Mean AUC Mean sensitivity Mean specificity PPV mean Mean NPV hsa-miR-223-3p+1 miRNA 0.645 0.606 0.655 0.283 0.897 hsa-miR-223-3p+2 miRNAs 0.661 0.640 0.638 0.277 0.901 hsa-miR-223-3p+3 miRNAs 0.678 0.657 0.646 0.287 0.906 hsa-miR-223-3p+4 miRNAs 0.697 0.655 0.674 0.306 0.909 hsa-miR-223-3p+5 miRNAs 0.719 0.681 0.682 0.317 0.915 hsa-miR-223-3p+6 miRNAs 0.742 0.673 0.723 0.342 0.917 hsa-miR-223-3p+7 miRNAs 0.766 0.689 0.773 0.383 0.924

[0176] Table 9 Performance of Cox model constructed with miRNA marker combination including hsa-miR-7-1-3p

[0177]

[0178]

[0179] Table 10 Performance of Cox model constructed with miRNA marker combination including hsa-miR-1280

[0180] miRNA marker panel Mean AUC Mean sensitivity Mean specificity PPV mean Mean NPV hsa-miR-1280+1 miRNA 0.559 0.540 0.618 0.234 0.871 hsa-miR-1280+2 miRNAs 0.607 0.592 0.634 0.262 0.889 hsa-miR-1280+3 miRNAs 0.645 0.581 0.689 0.291 0.892 hsa-miR-1280+4 miRNAs 0.677 0.580 0.725 0.316 0.897 hsa-miR-1280+5 miRNAs 0.706 0.636 0.710 0.326 0.908 hsa-miR-1280+6 miRNAs 0.736 0.686 0.701 0.330 0.917 hsa-miR-1280+7 miRNAs 0.766 0.689 0.773 0.383 0.924

[0181] It can be seen from Tables 8 to 10 that the Cox model constructed using a combination of multiple miRNA markers including hsa-miR-223-3p, hsa-miR-7-1-3p and hsa-miR-1280 has excellent performance in evaluating the prognostic risk of early non-small cell lung cancer.

[0182] All documents mentioned in the present invention are cited as references in this application, just as each document is cited as reference individually. In addition, it should be understood that after reading the above teachings of the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the claims attached to this application.

Claims

1. A miRNA marker combination for assessing the prognostic risk of non-small cell lung cancer, It is characterized in that The miRNA marker combination includes at least one of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280.

2. The miRNA marker combination according to claim 1, It is characterized in that The miRNA marker combination includes hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280.

3. The miRNA marker combination according to claim 1 or 2, It is characterized in that The miRNA marker combination consists of hsa-miR-223-3p, hsa-miR-324-5p, hsa-miR-596, hsa-miR-7-1-3p, hsa-miR-320a, hsa-miR-320b, hsa-miR-1275 and hsa-miR-1280.

4. Use of the miRNA marker combination according to any one of claims 1 to 3 in the preparation of a kit for assessing the prognostic risk of non-small cell lung cancer.

5. Use of the detection reagent of the miRNA marker combination according to any one of claims 1 to 3 in the preparation of a kit for assessing the prognostic risk of non-small cell lung cancer.

6. The use according to claim 5, It is characterized in that The detection reagent includes primers and / or probes that specifically bind to each miRNA marker in the miRNA marker combination.

7. A kit for assessing the prognostic risk of non-small cell lung cancer, It is characterized in that It comprises the miRNA marker combination and / or its detection reagent according to any one of claims 1 to 3.

8. The kit according to claim 7, It is characterized in that Also included are RNA extraction reagents, reverse transcription reagents and / or universal reagents for qPCR amplification.

9. A system for assessing the prognostic risk of non-small cell lung cancer, It is characterized in that Includes the following modules: A data input module for receiving the expression level of the miRNA marker combination according to any one of claims 1 to 3 in a blood sample of a subject, wherein the subject suffers from non-small cell lung cancer; A database storage module, used to store population data, wherein the population data includes the expression level of the miRNA marker combination of a plurality of non-small cell lung cancer patients with different prognostic risks; Evaluation module: connected to the data input module and the database storage module respectively, for constructing a machine learning classifier using population data, and evaluating the prognostic risk of the subject using the machine learning classifier and the expression level of the miRNA marker combination of the subject.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, which implements the functions of the system according to claim 8 or 9 when executed by a processor.

Images