CN111218513A - Peripheral blood extracellular vesicle microRNA biomarker for early diagnosis of lung cancer and application thereof - Google Patents

Abstract

The invention discloses a peripheral blood extracellular vesicle miRNA biomarker for early diagnosis of lung cancer, which is characterized by comprising hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p and hsa-miR-3168. The invention is based on the peripheral blood of the lung nodule patients detected by the LDCT in clinic as a research sample, has higher specificity and is more reliable, and the research result has important significance for accurately identifying the false-positive lung cancer patients from the lung nodule patients detected by the LDCT.

Description

Peripheral blood extracellular vesicle microRNA biomarker for early diagnosis of lung cancer and application thereof

Technical Field

The invention relates to the technical field of early detection of diseases, in particular to a peripheral blood extracellular vesicle microRNA biomarker for early diagnosis of lung cancer.

Background

Lung cancer is the most common malignancy with the highest incidence and mortality worldwide. Early stage lung cancer is insidious and usually has no symptoms, but most patients are in the middle and late stage at the initial diagnosis and lose the chance of surgical resection. The five-year survival rate of patients with advanced lung cancer is less than 5 percent, and the five-year survival rate of patients with early lung cancer can reach more than 90 percent. Therefore, early diagnosis is an important opportunity for a good prognosis for lung cancer patients.

Lung cancer usually occurs in early stages in the form of lung nodules. The rapid development of imaging techniques has increased the detection rate of lung nodules. However, the pulmonary nodule disease is complex, and the clinical manifestation lacks specificity, so that the differential diagnosis of benign and malignant pulmonary nodules has certain difficulty. Generally, 40-50% of nodules with a diameter >3 cm are malignant, 25% -30% of nodules with a diameter between 0.5-1 cm are malignant, and only 1% of nodules with a diameter <0.5 cm are malignant. Studies have shown that a significant proportion of lung nodule patients exhibit post-operative pathology as benign disease. Surgical resection of such benign lung nodules can result in over-treatment and increase pain for this portion of the patient. Therefore, how to accurately diagnose lung cancer patients carrying malignant nodules from patients carrying different lung nodules is clinically significant.

At present, early diagnosis methods of lung cancer include breast imaging, bronchoscopy technology, sputum cast-off cytology detection and the like, but the detection effects of the methods are not ideal. The sensitivity of sputum examination to central lung cancer is about 50%, while it is less than 20% for peripheral lung cancer. The bronchoscopy technique has a diagnosis rate of about 90% for central lung cancer, but less than 30% for peripheral lung cancer, especially for precancerous lesions. Chest imaging examination methods include X-ray chest radiography (CxR), low dose helical CT (LDCT), and PET-CT. CxR the misdiagnosis rate is as high as 50-90%, and the specificity of detecting the nodules in the lung by LDCT and PET-CT is poor. It is reported that the detection of LDCT technology widely used in early screening and diagnosis of lung cancer can generate over 21% of false positives. Therefore, it is difficult to accurately diagnose early lung cancer by using imaging method alone, and there is no effective biomarker combined with imaging technology (LDCT) to improve the specificity of early lung cancer diagnosis.

Extracellular Vesicles (EVs) are vesicles produced by different cells and released to the outside of the cells, and can be classified into various types. Among them, small extracellular vesicles, also known as exosomes, are membrane vesicles of about 30-150 nm in diameter. It is derived from vesicles of late endocytosis bodies (MVB), which are formed by the inward recession of endocytosis vesicle membranes of cells to form a multivesicular body containing a plurality of vesicles, and the multivesicular body is fused with cell membranes and released into extracellular matrix. The extracellular vesicle-carried contents include proteins, lipids, mRNA, rRNA, miRNA, etc. Cells secrete extracellular vesicles under both normal and pathological conditions, which can participate in the transfer of information between cells. The content of the extracellular vesicles can represent certain physiological and pathological conditions. Among various diseases (e.g., malignant tumor, immune disease, etc.), extracellular vesicles (exosomes) free in peripheral blood have been widely noticed and studied as an important form of liquid biopsy.

Micro nucleic acids (microRNAs, miRNAs) are a class of non-coding small-molecule RNAs with the length of 19-25 nucleotides discovered in recent years. It mainly degrades the target gene mRNA or inhibits its translation through the complete or incomplete pairing with the target gene 3' UTR, thereby participating in the regulation of individual development, apoptosis, proliferation and differentiation and other life activities, and playing a role similar to oncogene or cancer suppressor gene in the process of generating and developing tumors. The expression profile of miRNA has obvious tissue specificity and specific expression pattern in different tumors. These characteristics make mirnas possible to become new biological markers and therapeutic targets for tumor diagnosis. Recently, there are studies suggesting that some differences exist between blood extracellular vesicle mirnas in healthy people and lung cancer patients, lung cancer adenocarcinoma and lung squamous carcinoma patients. However, the blood extracellular vesicle miRNA is rarely studied as a biomarker in identifying good and malignant lung nodules, and no biomarker capable of improving the specificity of early diagnosis of lung cancer is found and verified.

According to the invention, based on the research, peripheral blood of a lung nodule patient detected by clinical LDCT is taken as a research sample, peripheral blood extracellular vesicles are extracted, the expression of peripheral blood extracellular vesicle miRNA of a benign patient and a malignant patient is further detected by small RNA sequencing technology, and a high-specificity peripheral blood extracellular vesicle miRNA biomarker diagnosis model for early diagnosis of lung cancer is found and verified in two independent groups.

Disclosure of Invention

The invention aims to provide a peripheral blood extracellular vesicle miRNA marker for early diagnosis of lung cancer, based on a large amount of sample verification, 5 specific diagnosis markers suitable for early diagnosis of lung cancer are determined, and compared with other reported miRNA markers, the sample provided by the invention is based on clinical LDCT to detect blood of a lung nodule patient as a research sample, so that the sample has higher specificity; the 5 miRNAs are provided for the first time as early diagnosis markers of lung cancer, and are more reliable compared with other miRNA molecular markers. The method has important significance for further identifying the false positive patients from the lung nodule patients detected by LDCT.

The technical scheme adopted by the invention for solving the technical problems is as follows:

a peripheral blood extracellular vesicle miRNA marker for early diagnosis of lung cancer comprises hsa-let-7b-3p, hsa-miR-101-3p and hsa-miR-3168.

Further, the miRNA biomarker also comprises one or more miRNAs selected from hsa-miR-125b-5p, hsa-miR-150-5p, hsa-let-7d-3p, hsa-miR-503-5p, hsa-miR-16-5p, hsa-miR-17-5p, hsa-miR-19b-3p, hsa-miR-532-5p and hsa-miR-98-5 p.

Preferably, the miRNA biomarker can be a combination of hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p and hsa-miR-3168.

Preferably, the miRNA biomarker can be a combination of hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p, hsa-miR-3168, hsa-let-7d-3p, hsa-miR-19b-3p and hsa-miR-503-5 p.

Preferably, the miRNA biomarker can also be a combination of hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p, hsa-miR-3168, hsa-miR-16-5p, hsa-miR-17-5p, hsa-miR-19b-3p, hsa-miR-532-5p and hsa-miR-98-5 p.

Further, the expression values of the peripheral blood extracellular vesicle miRNA markers in the peripheral blood extracellular vesicles of patients diagnosed with lung cancer are statistically significantly different from the expression values in the control samples.

Further, the control sample is a benign lung nodule patient, not the general normal population.

A kit for early diagnosis of lung cancer, which comprises the detection reagent of the peripheral blood extracellular vesicle miRNA marker.

Use of a peripheral blood extracellular vesicle miRNA marker for the preparation of a diagnostic agent for predicting the early stage lung cancer of a subject by a method comprising:

-collecting a sample of peripheral blood extracellular vesicles, detecting the presence of miRNA in peripheral blood extracellular vesicles obtained from the subject;

-measuring the expression level of said miRNA biomarker in said peripheral blood extracellular vesicle sample;

-using a risk scoring model based on previously measured expression levels of mirnas to predict the likelihood of the subject having early stage lung cancer.

Further, the expression level of the miRNA is obtained by quantification based on the result of the small RNA sequencing detection technology.

Further, the risk scoring model includes a mathematical model formulated as follows:

wherein Risk-Score is a lung nodule benign and malignant Risk prediction value, Gi represents an expression value of an ith miRNA, β i represents a Risk Score parameter value of the ith miRNA, i = 1, 2 … n, n is the total number of lung nodule benign and malignant biomarkers, and α represents a model correction parameter.

Drawings

FIG. 1 shows the results of electron microscopy of extracellular vesicles.

FIG. 2 protein expression characteristic of extracellular vesicles.

FIG. 3. training queue ROC curve.

FIG. 4 is a validation queue ROC curve.

FIG. 5. construction of a risk scoring model with 3 miRNAs as biomarkers, ROC curve analysis in training trains.

FIG. 6. construction of a risk scoring model with 8 miRNAs as markers, ROC curve analysis in training trains.

FIG. 7. construction of a risk scoring model using 10 miRNAs as markers, ROC curve analysis in training trains.

Detailed Description

All of the miRNAs involved in the present invention are mature miRNAs annotated in the miRBase database (http:// www.mirbase.org /).

Screening process of peripheral blood extracellular vesicle miRNA marker

(1) Research cohort and clinical information

The study included a total of 109 patients with LDCT as lung nodules from a total of two study cohorts, and plasma samples were taken from the patients prior to surgery. Each patient in the group gives an accurate diagnosis after surgery based on the results of the pathological examination. The training cohort was enrolled into 47 patients with a cohort time of 2018 months 9 to 2018 months 10, including 17 patients with benign lung nodules and 30 patients with malignant lung nodules (table 1). The validation cohort was 62 patients with a cohort time of 4 months to 5 months 2019, including 24 patients with benign lung nodules and 38 patients with malignant lung nodules (table 1). The benign specimen is classified into granuloma, inflammatory nodule, hamartoma, etc. Table 1 shows clinical information of patients, including sex, age, smoking history, and pathological diagnosis results. The analysis result shows that the sex, the age and the smoking history of the two groups of patients have no significant difference with the good and the malignant patient proportion.

TABLE 1.109 patient clinical information

(2) Extraction and characterization of peripheral blood extracellular vesicles

① collection of peripheral blood and extraction of extracellular vesicles

Peripheral blood samples were collected in anticoagulated vacuum blood collection tubes (REF367863, BD, USA) prior to surgery or drug treatment and delivered at 4 ℃ within one hour after collection. And (3) centrifuging at 4 ℃ for 10min at 1600g after receiving the blood sample, judging the hemolysis grade of the blood sample after centrifuging, transferring the supernatant into a 1.5ml centrifuge tube for centrifuging 16000g again, centrifuging at 4 ℃ for 15min, subpackaging the supernatant into 1ml per tube, and storing in a refrigerator at minus 80 ℃. The plasma sample is taken out and placed in a metal bath at 37 ℃ for incubation until the plasma sample is completely melted, 12000g is centrifuged at 4 ℃ for 10min, 500 mul of supernatant is transferred to 0.45 mul of tube filter (Costar, CLS8163-100EA, Corning, USA), 12000g is centrifuged at 4 ℃ for 5min, and filtrate is transferred to 0.22 mul of tube filter (Costar, CLS8161-100EA, USA) for 12000g and centrifuged at 4 ℃ for 5 min. The filtrate was transferred to a 1.5ml centrifuge tube and 1/4 volumes of extracellular vesicle precipitant (cat # REX015S, 3D Med, Shanghai, China) were added, after mixing, incubation was carried out at 4 ℃ for 30min, 4700g was carried out, centrifugation was carried out at 4 ℃ for 30min, and 200. mu.l of PBS (phosphate buffersaline) was added to discard the supernatant to suspend extracellular vesicle precipitation.

② characterization of peripheral blood extracellular vesicles

In order to detect the characteristics of the extracellular vesicles of the peripheral blood, the method adopts a scanning electron microscope to detect the morphology of the extracellular vesicles, and further utilizes an immune hybridization experiment to detect the expression level of characteristic proteins of the extracellular vesicles. And (3) identifying the morphological characteristics of the extracellular vesicles: after precipitation, it was resuspended in PBS, then fixed with 5% glutaraldehyde, washed with PBS for 5min, then fixed with 1% osmic acid, then dehydrated with ethanol of various concentration gradients of 40%, 60%, 80%, 96-98%, and dried for 24 hours at room temperature after plating a silicon metal film on the sample surface for analysis by scanning electron microscope (SU8020, High-Technologies, Japan). The detection result shows that the extracellular vesicle has a typical horseshoe-shaped form (see figure 1).

Detecting extracellular vesicle characteristic protein: the extracellular vesicles were precipitated with extracellular vesicle precipitant (cat # REX015S, 3DMEd, Shanghai, China) and then lysed with RIPA lysate (P0013B, Beyotime, Shanghai, China) on ice for 30 min. Electrophoresis was performed at 4% -20% SDS-PAGE gel (#4561095, Bio-Rad, USA) for about 1h at constant pressure, followed by constant current transformation on PVDF membrane (Millipore) for 45min and blocking with 5% skim milk overnight. First-order information: TSG101 (1:1000diluted, ab125011, Abcam, England), CD63 (1:1000diluted, ab216130, Abcam, England), CD9(1:1000 diluted, ab92726, Abcam, England), Alix (1:1000diluted, 2171, Cell Signaling Technology, Danvers, MA, USA) and Syntenin (1:1000diluted, ab19903, Abcam, England) antibodies were incubated for 2h at room temperature, TBST was washed for 10min four times, rabbit secondary antibody (A0208, Biyun day) or mouse secondary antibody (A0216, Biyun day) was incubated for 1h at room temperature, TBST was washed for 10min four times, and developed with chemiluminescence system (Tanon-5200Multi, Shanghai, China). The detection result shows that the extracellular vesicle characteristic proteins TSG101, CD63, CD9, Alix and Syntenin are all expressed in the representative sample extracted in the patent (see figure 2).

(3) Extraction and expression quantity of extracellular vesicle miRNA

① extraction of miRNA from extracellular vesicle of peripheral blood

MiRNeasy Serum/Plasma Kit (217184, QIAGEN, Shanghai, China) is selected for separating the miRNA of the extracellular vesicle of the peripheral blood, and the specific operation flow is according to the product instruction. miRNA quantification and fragment distribution detection are carried out by using an Agilent 2100 analyzer matched with a corresponding chip (5067-.

② expression detection of extracellular vesicles of peripheral blood

This patent adopts small RNA sequencing to detect the expression level of peripheral blood extracellular vesicle miRNA. The Library was constructed using NEBNext, Multiplex SmallRNA Library Prep Set for Illumina (E7300L, NEB, USA) kit, and the specific procedures were according to the product instructions. The loading amount of miRNA in each plasma sample is 100ng, the total volume is not more than 6 mu l, and the miRNA is connected with a 3 'joint, hybridized reverse transcription primers, connected with a 5' joint, reverse transcription and added with Illumina indexprimers for PCR amplification for 18 cycles. The PCR product was purified using a NucleoSpin Gel and PCR Clean-up (740609.50, QIAGEN, Shanghai, China) kit, and DNA was eluted with 30. mu.l of enzyme-free water. DNA quantification and fragment distribution detection were performed using LabChip GX Touch HT nucleic acid analyzer and its associated chip (CLS138948, PerkinElmer, USA) and reagents (CLS760672, PerkinElmer, USA). Typically 20-25 libraries were sequenced at equimolar ratios mixed melane, using Illumina HiSeq PE150 analyzer.

(4) Sequencing data analysis process

Based on small RNA sequencing detection technology, the expression quantity of miRNA in the patient peripheral blood extracellular vesicles is obtained. The analysis flow of the sequencing data is as follows:

① alignment of sequencing data after removing the small RNA sequencing data sequencing linker, BWA software (version: 0.7.12-r1039) was used to align the sequencing data to the human reference genome hg19 (genome download link: http:// hgdownload. soe. ucsc. edu/goldenPath/hg19/bigZips /), and the number of reads aligned to miRNA was counted.

② miRNA annotation mirnas were annotated using the Gencodev25 and miRBasev21 databases, reserving the annotated mature mirnas as known for subsequent analysis.

③ miRNA filtering, keeping mature miRNA with length less than or equal to 30nt and at least covering 2 reads for each sample in the training queue data for subsequent analysis, and keeping miRNA obtained by screening by the training queue and at least covering 2 reads for each sample in the validation queue data for subsequent analysis.

And (④) normalizing the miRNA expression quantity, namely performing miRNA expression quantity normalization on the training queue sample and the verification queue sample by using an M value weighted mean of M-values (TMM) and a limma-voom method in a limma analysis package in the R language.

(5) Discovery of biomarkers

Based on the expression quantity of miRNAs in the training queue, grouping samples according to pathological detection results, and using a statistical method to discover miRNA which can be used for distinguishing benign and malignant lung nodules as biomarkers. The process is as follows:

① the training cohorts are divided into two groups, benign lung nodule group and malignant lung nodule group according to the lung nodule pathological detection result.

② candidate biomarkers, miRNA with expression amount different in two groups of benign and malignant are analyzed by using a fitting linear model (limma-voom) method in an R language limma analysis package, wherein the expression amount is more than 50 Per Million of reads (CPM, Counts Per Million), the variation amount between the two groups is more than 1.5 times, and the test result P is less than or equal to 0.01, and the miRNA serves as candidate biomarkers, including hsa-let-7b-3P, hsa-miR-125b-5P, hsa-miR-150-5P, hsa-miR-101-3P, hsa-miR-3168 and has-miR-144-3P, and 6 in total, and is used for subsequent analysis.

③ biomarkers, in a training team, using a minimum absolute contraction and selection operator (LASSO) model to calculate The contribution degree of 6 candidate biomarkers to The prediction of The benign and malignant pulmonary nodules, and removing The candidate biomarker has-miR-144-3p with The contribution degree of 0. finally, selecting 5 miRNAs as The biomarkers, including hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p and hsa-miR-3168, for The subsequent construction of a pulmonary nodule benign and malignant pulmonary nodule risk scoring model.

Example 1: establishment and verification of risk scoring model with 5 miRNAs as biomarkers

(1) Lung nodule benign and malignant risk scoring model

And (3) constructing a benign and malignant risk scoring model of the pulmonary nodules by using the miRNA expression data of the training queue and 5 biomarkers as variables and combining pathological detection results. The model consists of three parts of parameters, a model formula and a reference value. The process is as follows:

① model parameters in the training cohort, 5 biomarkers were used as variables and the model coefficients for the biomarkers were obtained using 100 replicates of the 10-fold cross-validation method (table 2).

TABLE 2 parameters of risk scoring model constructed with 5 miRNAs as biomarkers

② risk score model the risk score model formula is as follows:

wherein Risk-Score is a lung nodule benign and malignant Risk prediction value, Gi represents an expression value of an ith miRNA, β i represents a Risk scoring model parameter of the ith miRNA, i = 1, 2 … n, n is the total number of lung nodule benign and malignant biomarkers, α represents a model correction value.

③ reference value, when the risk value is less than or equal to the reference value, the sample is predicted to be benign pulmonary nodule, otherwise, the sample is predicted to be malignant pulmonary nodule, according to the risk value and the pathological detection result of each patient in the training queue, a receiver operating characteristic curve (ROC curve) of the training queue is drawn, in clinical application, the high specificity risk assessment model can be used for assisting in reducing the false positive rate of LDCT detection, reducing the operation open rate of a clinician, and improving the specificity of early lung cancer diagnosis, therefore, the model emphasizes the high specificity of the model, the reference value is calculated according to the condition that the specificity value in the ROC curve is more than 0.9 and the sensitivity value is more than 0.5, in the example, the reference value is 0.79.

④ model efficiency evaluation, based on the reference Value of 0.79, the training queue samples are divided into a low risk group (i.e. predicted benign nodules) and a high risk group (i.e. predicted malignant nodules), the evaluation model predicts efficiency by taking the pathological detection result as the true Value, the evaluation method of model prediction efficiency comprises the Area Under the ROC Curve (AUC, Area Under the Curve, the Value range of 0-1), the Positive Predictive Value (PPV, Positive Predictive Value, the Value range of 0-1), the specificity (Value range of 0-1) and the sensitivity (Value range of 0-1), the higher the Value is, the better the effect is, the results are respectively 0.920 (figure 3), 93.8%, 94.1% and 50.0% (table 3), the results show that in the training queue, the risk prediction model has higher AUC, PPV and specificity, and the model has better prediction efficiency on the benign nodules.

(2) Verification of prediction efficiency of risk scoring model

And in the verification queue, verifying the efficiency of predicting the benign and malignant pulmonary nodules of the model according to the risk scoring model and the reference value determined in the training queue. The process is as follows:

① risk value in the validation queue, a risk value is calculated for each sample.

② model efficacy validation, dividing the validation cohort of patients into a low risk group and a high risk group (the same training cohort) based on the reference value of 0.79. plotting a ROC curve (fig. 4) with the pathological examination results as the true value, evaluating the model predictive efficacy including AUC, PPV, specificity and sensitivity of 0.763, 91.3%, 91.7% and 55.3%, respectively (table 3). the results show that in the validation cohort, the risk predictive model also has a higher AUC, PPV and specificity, i.e. the model predictive efficacy is validated in the validation cohort.

TABLE 3 evaluation of potency of five molecular marker models

Example 2: development of risk scoring model research by taking 3 miRNAs as biomarkers

The risk score model development research shows that when the biomarker takes 5 miRNAs (hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p and hsa-miR-3168) as the core, the number of the miRNAs is reduced or increased, and higher prediction efficiency can be obtained.

By using 3 miRNAs as markers and hsa-let-7b-3p, hsa-miR-101-3p and hsa-miR-3168 as biomarkers, a risk scoring model (shown in the table 4) is constructed in the same way, and by using 0.77 as a reference value, higher AUC (0.914) and specificity (94.1%) can be obtained (shown in figure 5).

TABLE 4 parameters of risk scoring model constructed with 3 miRNAs as markers

Example 3 expansion of Risk Scoring model Studies-8 miRNAs as markers

Based on 5 core biomarkers, hsa-let-7d-3p, hsa-miR-19b-3p and hsa-miR-503-5p are newly added to be 8-variable biomarkers, a risk score model (Table 5) is constructed, and a reference value of 0.79 is used, so that higher AUC (0.920) and specificity (94.1%) can be obtained (FIG. 6).

TABLE 5 parameters of risk scoring model constructed with 8 miRNAs as markers

Example 4 expansion of Risk scoring model Studies-10 miRNAs as markers

Based on 5 core biomarkers, hsa-miR-16-5p, hsa-miR-17-5p, hsa-miR-19b-3p, hsa-miR-532-5p and hsa-miR-98-5p are newly added as 10-variable biomarkers, a risk score model (Table 6) is constructed, and higher AUC (0.961) and specificity (94.1%) can be obtained by taking 0.68 as a reference value (figure 7).

TABLE 6 parameters of risk scoring model constructed with 10 miRNAs as markers

According to the invention, a model can be established by using fewer miRNA locus combinations for risk evaluation, and more miRNA locus combinations can be used for risk evaluation; when the number of miRNA sites used is increased, benign and malignant lung nodules can be detected and distinguished from each other from more dimensions, and therefore a better detection result is obtained. Based on the miRNA locus or locus combination of the present invention, one skilled in the art can establish other risk assessment models and select corresponding parameters, and is not limited to the risk scoring models and parameter selections used in the embodiments of the present invention.

Claims (9)

1. A peripheral blood extracellular vesicle miRNA biomarker for early diagnosis of lung cancer, wherein the peripheral blood extracellular vesicle miRNA biomarker comprises hsa-let-7b-3p, hsa-miR-101-3p and hsa-miR-3168.

2. The miRNA biomarker of claim 1, further comprising one or more miRNAs selected from the group consisting of hsa-miR-125b-5p, hsa-miR-150-5p, hsa-let-7d-3p, hsa-miR-503-5p, hsa-miR-16-5p, hsa-miR-17-5p, hsa-miR-19b-3p, hsa-miR-532-5p, and hsa-miR-98-5 p.

3. The miRNA biomarker of claim 2, consisting of hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p, hsa-miR-3168; or consists of hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p, hsa-miR-3168, hsa-let-7d-3p, hsa-miR-19b-3p and hsa-miR-503-5 p; or consists of hsa-let-7b-3p, hsa-miR-125b-5p, hsa-miR-150-5p, hsa-miR-101-3p, hsa-miR-3168, hsa-miR-16-5p, hsa-miR-17-5p, hsa-miR-19b-3p, hsa-miR-532-5p and hsa-miR-98-5 p.

4. The miRNA biomarker of any one of claims 1 to 3, wherein the expression value of the peripheral blood extracellular vesicle miRNA marker in a peripheral blood extracellular vesicle of a lung cancer patient is statistically significantly different from the expression value in a control sample.

5. The miRNA biomarker of claim 4, wherein the control sample is a benign lung nodule patient.

6. A kit for early diagnosis of lung cancer, comprising: the kit comprises a detection reagent of the miRNA marker of the extracellular vesicle of the peripheral blood of any one of claims 1-3.

7. Use of a peripheral blood extracellular vesicle miRNA marker according to any one of claims 1-3 for the preparation of a diagnostic agent for predicting early stage lung cancer by a method comprising:

(1) aiming at a subject with lung nodules detected by LDCT in clinic, collecting a peripheral blood sample, and detecting the existence of miRNA in the extracellular vesicles of peripheral blood;

(2) measuring the expression level of a miRNA biomarker as in claim 1 or 2 in the peripheral blood extracellular vesicle sample;

(3) a risk scoring model based on previously measured expression levels of mirnas is used to predict the likelihood of LDCT detecting whether a subject with existing lung nodules carries malignant nodules.

8. The use according to claim 7, wherein the expression level of the miRNA is quantified based on small RNAseq detection results.

9. The use of claim 8, wherein the risk scoring model comprises a mathematical model formulated as:

wherein Risk-Score is a lung nodule benign and malignant Risk prediction value, Gi represents an expression value of an ith miRNA, β i represents a Risk Score parameter value of the ith miRNA, i = 1, 2 … n, n is the total number of lung nodule benign and malignant biomarkers, and α represents a model correction parameter.

Images