Evaluation system for aided prediction of lung cancer metastasis and application

Abstract

According to the invention, a group of ctDNA fragmentation characteristics capable of predicting lung cancer remote site metastasis are screened out, and on the basis, a prediction model for evaluating the risk of lung cancer remote site metastasis is successfully constructed; according to the non-invasive liquid biopsy technology based on the ctDNA fragmentation characteristics, on the premise of non-invasive detection, whether a patient has the risk of merging the metastasis of the remote site or not can be evaluated in a short time, and the next treatment scheme can be made in an auxiliary mode; according to the method, an analysis strategy based on fragment omics is adopted, the required sequencing depth is low, and the related detection cost is remarkably reduced.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of biological medicine, and particularly relates to a lung cancer distant site metastasis evaluation system. BACKGROUND

[0002] Lung cancer is the cancer type with the highest incidence and mortality rate in the world at present, and is one of the important factors threatening people's health. It is mainly divided into two pathological types of non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). Among them, non-small cell lung cancer accounts for about 80%-85% of the total lung cancer incidence, is the main pathological subtype of lung cancer, and mainly includes squamous cell carcinoma (squamous carcinoma), adenocarcinoma, large cell carcinoma and other pathological subtypes. Small cell lung cancer accounts for about 15%-20% of the total lung cancer, grows faster, and has very strong invasion and metastasis ability, so that when it is found clinically, it is often combined with regional or distant site metastasis, and surgical treatment often cannot benefit such patients. However, for small cell lung cancer patients without distant site metastasis, surgical treatment also has good effect, and can even achieve a radical effect. Therefore, for both small cell lung cancer and non-small cell lung cancer patients, it is of great clinical significance to determine the tumor progression and whether it is combined with distant site metastasis for the selection of subsequent treatment plan and the overall prognosis of patients.

[0003] At present, the evaluation of potential metastasis of lung cancer mainly relies on plain CT, PET-CT or bone scan and other examination methods. However, due to the technical limitations of imaging detection, although the plain CT has lower cost and smaller radiation, it depends on experienced interpreters and mainly scans the key areas, so it is difficult to evaluate the whole body. In addition, the resolution of plain CT for metastasis smaller than 5mm is limited. Bone scan examination is based on the principle of radioisotope labeling, which can scan the whole body for potential bone metastasis. However, it can only screen the parts with active bone metabolism, and the effect of distinguishing benign bone diseases, bone degenerative changes and local inflammation is limited. At the same time, it can only distinguish metastasis of 5-10mm, and lacks enough resolution for smaller metastasis. PET-CT uses FDG-18 labeled glucose to scan the whole body for active metabolism sites, and has high screening efficiency. However, this method has high cost and is a radiative examination, which has great influence on patients and the surrounding environment. In addition, the detection resolution of this method is also between 5-10mm, which makes it difficult to distinguish smaller potential metastasis, resulting in rapid recurrence of some patients with distant site micro-metastasis after surgery, leading to poor overall survival. In addition to the above imaging methods, the current clinical method is to obtain pathological tissue through invasive puncture for further diagnosis, but the sensitivity of the related operation is low, and it is limited by the position of the lesion, the relationship with the surrounding tissue and other factors, so it has great limitations. There is an urgent need for an auxiliary detection method which can evaluate the risk of distant site metastasis, is minimally invasive, efficient and low-cost.

[0004] Cell-free DNA (cfDNA) is the cell-free and partially degraded DNA in the body fluids, especially in the blood circulation, which is released into the blood circulation after the shedding or apoptosis of the body tissue cells. The cfDNA derived from tumor cells is the circulating tumor DNA (ctDNA), which carries the genetic characteristics of tumor cells and can be detected in peripheral blood (Liquid biopsies come of age: towards implementation of circulating tumour DNA. Nat Rev Cancer, 2017, 17(4): 223~238.). During the development and metastasis of tumors, there are mainly two categories of molecular mechanisms, including sequence mutations and epigenetic changes. Epigenetic changes are important factors driving tumor metastasis (Protein arginine methyltransferase 5 promotes lung cancer metastasis via the epigenetic regulation of miR-99 family / FGFR3 signaling[J]. Cancer Lett, 2018, 427: 38~48.). It is generally recognized in the academic field that epigenetic changes often occur earlier than sequence mutations, and ctDNA carrying related epigenetic changes can be detected earlier. The related detection methods show superior clinical application performance in tumor screening (Mouliere, Toward the Early Detection of Cancer by Decoding the Epigenetic and Environmental Fingerprints of Cell-Free DNA. Cancer cell, 2019. 36(4): p. 350-368.). Epigenetic changes of tumors are not only one of the important factors driving tumor metastasis, but also the important molecular basis for tumor cells to adapt to the local microenvironment after colonization at the metastatic site. Many studies have confirmed that there are significant differences in epigenetic modifications between the primary tumor and metastatic tumor (Viability of Lymph Node Samples Obtained by Echobronchoscopy in the Study of Epigenetic Alterations in Patients With Lung Cancer. Archivos de Bronconeumología (English Edition), 2014. 50(6): p. 213-220.).

[0005] Methods based on ctDNA epigenetic detection have been widely used in clinical diagnosis. Related studies have shown that this method has good clinical detection efficiency, and the sensitivity and specificity can meet the clinical detection requirements (Epigenetics, fragmentomics, and topology of cell-free DNA in liquid biopsies[J]. Science, 2021, 372(6538).). But the current clinical detection application scenarios based on this biological principle mainly focus on tumor screening, benign and malignant lung nodule differentiation, etc. Only a small number of literatures analyze the differences in ctDNA level between patients with and without distant metastasis. Such studies mainly focus on mutation group analysis and do not conduct in-depth analysis of the epigenetic group of patients. Also, no corresponding detection system has been constructed. SUMMARY

[0006] In this study, the ctDNA fragmentation characteristics of lung cancer patients with and without distant metastasis were analyzed, and a group of characteristic ctDNA fragmentation characteristics of lung cancer distant metastasis was screened. On this basis, a prediction model for evaluating the risk of lung cancer distant metastasis was constructed. Based on this, the present application is completed.

[0007] In a first aspect, the present application provides a biomarker for predicting distant metastasis of lung cancer patients, which comprises ctDNA fragments with the following fragment characteristics: chr7_11000001_11500000, chrll_55000001_60000000, chrll_45000001_50000000, chrll_40000001_45000000, chrll_35000001_40000000, chrll_30000001_35000000, chrll_25000001_30000000, chrll_20000001_25000000, chrll_15000001_20000000, chrll_10000001_15000000, chrll_50000001_10000000, chr10_130000001_135000000, chr10_125000001_130000000, chr10_120000001_12500000C, chr10_115000001_120000000, chr10_110000001_115000000, chr10_105000001_110000000, chr10_950000001_100000000, chr10_90000001_95000000, and chr10_85000001_90000000; and ctDNA fragments with ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC, or GGCG end structure; wherein, when the sum of the total segment characteristic entropy value of the (1) part ctDNA fragments and the total end characteristic entropy value of the (2) part ctDNA fragments is greater than 0.5155907, it is suggested that the lung cancer patient has a risk of distant metastasis.

[0008] Further, the total segment characteristic entropy value is obtained by adding the segment characteristic value of each ctDNA fragment in the (1) part, wherein the segment characteristic value of each ctDNA fragment is the quotient value obtained by dividing the number of short fragments in the fragment by the number of long fragments.

[0009] Further, the long fragment is a ctDNA fragment of 151-220 bp, and the short fragment is a ctDNA fragment of 100-150 bp.

[0010] Furthermore, the total terminal entropy value refers to the total terminal entropy value of the ctDNA fragments in part (2) obtained by dividing the number of ctDNA fragments with ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC or GGCG terminal structures in part (2) by the total number of ctDNA fragments.

[0011] Furthermore, the terminology refers to the four bases at the end of the ctDNA fragment, specifically the two bases at the 5' end and the two bases at the 3' end.

[0012] Furthermore, the lung cancers include small cell lung cancer and non-small cell lung cancer.

[0013] Furthermore, the non-small cell lung cancer includes adenocarcinoma, squamous cell carcinoma, and other pathological subtypes of lung cancer.

[0014] Furthermore, the remote site transfer includes unmerged remote site transfer and / or merged remote site transfer.

[0015] Furthermore, the absence of distant metastasis refers to the tumor being confined to the primary lesion in the lung and its surrounding area, without spreading to distant organs or tissues outside the lung via the blood or lymphatic system; the presence of distant metastasis refers to the tumor cells having broken through local limitations and spreading to distant organs or tissues outside the lung via the blood or lymphatic system.

[0016] Secondly, the present invention provides a liquid biopsy system for predicting distant metastasis in lung cancer patients, the system comprising a data input module, a data processing module, and a data output module; wherein... The data input module is for inputting ctDNA sequencing data from lung cancer patients; The data processing module performs calculations on the ctDNA sequencing data input from the data input module, specifically including the following steps: S01, Obtain the (1)th part of the ctDNA fragment from the ctDNA sequencing data. The (1)th part of the ctDNA fragment includes: chr7_11000001_11500000, chr11_55000001_60000000, chr11_45000001_50000000, chr11_40000001_45000000, chr11_35000001_40000000, chr11_30000001_35000000, chr11_25000001_30000000, chr11_20000001_ 25000000,chr11_15000001_20000000,chr11_10000001_15000000,chr11_50000001_10000000,chr10_130000001_ 135000000,chr10_125000001_130000000,chr10_120000001_12500000C,chr10_115000001_120000000,chr10_110000001 _115000000,chr10_ 105000001 _ 110000000, chr10 _ 950000001 _ 100000000, chr10 _ 90000001 _ 95000000 and chr10 _ 85000001 _ 90000000; S02, Obtain the (2nd) part of ctDNA fragment from ctDNA sequencing data, wherein the (2nd) part of ctDNA fragment includes ctDNA fragments with terminal structures of ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC or GGCG; S03, calculate the total segment feature entropy value of the ctDNA fragment in part (1); S04, calculate the total terminal entropy value of the ctDNA fragment in part (2); S05, sum the total segment feature entropy value of the (1) part of the ctDNA fragment obtained in S03) with the total terminal feature entropy value of the (2) part of the ctDNA fragment to obtain the ctDNA entropy value of the lung cancer patient; The data output module compares the ctDNA entropy value obtained in S05 of the data processing module with 0.5155907. When the ctDNA entropy value is greater than 0.5155907, the system outputs that the risk of distant metastasis of cancer cells is high.

[0017] Furthermore, the ctDNA sequencing data of the lung cancer patients was obtained by collecting blood from lung cancer patients, extracting ctDNA, and sequencing it. The blood from the lung cancer patients is preferably peripheral blood.

[0018] Furthermore, in S03, the total segment feature entropy value is obtained by adding the segment feature values ​​of each ctDNA fragment in part (1), wherein the segment feature value of each ctDNA fragment is the quotient obtained by dividing the number of short fragments in that fragment by the number of long fragments.

[0019] Furthermore, the long fragment is a ctDNA fragment of 151-220 bp, and the short fragment is a ctDNA fragment of 100-150 bp.

[0020] Furthermore, the total terminal entropy value refers to the total terminal entropy value of the ctDNA fragments in part (2) obtained by dividing the number of ctDNA fragments with ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC or GGCG terminal structures in part (2) by the total number of ctDNA fragments.

[0021] Furthermore, the terminology refers to the two bases at the 5' end and the two bases at the 3' end of the ctDNA fragment sequence, which are the four bases at the end of the ctDNA fragment.

[0022] Thirdly, the present invention provides an application of a reagent for detecting the expression level of the biomarker described in the first aspect in the preparation of a product for predicting distant metastasis in lung cancer patients.

[0023] Furthermore, the lung cancers include small cell lung cancer and non-small cell lung cancer.

[0024] Furthermore, the non-small cell lung cancer includes adenocarcinoma, squamous cell carcinoma, and other pathological subtypes of lung cancer.

[0025] Furthermore, the remote site transfer includes unmerged remote site transfer and / or merged remote site transfer.

[0026] Furthermore, the absence of distant metastasis refers to the tumor being confined to the primary lesion in the lung and its surrounding area, without spreading to distant organs outside the lung via the blood or lymphatic system; the presence of distant metastasis refers to the tumor cells having broken through local limitations and spreading to distant organs or tissues outside the lung via the blood or lymphatic system.

[0027] Furthermore, the product includes reagents or kits.

[0028] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, the program being executed by a processor according to the liquid biopsy system described in the second aspect.

[0029] Beneficial effects This invention screened a set of characteristic ctDNA fragmentation features of lung cancer distant metastasis. Based on these features, a predictive model for assessing the risk of distant metastasis in lung cancer patients was constructed. Using the non-invasive liquid biopsy technology based on ctDNA fragmentation features provided by this invention, the risk of distant metastasis in patients can be assessed in a short time without invasive testing, assisting in the formulation of the next treatment plan. This invention adopts a fragmentomics-based analysis strategy, which requires low sequencing depth and significantly reduces related testing costs. Attached Figure Description

[0030] Figure 1 This is a schematic diagram illustrating the technology used in a lung cancer distant metastasis prediction system to assess the risk of distant metastasis in lung cancer patients based on cfDNA fragmentomics.

[0031] Note: A. Due to apoptosis, necrosis, and active release pathways, tumor cells in metastatic lesions release their nuclear chromatin into the bloodstream. Due to blood shearing forces and enzymatic degradation, the corresponding genome is fragmented into cfDNA fragments, mainly in the form of nucleosome monomers / oligomers, which participate in circulation in the plasma; B. Free DNA fragments are extracted from the plasma and subjected to fragmentomics analysis to extract fragmentomics characteristics of the test samples; C. The fragmentomics characteristics extracted in the above steps are input into the lung cancer distant metastasis prediction system for calculation, and the prediction results are output.

[0032] Figure 2 The difference in ctDNA fragmentation levels between N2M0 and N2M1 patients.

[0033] Figure 3 This describes the segmental characteristics related to distant metastasis of lung cancer.

[0034] Figure 4 This is a characteristic terminal pattern associated with distant sites of lung cancer.

[0035] Figure 5 This study analyzed ctDNA fragments in traditional lung cancer screening systems for patients with and without distant metastases.

[0036] Figure 6 This study analyzed patients with and without distant metastasis using a ctDNA fragment group risk assessment model.

[0037] Figure 7The receiver operating characteristic (ROC) curve is for a lung cancer distant site metastasis assessment system based on plasma ctDNA detection and independent test cohorts.

[0038] Figure 8 The difference in metastasis index represents the degree of invasion, where AIS represents adenocarcinoma in situ; MIA represents minimally invasive adenocarcinoma; and IA represents invasive adenocarcinoma. Detailed Implementation

[0039] The specific embodiments of the present invention will be further described below. It should be noted that these descriptions are for the purpose of aiding understanding the present invention, but do not constitute a limitation thereof. Furthermore, the technical features involved in the embodiments described below can be combined with each other as long as they do not conflict with each other.

[0040] Unless otherwise specified, the experimental methods used in the following embodiments are conventional methods, and the experimental materials used in the following embodiments are all available through conventional commercial channels.

[0041] Example 1: Screening for biomarkers associated with distant metastasis of lung cancer Case enrollment information and data collection A retrospective clinical cohort was constructed. All lung cancer patients included in this invention underwent clinical evaluation and obtained definitive pathological diagnoses, staging, and metastasis status. Based on pathology reports, the enrolled patients included 32 small cell lung cancer patients and 294 non-small cell lung cancer patients. The non-small cell lung cancer patients included 199 adenocarcinoma patients, 85 squamous cell carcinoma patients, and 10 patients with other pathological subtypes of lung cancer. All lung cancer patients with confirmed diagnoses were randomly assigned to the training set and the validation set, with 246 patients in the training set and 80 patients in the validation set.

[0042] Peripheral blood samples from 246 lung cancer patients in the training set were collected as raw samples and placed in ctDNA preservation tubes for ctDNA extraction. 1) Sample pretreatment: EDTA-anticoagulated whole blood was centrifuged, and the supernatant plasma was aspirated and centrifuged again. The clarified plasma was transferred to EP tubes; 2) Lysis reaction: Lysis buffer and proteinase K were added to the EP tubes containing clarified plasma, and the mixture was incubated in a water bath; 3) DNA binding: Binding buffer was added, and the mixture was centrifuged; 4) Washing and purification; 5) Elution and collection yielded ctDNA samples. Lung cancer patients with substandard DNA quality were removed from the extracted ctDNA samples, resulting in ctDNA samples from lung cancer patients with satisfactory DNA quality.

[0043] WGS libraries were constructed using the RainbowOne Universal DNA Library Prep Kit for MGI. The basic principles of WGS library preparation were followed, including molecular end repair, sequencing aptamer ligation, and library purification. The libraries were then amplified and purified. Finally, 24 libraries were sequenced using DNBSEQ™ technology and sequencing mode on a single channel of an MGI-2000 sequencer.

[0044] Based on the XGboost machine learning algorithm as the underlying architecture for screening and training, ctDNA fragment omics feature analysis is performed by inputting ctDNA sequencing data from lung cancer patients.

[0045] Raw sequencing data was filtered using FASTP as part of quality control. After PCR repeats were removed using Samtools, qualified reads were aligned to the reference genome hg19 using the BWA sequence alignment tool with default parameters.

[0046] Following alignment, the data underwent cell-free DNA fragmentation mapping analysis. Specifically, the genome was divided into 5 Mb segments. Based on segment length, ctDNA fragments were further categorized into short and long fragments (151-220 bp for long fragments and 100-150 bp for short fragments). Dividing the short fragment by the long fragment yielded the segmental entropy value for each region of the genome. After whole-genome analysis, 472 ctDNA segment features were obtained. Using the 5' and 3' ends of each ctDNA fragment as motif codes, 256 features were identified and used as ctDNA fragment terminal features. Combined analysis of ctDNA segment and terminal features yielded a total of 738 fragmentomics features.

[0047] Considering the interference of local lymph node metastasis, this invention performs grouping analysis on the ctDNA fragment omics status of non-small cell lung cancer patients with local metastasis (especially those with only regional lymph node metastasis) and those with distant metastasis. The ctDNA fragment omics indicators of TxN2M0 (clinical stage III, lung cancer patients with regional metastasis) and TxNxM1 (clinical stage IV, patients with distant metastasis in one or more organs) candidates were analyzed. The results showed that, genomically, both M0 patients with local metastasis but no distant metastasis and M1 patients with distant metastasis exhibited significant fluctuations and high disorder (see...). Figure 2 Among these, in certain characteristic segments, M1 patients with distant metastases exhibited higher fragmentation entropy values ​​(see [link to relevant documentation]). Figure 3 In addition, the above differences also exist in the terminal patterns of some ctDNA (see...). Figure 4 ).

[0048] Based on the above results, this invention utilizes the XGBoost machine learning algorithm and, according to the DELFI method (Genome-widecell-free DNA fragmentation in patients with cancer[J]. NATURE 2019;570: 385-9.x), compares and analyzes the ctDNA segment features and terminal features of 246 lung cancer patients in the training set, including those without distant metastases and those with distant metastases. This analysis identifies core features related to distant metastases in lung cancer, including distant metastasis-related fragment features (chr7_11000001_11500000, chr11_55000001_60000000, chr11_45000001_50000000, chr11_40000001_45000000, chr11_35000001_ 40000000,chr11_30000001_35000000,chr11_25000001_30000000,chr11_20000001_25000000,chr11_15000001_ 20000000, chr11 _10000001 _ 15000000, chr11 _ 50000001 _ 10000000, chr10 _ 130000001 _135000000, chr10 _ 125000001 _ 130000000,chr10_120000001_ 12500000C, chr10 _115000001 _ 120000000, chr10 _ 110000001 _ 115000000, chr10 _ 105000001 _110000000, chr10 _ 950000001 _ 100000000, chr10 _ 90000001 _ 95000000 and chr10 _85000001 _ 90000000) and characteristic terminal patterns (ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC and GGCG).

[0049] Example 2: Construction of a lung cancer distant metastasis prediction system Based on the remote site transfer-related segment features obtained in Example 1 (chr7_11000001_11500000, chr11_55000001_60000000, chr11_45000001_50000000, chr11_40000001_45000000, chr11_35000001_40000000, chr11_30000001_35000000, chr11_25000001_30000000, chr11_20000001_25000000, chr11_15000001), _20000000, chr11 _ 10000001 _ 15000000, chr11 _ 50000001 _ 10000000, chr10 _130000001 _ 135000000, chr10 _ 125000001 _ 130000000,chr10_120000001_12500000C,chr10_115000001_120000000,chr10_110000001_115000000,chr10_105000001_ 110000000, chr10 _ A lung cancer distant metastasis prediction system was constructed using 950000001 _ 100000000, chr10 _ 90000001 _ 95000000 and chr10 _ 85000001 _ 90000000) and characteristic terminal patterns (ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC and GGCG).

[0050] Peripheral blood samples from 80 patients in the validation set were collected as raw samples and placed in ctDNA storage tubes for ctDNA extraction. 1) Sample pretreatment: EDTA-anticoagulated whole blood was centrifuged, and the supernatant plasma was aspirated and centrifuged again. The clarified plasma was transferred to EP tubes. 2) Lysis reaction: Lysis buffer and proteinase K were added to the EP tubes containing clarified plasma, and the mixture was incubated in a water bath. 3) DNA binding: Binding buffer was added, and the mixture was centrifuged. 4) Washing and purification. 5) Elution and collection yielded ctDNA samples. Lung cancer patients with substandard DNA quality were excluded from the extracted ctDNA samples, resulting in 83 lung cancer patients with acceptable DNA quality. The ctDNA samples were sequenced to obtain ctDNA sequencing data.

[0051] The total segment feature entropy value of the ctDNA fragment in part (1) is calculated by summing the quotients obtained by dividing the number of short fragments by the number of long fragments in each ctDNA fragment in part (1). Among them, long fragments are fragments with a size of 151-220 bp, and short fragments are fragments with a size of 100-150 bp.

[0052] The total terminal entropy value of the ctDNA fragment in part (2) is calculated by dividing the number of ctDNA fragments in part (2) with ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC, or GGCG terminal structures by the total number of ctDNA fragments. The terminal structure refers to the motif composed of the two bases at the 5' end and the two bases at the 3' end of the ctDNA fragment.

[0053] The total segment feature entropy value of the obtained (1) part ctDNA fragment and the total terminal feature entropy value of the obtained (2) part ctDNA fragment were summed to obtain the ctDNA entropy value (score) for each lung cancer patient, resulting in a total of 80 sets of ctDNA entropy values. At the same time, the sensitivity, specificity and Youden index corresponding to each set were obtained, and the receiver operating characteristic curve (ROC curve) was plotted. The sensitivity and "1-specificity" values ​​corresponding to each ctDNA fragment entropy value node were output, and the corresponding Youden index was calculated, i.e., sensitivity + specificity - 1. When the Youden index is the largest, the sensitivity and specificity are relatively balanced, and XGBoost machine learning is more stable in predicting the risk of lung cancer patients with distant metastasis, and its diagnostic effect is the best. The corresponding feature entropy value at this time is 0.5155907. Furthermore, 0.5155907 is used as the cut-off value for judging the risk of lung cancer patients with distant metastasis. When a patient's ctDNA entropy value is greater than 0.5155907, the patient is diagnosed with a high risk of distant metastasis of cancer cells.

[0054] A tumor classification model trained using patent CN116452559B was used to analyze patients with and without distant metastases. In the N0 and N1-3 stratifications, traditional models for lung cancer screening failed to successfully distinguish between M0 and M1 stage patients (see...). Figure 5 Based on this, the tumor classification model was optimized using the method described in patent CN116452559B, and a liquid biopsy system for predicting distant metastases of lung cancer was successfully constructed. This system can successfully distinguish between M0 and M1 stage patients (see...). Figure 6 ).

[0055] Example 3: Performance Validation of a Lung Cancer Distant Metastasis Prediction System An external validation cohort, independent of the samples selected in Example 1, was established, including 46 patients: 7 with small cell lung cancer, 25 with adenocarcinoma, and 14 with squamous cell carcinoma. The external validation cohort was used to evaluate the predictive performance of the aforementioned liquid biopsy system for distant metastases of lung cancer.

[0056] The verification results are as follows Figure 7 As shown, the AUC of the external validation cohort was 0.959, the sensitivity was 0.960, and the specificity was 0.909. The positive predictive value (PPV) and negative predictive value (NPV) were 0.827 and 0.980, respectively. Although the specificity has certain limitations, imaging techniques such as CT and PET-CT can compensate for this deficiency. Conversely, the high specificity and negative predictive value suggest that when this system assesses a low risk of distant metastasis, surgical treatment can be performed directly, and subsequent adjuvant therapy targeting potential metastases may not be necessary if postoperative pathological lymph node metastasis is negative.

[0057] Example 4: Relationship between the assessment index of the lung cancer distant metastasis prediction system and the degree of lung cancer invasion. A retrospective clinical cohort of adenocarcinoma subgroups was constructed, including 152 adenocarcinoma patients. The distant metastasis prediction assessment system for lung cancer was used to evaluate the distant metastasis index in these 152 patients across different invasive and differentiated adenocarcinoma subgroups. The results showed that the distant metastasis prediction index of the lung cancer prediction system increased with the degree of lung cancer invasion (see...). Figure 8 The number of metastasis indices is increasing and showing an upward trend. A high metastasis index indicates a higher risk of distant metastasis, which is consistent with established academic understanding and supported by numerous related studies.

Claims

1. A biomarker for predicting distant metastasis of a lung cancer patient, the biomarker comprising ctDNA fragments with the following fragment characteristics: chr7_11000001_11500000, chr11_55000001_60000000, chr11_45000001_50000000, chr11_40000001_45000000, chr11_35000001_40000000, chr11_30000001_35000000, chr11_25000001_30000000, chr11_20000001_25000000, chr11_15000001_20000000, chr11_10000001_15000000, chr11_50000001_10000000, chr10_130000001_135000000, chr10_125000001_130000000, chr10_120000001_12500000C, chr10_115000001_120000000, chr10_110000001_115000000, chr10_105000001_110000000, chr10_950000001_100000000, chr10_90000001_95000000, and chr10_85000001_90000000; and ctDNA fragments with ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC, or GGCG end structure. When the sum of the total segment characteristic entropy value of the ctDNA fragments in part (1) and the total end characteristic entropy value of the ctDNA fragments in part (2) is greater than 0.5155907, it indicates that the lung cancer patient has a risk of distant metastasis. The segment characteristic value of each ctDNA fragment is the quotient obtained by dividing the number of short fragments in the fragment by the number of long fragments; and the total end characteristic entropy value refers to the total end characteristic entropy value of the ctDNA fragments in part (2) obtained by dividing the number of ctDNA fragments with ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC, or GGCG end structure in part (2) by the total number of ctDNA fragments. wherein ​ 2. The biomarker of claim 1, wherein the total segment feature entropy value is obtained by adding the segment feature values of each of the ctDNA fragments of part (1), wherein, ​ 3. The biomarker of claim 1, wherein the long fragment is a 151-220 bp ctDNA fragment, the short fragment is a 100-150 bp ctDNA fragment, the terminal structure refers to four bases at the end of the ctDNA fragment, and the four bases are two bases at the 5' end and two bases at the 3' end.

4. The biomarker of claim 1, wherein the lung cancer comprises small cell lung cancer and non-small cell lung cancer.

5. The biomarker of claim 1, wherein the distant metastasis comprises non- combined distant metastasis and / or combined distant metastasis.

6. A liquid biopsy system for predicting distant metastasis of a lung cancer patient, the system comprising a data input module, a data processing module, and a data output module, wherein the data input module is configured to input ctDNA sequencing data of the lung cancer patient; the data processing module is configured to calculate the ctDNA sequencing data input by the data input module, and specifically comprises the following steps: S01, obtaining a first part of ctDNA fragments from the ctDNA sequencing data, wherein the first part of ctDNA fragments comprises chr7_11000001_11500000, chr11_55000001_60000000, chr11_45000001_50000000, chr11_40000001_45000000, chr11_35000001_40000000, chr11_30000001_35000000, chr11_25000001_30000000, chr11_20000001_25000000, chr11_15000001_20000000, chr11_10000001_15000000, chr11_50000001_10000000, chr10_130000001_135000000, chr10_125000001_130000000, chr10_120000001_12500000C, chr10_115000001_120000000, chr10_110000001_115000000, chr10_105000001_110000000, chr10_950000001_100000000, chr10_90000001_95000000, and chr10_85000001_90000000. ​ ​ ​ S02, obtaining the second part of ctDNA fragments from the ctDNA sequencing data, wherein the second part of ctDNA fragments comprises ctDNA fragments with end structure of ACCG, CCCG, CCTG, GCCC, GCCG, GCCT, GCTC, GGCC or GGCG; S03, calculating the total segment feature entropy value of the first part of ctDNA fragments; S04, calculating the total end feature entropy value of the second part of ctDNA fragments; S05, summing the total segment feature entropy value of the first part of ctDNA fragments obtained in S03 and the total end feature entropy value of the second part of ctDNA fragments to obtain the ctDNA entropy value of the lung cancer patient; The data output module refers to comparing the ctDNA entropy value obtained in S05 in the data processing module with 0.5155907, and when the ctDNA entropy value is greater than 0.5155907, the system outputs that the risk of cancer cell distant metastasis is high.

7. The liquid biopsy system of claim 6, wherein the ctDNA sequencing data of the lung cancer patient is obtained by collecting blood of the lung cancer patient, extracting ctDNA and sequencing.

8. Use of a reagent for detecting the expression amount of the biomarker of claim 1 in the preparation of a product for predicting distant metastasis of a lung cancer patient.

9. The use of claim 8, wherein the lung cancer comprises small cell lung cancer and non-small cell lung cancer; and the distant metastasis comprises no distant metastasis and / or distant metastasis.

10. A computer readable storage medium having stored thereon a computer program, which is executed by a processor according to the liquid biopsy system of claim 6.

Images