CN116413430B - Autoantibody/antigen combination and detection kit for early prediction of liver cancer - Google Patents

Abstract

The invention provides an autoantibody/antigen combination and a detection kit for early liver cancer prediction, which finally identify a group of autoantibody/antigen molecular combinations which can be used for screening liver cancer, especially early liver cancer patients, and establish an early liver cancer prediction system by detecting autoantibodies and antigen markers aiming at different targets in blood of liver cancer patients and healthy people. The autoantibody/antigen molecule combination in the system is used as a biomarker, has high sensitivity and detection specificity in early liver cancer, is particularly suitable for Chinese people, can be used for prediction of disease risk, screening, prognosis evaluation, treatment effect monitoring or recurrence monitoring of liver cancer, and the like, and meets clinical requirements.

Description

Autoantibody/antigen combination and detection kit for early prediction of liver cancer

Technical Field

The invention relates to the technical field of biology, in particular to an autoantibody/antigen combination and detection kit for early prediction of liver cancer.

Background

Hepatocellular carcinoma (HCC) accounts for 70-85% of total liver cancer cases, is the major cancer burden for liver cancer, and generally develops in the context of advanced chronic liver disease, mainly associated with Hepatitis B Virus (HBV), hepatitis C Virus (HCV) and alcoholism. According to the data of Chinese tumor journal, about 37 ten thousand new cases of liver cancer occur in China, wherein about 27 ten thousand men and 10 ten thousand women; about 32.6 tens of thousands of people die from liver cancer each year, of which about 24.2 tens of thousands are men and 8.4 tens of thousands are women; the incidence rate of liver cancer in rural areas is higher than that in urban areas and in western areas than in eastern areas.

The death rate of liver cancer is as high as 95%, and most patients die within 1 year of diagnosis, which is related to imperfect early detection means of liver cancer. Liver ultrasound, CT and serum Alpha Fetoprotein (AFP) detection are the main methods for finding liver cancer at present, but ultrasound has low sensitivity to small liver cancer (< 2 cm) detection and is greatly influenced by the skill level of operators, and CT detection has the problems of radiation exposure, high cost and the like. AFP is the most widely used serum marker of primary liver cancer at present, the sensitivity is 40% -65%, but the detection rate of early liver cancer is low, and about 30% -40% of AFP of early liver cancer patients is at normal level. In addition, AFP is elevated in some chronic liver diseases and reproductive system tumors, and even if detected in combination with liver-specific AFP heteroplasmic AFP-L3, its efficacy is not improved. Currently, there is no clinically effective molecular marker associated with early detection of liver cancer. Therefore, searching for effective serum markers to improve the ability of early detection of liver cancer and accurate prediction of prognosis of liver cancer patients is always a research hotspot in the liver cancer field.

At present, there are five main modes for early detection of liver cancer: imaging, tumor antigens, gene sequences, pathology detection, tumor autoantibodies. The imaging detection is used as the defect of early detection mode of liver cancer: 1) The false positive rate is high, so that a large number of potential patients can be subjected to re-diagnosis, and the phenomenon of excessive diagnosis is formed; 2) Experienced physicians are required, with a great impact on sensitivity and specificity; 3) The detection cost is high, the time consumption is long, and the method cannot be used for large-scale crowds; 4) There is a risk of radiation exposure; the disadvantage of using tumor antigen alone as early detection mode of liver cancer: 1) The specificity is not high, and can be observed in other malignant tumors, benign hyperplasia and certain inflammatory conditions; 2) Sensitivity to early liver cancer is limited, and some benign diseases of liver, kidney and the like and other cancers can be observed; 3) Antigen levels are affected by tumor burden, and single index sensitivity is relatively low, with limited value in early screening; the deficiency of using gene sequence as early detection mode of liver cancer: 1) Because the early stage of liver cancer has smaller tumor load, the concentration of cfDNA after entering blood is very low, and amplification is usually needed, so that the detection scheme is complicated, the detection time is prolonged, and the standardization degree is insufficient; 2) Among many microRNA molecules, there may be a phenomenon of homology with other RNA sequences, and there are alterations in other cancers, while there is a disadvantage of low abundance in serum, and microRNA detection lacks uniform standards; the pathological detection of liver cancer causes trauma and pain to patients, and is also not suitable for early detection of crowds. The tumor autoantibody is also used as a detection mode for early liver cancer, which has the following defects: 1) Due to tumor heterogeneity, single autoantibodies cannot independently serve as powerful clinical biomarkers; 2) The detection sensitivity of the pure autoantibody is limited, and the pure autoantibody needs to be combined with other types of markers, such as antigens, under the condition of ensuring the specificity, a sensitive detection molecule combination aiming at liver cancer is developed; 3) Detection combinations developed for the chinese population are required, as genetic background of different ethnic groups makes the molecular combinations have different applicability. However, compared with other detection methods, the autoantibody detection is the most effective early liver cancer detection method at present.

The structural changes, abnormal post-translational modifications, abnormal over-or ectopic expression of proteins within tumor cells, etc. may produce tumor-associated antigens (tumour-associated antigens, TAA), which may stimulate the immune system to produce tumor-associated autoimmune antibodies (autoantibodies, AAB). AAB, as a "monitor" of TAA, reflects tumor development, progression and immune status in the body, while AAB is produced clonally for B cells and has a longer half-life, so it is easier to detect as a "reporter" of TAA. Research shows that liver cancer related AAB appears in serum several months or even years before clinical diagnosis, and has potential value for early liver cancer discovery. Meanwhile, the AAB of the target TAA can correspondingly change along with the change of the biological behaviors of tumor cells, so the target TAA has the potential value of monitoring tumor progress and predicting prognosis.

Currently, more than 70 liver cancer related AAB are reported, and related researches are mainly focused on liver cell liver cancer (hepatocellualr cancer, HCC) and Asian population, which are consistent with pathological type distribution and disease area distribution of liver cancer. The most studied are the AAB's against the p53, insulin-like growth factor II messenger RNA binding protein (insulin like growth factor-II mRNAbinding protein, IMP) family and glucose regulatory protein 78 (glucose regulated protein 78kD, GRP 78). In recent years, with the development of high-throughput detection technology for serum proteins, simultaneous screening of a plurality of AABs by using a plurality of TAAs has become the mainstream of research in this field. The combination detection of the plurality of AAB can mutually make up for the defect of low single AAB positive rate, and the liver cancer detection rate is improved to more than 40%.

AAB is closely related to the occurrence and development of liver cancer, is another important serum molecular marker except for circulating tumor cells, circulating tumor cell DNA and serum TAA, has diagnostic and early diagnosis values, generally has slightly reduced specificity compared with single AAB diagnosis, but has greatly improved sensitivity, and simultaneously, the detection rate can be further improved by combining AAB with AFP detection, so that the defect of low sensitivity of AFP to small liver cancer and early liver cancer is overcome.

In the future research in the liver cancer AAB field, new AAB with high diagnostic value and prognostic evaluation value is continuously searched, meanwhile, the clinical transformation of AAB combination is also emphasized, and prospective clinical research of more large samples is developed to define the clinical application value of the AAB. Aiming at high-risk liver cancer groups such as chronic liver disease patients, dynamic monitoring of AAB level is carried out, the contribution degree of AAB in diagnosis and prognosis prediction of liver cancer is evaluated by combining clinical pathological features, and a multi-factor diagnosis model and a curative effect prediction model are constructed, so that the occurrence and progress of liver cancer can be comprehensively and truly reflected, and the method has important significance for early diagnosis and early treatment of liver cancer.

The clinical significance of liver cancer detection is that the earlier the tumor is found, the better the prognosis of the patient. In order to more accurately predict liver cancer in early stage, a more effective liver cancer early stage prediction system based on autoantibody/antigen marker combination is also required to be constructed.

Disclosure of Invention

Aiming at the problems existing in the prior art, the invention provides an autoantibody/antigen combination and a detection kit for early liver cancer prediction, and finally identifies a group of autoantibody/antigen molecule combinations which can be used for screening liver cancer, especially early liver cancer patients, by detecting autoantibodies and antigen markers aiming at different targets in blood of liver cancer patients and healthy people, and establishes an early liver cancer prediction system. The autoantibody/antigen molecule combination in the system is used as a biomarker, has high sensitivity and detection specificity in early liver cancer, and is particularly suitable for Chinese people.

In one aspect, the present invention provides a biomarker panel for predicting whether an individual is liver cancer, the biomarker panel comprising any two or more of the following autoantibodies ：anti-HNRPA1、anti-SCYL3、anti-Trim21、anti-PIM1、anti-CEA、anti-SSX2、anti-CAGE、anti-AIF、anti-IMPDH1、anti-CTAG2.

The invention mainly aims to develop an autoantigen/antibody combination detection method which can sensitively discover early liver cancer by taking blood as a sample. The clinical significance of early detection of liver cancer is that the earlier the tumor is found, the better the prognosis of the tumor patient.

The antigen of the autoantibody is respectively: HNRPA1, SCYL, trim21, PIM1, CEA, SSX2, CAGE, AIF, IMPDH1, CTAG2.

The serial numbers of the 10 autoantibodies are ：HNRPA1：P09651;SCYL3：Q8IZE3;Trim21：P19474;PIM1：P11309;CEA：P06731;SSX2：Q16385;CAGE：Q86TM3;AIF：P55008;IMPDH1：P20839;CTAG2：O75638. in the Uniprot database, and the website of the Uniprot database is www.uniprot.org.

According to the invention, the autoantibodies aiming at the purified antigen proteins in the liver cancer patients are detected, a large amount of public data is synthesized, and the autoantibody contents aiming at different antigen targets in the blood of the liver cancer patients and healthy people are compared to find the autoantibodies capable of predicting the liver cancer. Through preliminary screening, a series of autoantibody biomarkers which can effectively distinguish liver cancer and healthy people are found, and 10 autoantibody markers with higher detection sensitivity and specificity are further obtained through screening.

The sensitivity difference of the individual autoantibody molecules is large, and the heterogeneity of the autoantigens among cancer patients is high, so that the individual autoantibodies are difficult to be used as independent detection basis, and are combined with other biomarkers to develop sensitive detection molecule combinations aiming at liver cancer under the condition of ensuring specificity. In addition, the group of the tested people should be finely divided, and a detection model established for the specific group of people should be developed to reduce adverse effects of various external factors and heterogeneity on early detection of liver cancer. There is therefore a need to develop high quality studies in large independent samples and to try different types of markers to be used in combination. The subjects should contain a large number of early stage patients, with reasonable choice of control population, and attention should be paid to the comparability of results between studies.

According to the invention, through detecting autoantibodies and antigen markers aiming at different targets in the blood of a liver cancer patient, a series of autoantibody/antigen molecule combinations which can be used for screening liver cancer, especially early-stage patients, are finally identified. The biomarker combinations have a sufficiently high sensitivity, in particular in early liver cancer, in particular in experimental chinese populations; while also having a sufficiently high detection specificity.

In some embodiments, the liver cancer comprises hepatocellular carcinoma, non-hepatocellular carcinoma.

In some embodiments, the biomarker combination comprises the following biomarkers: anti-HNRPA1 and Anti-SCYL3.

In some embodiments, the biomarker combination comprises the following biomarkers: one or more of anti-TRIM21, anti-PIM1 and anti-CEA.

In some embodiments, the biomarker combination comprises the following biomarkers: one or more of anti-SSX2, anti-CAGE, anti-CTAG2, anti-AIF and anti-IMPDH 1.

In some embodiments, the biomarker combination comprises the following biomarkers:

(1) Anti-HNRPA1 and Anti-SCYL3; or (b)

(2) Anti-HNRPA1, anti-SCYL, and Anti-TRIM21; or (b)

(3) Anti-HNRPA1, anti-SCYL, anti-TRIM21 and Anti-PIM1; or (b)

(4) Anti-HNRPA1, anti-SCYL, anti-TRIM21, anti-PIM1 and Anti-CEA; or (b)

(5) Anti-HNRPA1, anti-SCYL, anti-TRIM21, anti-PIM1, anti-CEA and Anti-SSX2; or (b)

(6) Anti-HNRPA1, anti-SCYL, anti-TRIM21, anti-PIM1, anti-CEA, anti-SSX2 and Anti-CAGE; or (b)

(7) Anti-HNRPA1, anti-SCYL, anti-TRIM21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE and Anti-CTAG2; or (b)

(8) Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE, anti-CTAG2, and Anti-AIF; or (b)

(9) Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE, anti-CTAG2, anti-AIF and Anti-IMPDH1.

Further, the biomarker combination comprises Anti-HNRPA1, anti-SCYL3, or Anti-HNRPA1, anti-SCYL3, anti-TRIM21, or Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, or Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, anti-CEA, or Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, anti-CEA, anti-SSX2, or Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE, or Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE, anti-CTAG2, or Anti-HNRPA1, anti-SCYL3, anti-TRIM21, anti-PIM1, anti-CEA+anti-SSX2, anti-CAGE, anti-CTAG2, anti-AIF, or Anti-HNRPA1、anti-SCYL3、anti-TRIM21、anti-PIM1、anti-CEA、anti-SSX2、anti-CAGE、anti-CTAG2、anti-AIF+anti-IMPDH1.

Further, the biomarker combination comprises anti-HNRPA1, anti-SCYL, anti-Trim21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE and anti-CTAG2.

The data of the clinical liver cancer detection sample show that when the 8 autoantibody biomarker combinations are only adopted to predict liver cancer, the sensitivity can be up to 60 percent, the specificity can be up to 90 percent, and the diagnosis sensitivity especially for non-liver cell cancer can be up to 72.73 percent, and the specificity can be up to 90 percent. .

Further, the biomarker combination further comprises a tumor antigen marker AFP (alpha fetoprotein).

Based on the autoantibody combination, the invention combines the tumor antigen marker AFP commonly used in clinic with the autoantibody combination, and unexpected discovery can further improve the screening of liver cancer, when a liver cancer detection model of the autoantibody combination and the AFP is established, the sensitivity of liver cancer detection can reach 88.57 percent, and the specificity can reach 97.67 percent. Therefore, the invention also provides application of the autoantibody combination and AFP combined model in liver cancer detection.

In another aspect, the invention provides a kit for predicting whether an individual is liver cancer, the kit comprising detection reagents for the biomarker combinations as described above.

It is another object of the present invention to provide a kit for predicting whether an individual is liver cancer, and a method of using the kit for disease risk prediction, screening, prognosis evaluation, treatment effect monitoring or recurrence monitoring, etc. accordingly.

Based on the autoantibody/antigen combination as a biomarker, the present invention provides a detection reagent, e.g. an antigen protein combination, for detecting the autoantibody/antigen combination; and provides the application of the autoantibody/antigen combination or the detection reagent in preparing products for predicting the risk of liver cancer, screening, prognosis evaluation, treatment effect monitoring or recurrence monitoring and the like.

In some embodiments, the kit is a kit for enzyme-linked immunosorbent assay (ELISA), protein/peptide fragment chip detection, immunoblotting, microbead immunoassay, or microfluidic immunoassay; preferably, the kit is for detecting the biomarker by antigen-antibody reaction, for example an ELISA kit or a fluorescent or chemiluminescent immunoassay kit.

Preferably, the kit is an enzyme-linked immunosorbent assay (ELISA) detection kit. That is, the kit is used to detect whether or not an autoantibody is positive in a sample of a subject by an enzyme-linked immunosorbent assay. Accordingly, the kit may further comprise other components required for ELISA detection of the autoantibody. For detection purposes, the antigenic protein in the kit may be linked to a tag peptide, e.g. His tag, streptavidin tag, myc tag; for another example, the kit may include a solid support, such as a support having microwells to which the antigen protein can be immobilized, such as an elisa plate; or a microbead or magnetic bead solid phase carrier. It may also include an adsorption protein for immobilizing an antigen protein on a solid carrier, a dilution of blood such as serum, a washing solution, a secondary antibody with an enzyme label or a fluorescent or chemiluminescent substance, a color development solution, a stop solution, etc. The concentration of the corresponding antibody in the body fluid is detected by the principle that the antigen protein indirectly or directly coated on the surface of the solid carrier reacts with the antibody in serum/plasma/tissue fluid and the like to form an antigen-antibody complex.

In yet another aspect, the invention provides a kit for predicting whether an individual is non-hepatocellular carcinoma, the kit comprising detection reagents for a biomarker combination as described above.

The study proves that the biomarker combination provided by the invention has higher sensitivity when being used for predicting whether the hepatocellular carcinoma is not caused.

In yet another aspect, the present invention provides a system for predicting whether an individual is liver cancer, the system comprising a data analysis module; the data analysis module is used for analyzing the detection condition of a biomarker, wherein the biomarker is any two or more than two autoantibodies selected from the following autoantibodies ：anti-HNRPA1、anti-SCYL3、anti-Trim21、anti-PIM1、anti-CEA、anti-SSX2、anti-CAGE、anti-AIF、anti-IMPDH1、anti-CTAG2.

Further, the biomarker is a biomarker combination comprising the following autoantibodies: anti-HNRPA1, anti-SCYL, anti-Trim21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE and anti-CTAG2.

In some embodiments, the biomarker combination further comprises a tumor antigen marker AFP.

Further, the analysis method of the data analysis module comprises the following steps: detecting whether a biomarker in a body fluid sample of an individual is positive; and comparing whether the biomarker is positive or not through a cutoff value, wherein the biomarker is positive or not, and the biomarker is negative or not.

In some embodiments, the cutoff value is an OD value determined by enzyme-linked immunosorbent. The cutoff value of the anti-HNRPA1 is 0.7, the cutoff value of the anti-SCYL is 1, the cutoff value of the anti-Trim21 is 0.5, the cutoff value of the anti-PIM1 is 0.3, the cutoff value of the anti-CEA is 0.3, the cutoff value of the anti-SSX2 is 1.2, the cutoff value of the anti-CAGE is 1.6, and the cutoff value of the anti-CTAG2 is 1.

Further, the analysis method of the data analysis module further comprises the following steps: when one or more biomarkers in the biomarker combination are positive, the biomarker combination is positive, and the individual is predicted to be a liver cancer patient; when all the biomarkers in the biomarker combination are negative, the biomarker combination is negative, and the individual is predicted to be a healthy population or a benign disease population.

Further, the analysis method of the data analysis module can be to detect the biomarker by an enzyme-linked immunosorbent assay (ELISA), protein/peptide fragment chip detection, immunoblotting, microbead immunodetection or microfluidic immunodetection, and compare with the cutoff value of the biomarker according to the detection value, so as to judge whether the biomarker is positive.

In some embodiments, the assay of the data analysis module is performed by detecting the biomarker of the invention by antigen-antibody reaction, for example by ELISA or fluorescent or chemiluminescent immunoassay.

Further, the body fluid sample is whole blood, serum, plasma, tissue or cells, interstitial fluid, cerebrospinal fluid or urine sample.

In some embodiments, the liver cancer is hepatocellular carcinoma and/or non-hepatocellular carcinoma.

In some embodiments, whole blood, serum, plasma from the subject is preferred.

In some embodiments, the subject is a mammal, preferably a primate mammal, more preferably a human.

In some embodiments, the autoantibody is IgA (e.g., igA1, igA 2), igM, or IgG (e.g., igG1, igG2, igG3, igG 4).

The autoantibodies can be detected in a sample (e.g., plasma or serum) from a subject. In the present invention, "presence" or "absence" of autoantibodies is used interchangeably with "positive" or "negative"; this is judged as conventional in the art. For example, detection can be by a tumor-associated antigen and antigen-antibody specific reaction therebetween that results in the presence of any autoantibody in the combination.

In yet another aspect, the present invention provides a method for predicting whether an individual is liver cancer, comprising administering to the individual any one or more of the following autoantibodies ：anti-HNRPA1、anti-SCYL3、anti-Trim21、anti-PIM1、anti-CEA、anti-SSX2、anti-CAGE、anti-AIF、anti-IMPDH1、anti-CTAG2.

Further, the biomarker comprises the following autoantibodies: anti-HNRPA1, anti-SCYL, anti-Trim21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE and anti-CTAG2.

The present invention provides reagents, e.g., antigen protein combinations, for detecting said autoantibodies; and provides the application of the autoantibody combination or the detection reagent in preparing products for predicting the risk of liver cancer, screening, prognosis evaluation, treatment effect monitoring or recurrence monitoring and the like.

In yet another aspect, the invention provides the use of a biomarker in the preparation of a reagent for predicting whether an individual is non-hepatocellular carcinoma, the biomarker being any one or more of the autoantibodies selected from the group consisting of ：anti-HNRPA1、anti-SCYL3、anti-Trim21、anti-PIM1、anti-CEA、anti-SSX2、anti-CAGE、anti-AIF、anti-IMPDH1、anti-CTAG2.

Further, the biomarker comprises the following autoantibodies: anti-HNRPA1, anti-SCYL, anti-Trim21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE and anti-CTAG2.

In yet another aspect, the present invention provides a method for predicting, screening, prognostic evaluation, treatment effect monitoring or recurrence monitoring of a liver cancer, comprising the steps of:

(1) Quantifying each autoantibody in the autoantibody combination provided by the invention from a sample from a subject;

(2) Comparing the amount of the autoantibody to a reference threshold, and determining that the subject is at risk of developing liver cancer, has a poor prognosis, or has a poor therapeutic effect when it is above the reference threshold.

In step (1), the quantification comprises detection of each autoantibody in the autoantibody combination using the reagent (i.e. antigen protein combination) or the kit comprising the reagent provided by the invention.

In step (2), the reference threshold may be a reference level from a healthy person or population of healthy people; for example, it may be defined as the mean value plus 2 standard deviations of the population confirmed by physical examination as not having cancer.

Compared with the prior art, the autoantibody combination sensitivity for detecting liver cancer provided by the invention can be up to 60%, and the specificity is up to 90%. In particular, the diagnosis sensitivity of the kit for non-hepatocellular carcinoma can reach 72.73%, and the specificity can reach 90%.

The autoantibody biomarker for predicting whether an individual is liver cancer has the following beneficial effects:

1. Screening a series of novel autoantibody biomarkers capable of early predicting liver cancer, and screening an optimal autoantibody combination from the autoantibody biomarkers to construct a liver cancer early prediction system, wherein the sensitivity of predicting liver cancer can be up to 60%, the specificity can be up to 90%, and the diagnosis sensitivity of diagnosing non-liver cell cancer can be up to 72.73% and the specificity can be up to 90%;

2. The screened autoantibodies are combined and combined with a tumor marker AFP, so that the screening performance of liver cancer is further improved, the sensitivity of liver cancer detection can reach 88.57%, and the specificity can reach 97.67%;

3. The method is suitable for Chinese people, and has no obvious preference for liver cancer detection at different positions;

4. The prediction result based on the liver cancer early prediction system can be used for prediction of the risk of the liver cancer, screening, prognosis evaluation, treatment effect monitoring or recurrence monitoring and the like, and meets clinical requirements.

Detailed Description

(1) Diagnosis or detection

Diagnostic or test herein refers to the detection or assay of a biomarker in a sample, or the level of the biomarker of interest, such as absolute or relative, and then indicating whether the individual providing the sample is likely to have or suffer from a disease, or the likelihood of having a disease, by the presence or amount of the biomarker of interest. The diagnostic and detection meanings are interchangeable herein. The result of such detection or diagnosis is not directly as a direct result of the disease, but is an intermediate result, and if a direct result is obtained, it is also necessary to confirm that the patient has a disease by other auxiliary means such as pathology or anatomy. For example, the present invention provides a variety of novel biomarkers that have relevance to liver cancer, and changes in the levels of these markers have direct relevance to predicting liver cancer.

(2) Association of markers or biomarkers with prediction of thyroid cancer

Markers and biomarkers have the same meaning in the present invention. The term "associated with" as used herein refers to the presence or amount of a marker in a sample that has a direct correlation with a particular disease, e.g., a relative increase or decrease in amount, indicating a higher or lower likelihood of having the disease.

If multiple different markers are present in the sample at the same time or in a relatively varying amount, this is indicative of a higher likelihood of suffering from the disease. That is, some markers have strong association with the disease, some markers have weak association with the disease, or some even no association with the disease, among the marker categories. One or more of those markers with strong association can be used as markers for predicting the disease, and those markers with weak association can be combined with the strong markers to predict the disease, so that the accuracy of the prediction result is improved.

For the numerous autoantibody biomarkers found in the liver cancer patients of the invention, the increase or decrease of the presence or the content of the autoantibody biomarkers is directly related to whether the liver cancer is suffered or not.

(3) Correlation of 10 autoantibodies with liver cancer and tumor:

Drawings

FIG. 1 is a plot of autoantibody Anti-CTAG2 levels in the training cohort of example 1 versus evaluation scatter points of liver cancer and healthy groups;

FIG. 2 is a plot of the level of autoantibodies Anti-HNRPA1 in the training cohort of example 1 versus the evaluation scatter of the liver cancer group and the healthy group;

FIG. 3 is a plot of the level of autoantibodies Anti-CAGE in the training cohort of example 1 versus the evaluation scatter of the liver cancer and healthy groups;

FIG. 4 is a plot of the level of autoantibodies Anti-Trim21 in the training cohort of example 1 versus the evaluation scatter of the liver cancer and healthy groups;

FIG. 5 is a plot of the scatter plot of autoantibody Anti-CEA levels in the training cohort of example 1 versus the evaluation of liver cancer and healthy groups;

FIG. 6 is a plot of the level of autoantibodies Anti-SSX2 in the training cohort of example 1 versus the evaluation scatter of the liver cancer and healthy groups;

FIG. 7 is a plot of the level of autoantibodies Anti-IMPDH1 in the training cohort of example 1 versus the evaluation scatter of liver cancer and healthy groups;

FIG. 8 is a plot of the level of autoantibodies Anti-AIF in the training cohort of example 1 versus the evaluation scatter of the liver cancer and healthy groups;

FIG. 9 is a plot of the level of autoantibodies Anti-PIM1 in the training cohort of example 1 versus the evaluation scatter of the liver cancer and healthy groups;

FIG. 10 is a plot of the level of autoantibodies Anti-SCYL3 in the training cohort of example 1 versus the evaluation scatter of the liver cancer and healthy groups;

FIG. 11 is a plot of the level of autoantibodies Anti-CTAG2 in the validation cohort of example 2 versus the evaluation scatter of the liver cancer and healthy groups;

FIG. 12 is a plot of the evaluation scatter plot of autoantibody Anti-HNRPA1 levels in the validation cohort versus liver cancer and healthy groups of example 2;

FIG. 13 is a plot of the evaluation scatter plot of autoantibody Anti-CAGE levels in the validation cohort versus liver cancer and healthy groups of example 2;

FIG. 14 is a plot of the evaluation scatter plot of autoantibody Anti-Trim21 levels in the validation cohort versus liver cancer and healthy groups of example 2;

FIG. 15 is a plot of the scatter plot of the levels of autoantibodies Anti-CEA in the validation cohort versus the evaluation of liver cancer and healthy groups of example 2;

FIG. 16 is a plot of the evaluation scatter plot of autoantibody Anti-SSX2 levels in the validation cohort versus liver cancer and healthy groups of example 2;

FIG. 17 is a plot of the evaluation scatter plot of autoantibody Anti-IMPDH1 levels in the validation cohort versus liver cancer and healthy groups of example 2;

FIG. 18 is a plot of the scatter plot of the levels of autoantibodies Anti-AIF in the validation cohort versus the evaluation of liver cancer and healthy groups of example 2;

FIG. 19 is a plot of the scatter plot of the levels of autoantibodies Anti-PIM1 in the validation cohort versus the evaluation of liver cancer and healthy groups of example 2;

FIG. 20 is a plot of the evaluation scatter plot of autoantibody Anti-SCYL3 levels in the validation cohort versus liver cancer and healthy groups of example 2;

FIG. 21 is a ROC curve showing the performance of the autoantibody combinations of example 4 in the training cohort of liver cancer patients and healthy physical examination populations;

FIG. 22 is a ROC curve showing the analytical performance of the autoantibody combinations of example 5 in a validated cohort of liver cancer patients and healthy physical examination populations;

FIG. 23 is a ROC curve of the performance of the autoantibody combinations of example 6 in hepatocellular carcinoma patients and healthy physical examination populations;

FIG. 24 is a ROC curve of the performance of the autoantibody combinations of example 6 in non-hepatocellular carcinoma patients and healthy physical examination populations;

FIG. 25 is a ROC curve showing the analytical performance of AFP in liver cancer patients and healthy physical examination population in example 7;

FIG. 26 is a ROC curve showing the performance of the autoantibody combinations of example 7 in liver cancer patients and healthy physical examination populations.

Detailed Description

In the present invention, the term "antigen" or the term "antigenic protein" is used interchangeably.

The terms "antibody" and "autoantibody" are interchangeable in the present invention.

Furthermore, the present invention is directed to the following experimental operations or definitions, and it should be noted that the present invention may also be implemented using other conventional techniques in the art, and is not limited to the following experimental operations.

Preparation of recombinant antigen proteins

The cDNA fragment of the tumor antigen was cloned into PET28 (a) expression vector containing the 6XHIS tag. At the N-or C-terminus of the antigen, streptavidin or the like (biotin-binding tag protein) is introduced. The obtained recombinant expression vector is transformed into escherichia coli for expression. The protein expressed by the supernatant was purified by Ni-NTA affinity column and ion column. When the protein is expressed in inclusion bodies, the protein is denatured by 6M guanidine hydrochloride, renaturated and folded in vitro according to a standard method, and then purified by a Ni-NTA affinity column through a 6XHIS tag, so that antigen protein is obtained.

(II) preparation and preservation of serum or plasma

Serum or plasma of a liver cancer patient is collected when the patient is initially diagnosed with liver cancer and has not received any radiotherapy and chemotherapy or surgical treatment. Plasma or serum was prepared according to standard clinical procedures and stored in a-80 ℃ refrigerator for long periods of time.

(III) ELISA detection

The concentration of autoantibody markers in the sample was quantified by enzyme-linked immunosorbent assay (ELISA). The purified tumor antigen is immobilized to the microwell surface by its tag streptavidin or the like. Microwells were pre-coated with biotin-labeled Bovine Serum Albumin (BSA). Serum or plasma samples were diluted 1:110 fold with phosphate buffer and reacted by adding microwells (50 ml/well). After washing unbound serum or plasma components with wash solution, horseradish peroxidase (HRP) -conjugated anti-human IgG was added to each well for reaction. Then, TMB (3, 3', 5' -tetramethylbenzidine) as a reaction substrate was added for color development. Stop solution (1 NHCl) was added and absorbance was read by a microplate reader (OD) using a single spectrum at 450 nm. In this case, the amount of enzyme carried on the solid support is positively correlated with the amount of the test substance in the specimen, and the enzyme catalyzes the substrate to be a colored product. Qualitative or quantitative determination of the autoantibody is performed according to the degree of color reaction. Serum autoantibody concentrations were quantified using a standard curve.

The concentration of the antigen marker in the sample is quantified by a sandwich enzyme-linked immunosorbent assay. Connecting the specific antibody with a solid phase carrier to form a solid phase antibody, and washing to remove unbound antibody and impurities; and (3) adding a sample to be tested, namely diluting a serum or plasma sample by 1:110 times by using phosphate buffer, adding micropores to react (50 ml/hole), enabling the sample to contact and react with the solid-phase antibody for a period of time, and combining the antigen in the sample with the antibody on the solid-phase carrier to form a solid-phase antigen complex. Washing to remove other unbound material. Horseradish peroxidase (HRP) -conjugated anti-human IgG was added for reaction. Then, TMB (3, 3', 5' -tetramethylbenzidine) as a reaction substrate was added for color development. Stop solution (1N HCl) was added and absorbance was measured at 450nm using a microplate reader (OD). The amount of enzyme carried on the solid support is now positively correlated with the amount of test substance in the sample. The enzyme in the sandwich complex catalyzes the substrate to a colored product. Qualitative or quantitative determination of the antigen is performed according to the degree of color reaction.

(IV) threshold value of autoantibody and antigen protein (cutoff value)

The cutoff values for autoantibody and antigen levels were defined as being equal to the average of the healthy control cohorts in the control group (the group of people confirmed to have no cancer by physical examination) plus 2 Standard Deviations (SDs).

Fifth, positive and negative judgment of single autoantibody and antigen protein

For each autoantibody and antigen protein assay, a positive reaction is defined as detecting after quantifying the level of autoantibody or antigen protein in the sample, comparing it with a cutoff value, and making the value equal to or greater than the cutoff value positive; accordingly, a negative response is defined as < cutoff value negative.

The cutoff value of anti-HNRPA1 is 0.7, the cutoff value of anti-SCYL is 1, the cutoff value of anti-Trim21 is 0.5, the cutoff value of anti-PIM1 is 0.3, the cutoff value of anti-CEA is 0.3, the cutoff value of anti-SSX2 is 1.2, the cutoff value of anti-CAGE is 1.6, the cutoff value of anti-CTAG2 is 1, the cutoff value of anti-AIF is 1, and the cutoff value of anti-IMPDH1 is 0.5.

Positive determination of autoantibody and/or antigen protein combinations

Since the single autoantibody and/or single antigen protein has a low positive rate, the result of the analysis is combined with a plurality of autoantibodies and/or a plurality of antigen proteins to determine the predictive effect in order to increase the positive rate of detection of the autoantibody and/or antigen protein. The rules are: (1) Detecting a plurality of autoantibodies in a sample, and judging that the combined result of the antibodies is positive as long as one or more autoantibodies show positive; if all the autoantibodies are negative, the judgment result is negative; (2) Detecting a plurality of antigen proteins in a sample, and judging that the result is positive as long as one or more antigen proteins are positive; if all the antigen proteins are negative, the judgment result is negative; (3) Detecting a plurality of autoantibodies and a plurality of antigen proteins in a sample at the same time, and judging that the result is positive as long as one or more autoantibodies and/or antigen proteins are positive; and if all antibodies and antigen proteins are negative, the judgment result is negative.

(Seventh) statistical analysis method

Both groups were statistically analyzed using the GRAPHPAD PRISM v.6 (GRAPHPAD PRISM software, san diego, california) and IBM SPSS STATISTICS (IBM, new york) against Windows using the Mann-Whitney U test. In analyzing the relationship between each parameter, a Spearman correlation analysis was performed.

Eighth) sensitivity and specificity determination

Sensitivity: among all cases diagnosed with the gold standard, the cases in which the detection results of autoantibodies, autoantibody combinations, antigen proteins, antigen protein combinations, and combinations of autoantibodies and antigen proteins were positive account for the proportion of all cases.

Specificity: among all subjects diagnosed with no disease by gold standard, the subjects whose detection results of autoantibodies, autoantibody combinations, antigen proteins, antigen protein combinations, and combinations of autoantibodies and antigen proteins were negative were the proportion of all subjects.

The invention is described below with reference to specific examples. It will be appreciated by those skilled in the art that these examples are for illustration of the invention only and are not intended to limit the scope of the invention in any way. Sample collection has informed consent of the subject or patient and is approved by regulatory authorities.

The experimental methods in the following examples are conventional methods unless otherwise specified. The raw materials, reagent materials and the like used in the examples described below are commercially available products unless otherwise specified.

Example 1 screening of autoantibody biomarkers related to liver cancer in training cohorts

In this example, by summarizing 195 antigen proteins counted by a large number of public data, and performing autoantibody detection against purified antigen proteins on serum samples of a total of 92 training queues of 30 patients diagnosed with liver cancer and 62 healthy physical examination populations, it is desirable to screen autoantibody biomarkers related to liver cancer.

The healthy physical examination population is from not less than 3 different physical examination centers. All the serum of liver cancer patients is collected when the patients are diagnosed as liver cancer and do not receive any radiotherapy and chemotherapy or operation treatment, and is stored in a refrigerator at the temperature of minus 80 ℃. Training queue liver cancer patient information is shown in table 1.

TABLE 1 training queue liver cancer patient information

Age of	39～74
		Average age of	57.49
Sex (sex)
		Man's body	24
Female	6

Serum from 92 training queues consisting of 30 liver cancer patients and 62 healthy people participating in the study is detected, and the content of 195 candidate autoantibodies in serum samples of the training queues is detected respectively. 195 antigens are coated on the surface of a 96-well plate after being expressed and purified, are subjected to serum reaction with training queue diluted 1:110 times after being blocked, are subjected to reaction with anti-human IgG antibody-HRP horseradish catalase, are subjected to color reaction, and are detected by an enzyme-labeled instrument OD450nm wavelength.

The detection sensitivity and the specificity of various autoantibodies are calculated, wherein the sensitivity is calculated by the following method: cases with positive autoantibody detection results account for the proportion of all liver cancer patients; the specific calculation method comprises the following steps: the proportion of subjects whose autoantibody detection results are negative is the proportion of all subjects.

According to the sensitivity and specificity detection results of 195 antigens, under the condition that the specificity is higher than 96%, selecting biomarkers with higher sensitivity, screening 10 autoantibodies with higher sensitivity and specificity from the biomarkers, and setting antigen proteins of the 10 autoantibodies and serial numbers of a Uniprot database thereof as shown in table 2, wherein the website of the Uniprot database is www.uniprot.org.

Table 2, antigen proteins of 10 autoantibodies obtained by preliminary screening

Antigen proteins	Uniprot database sequence number	Antigen proteins	Uniprot database sequence number
				HNRPA1	P09651	SSX2	Q16385
SCYL3	Q8IZE3	CAGE	Q86TM3
				Trim21	P19474	AIF	P55008
PIM1	P11309	IMPDH1	P20839
				CEA	P06731	CTAG2	O75638

Correlation data for detection sensitivity, specificity, etc. for 10 individual autoantibodies are shown in table 3.

TABLE 3 detection sensitivity, specificity and about Density index of 10 individual autoantibodies in training cohorts

Molecular name	Sensitivity (%)	Specificity (%)	About sign index
				Anti-HNRPA1	23.33(7/30)	96.77(60/62)	0.201
Anti-SCYL3	26.67(8/30)	96.77(60/62)	0.2344
				Anti-Trim21	20.00(6/30)	100.00(62/62)	0.2
Anti-PIM1	26.67(8/30)	98.39(61/62)	0.2506
				Anti-CEA	16.67(5/30)	98.39(61/62)	0.1506
Anti-SSX2	13.33(4/30)	98.39(61/62)	0.1172
				Anti-CAGE	10.00(3/30)	100.00(62/62)	0.1
Anti-AIF	10.00(3/30)	98.39(61/62)	0.0839
				Anti-IMPDH1	20.00(6/30)	100.00(62/62)	0.2
Anti-CTAG2	10.00(3/30)	98.39(61/62)	0.0839

From Table 3, it can be seen that among the 10 autoantibodies in the training cohort, the about dengue index was the highest was Anti-PIM1, the second Anti-SCYL3, and the third Anti-HNRPA1.

The horizontal distribution scatter diagrams of the 10 screened autoantibodies in the liver cancer group and the healthy control group of the training queue are shown in fig. 1-10, wherein fig. 1-10 are evaluation scatter relation diagrams of the levels of the autoantibodies Anti-CTAG2、Anti-HNRPA1、Anti-CAGE、Anti-Trim21、Anti-CEA、Anti-SSX2、Anti-IMPDH1、Anti-AIF、Anti-PIM1、Anti-SCYL3 and the liver cancer group and the healthy group respectively.

As can be seen from fig. 1 to 10, the distribution sensitivity of the individual tumor autoantibodies in the tumor patients is low due to the difference in the immune system of the liver cancer patients and the diversity of the tumor generation mechanism. Statistical analysis of the horizontal distribution of autoantibodies in liver cancer group and healthy control group using Mann-WHITNEY TEST found that the horizontal distribution of antibodies against CTAG2, HNRPA1, CAGE, TRIM21, CEA, SSX2, AIF, PIM1 and SCYL3 was significantly different in training queue liver cancer group and healthy control group (p < 0.05), and IMPDH1 antibody molecules also had an ascending trend in liver cancer group.

Example 2 Single autoantibody sensitivity and specificity detection in validation cohorts

In this example, the 10 autoantibody markers obtained by the screening in example 1 were screened and validated by another independent group of 92 healthy physical examination groups and 92 liver cancer patients as validation queues. The healthy physical examination population is from not less than 3 different physical examination centers. All the serum of liver cancer patients is collected when the patients are diagnosed as liver cancer and do not receive any radiotherapy and chemotherapy or operation treatment, and is stored in a refrigerator at the temperature of minus 80 ℃. The information of the liver cancer patients in the verification queue is shown in Table 4.

TABLE 4 training queue liver cancer patient information

Age of	42～78
		Average age of	60.36
Sex (sex)
		Man's body	75
Female	17

And similarly, the liver cancer antigen is coated on the surface of a 96-well plate after being expressed and purified, is subjected to reaction with liver cancer serum diluted by 1:110 times or serum of a physical examination control group, is subjected to reaction with anti-human IgG antibody-HRP horseradish peroxidase, is subjected to color reaction, and is detected by using an enzyme-labeled instrument OD450nm wavelength. The detection sensitivity and specificity of each autoantibody were calculated, and the detection sensitivity, specificity and about log index of 10 individual antibodies are shown in table 5.

TABLE 5 detection sensitivity and specificity of 5 individual autoantibodies in the validation cohort

Molecular name	Sensitivity (%)	Specificity (%)	About sign index
				Anti-HNRPA1	10.87(10/92)	100.00(92/92)	0.1087
Anti-SCYL3	6.52(6/92)	97.83(90/92)	0.0435
				Anti-Trim21	13.04(12/92)	100.00(92/92)	0.1304
Anti-PIM1	19.57(18/92)	98.91(91/92)	0.1848
				Anti-CEA	5.43(5/92)	98.91(91/92)	0.0434
Anti-SSX2	13.04(12/92)	97.83(90/92)	0.1087
				Anti-CAGE	5.43(5/92)	98.91(91/92)	0.0434
Anti-AIF	6.52(6/92)	98.91(91/92)	0.0543
				Anti-IMPDH1	9.78(9/92)	97.83(90/92)	0.0761
Anti-CTAG2	9.78(9/92)	98.91(91/92)	0.0869

As can be seen from table 5, the specificity of 10 autoantibodies was higher than 97% for the validation cohort, with Anti-PIM1 being the highest about the dengue index, indicating better correlation with liver cancer, followed by Anti-Trim21.

The horizontal distribution scatter diagrams of the 10 screened autoantibodies in the liver cancer group and the healthy control group of the verification queue are shown in fig. 11-20, wherein fig. 11-20 are evaluation scatter relation diagrams of the levels of the autoantibodies Anti-CTAG2、Anti-HNRPA1、Anti-CAGE、Anti-Trim21、Anti-CEA、Anti-SSX2、Anti-IMPDH1、Anti-AIF、Anti-PIM1、Anti-SCYL3 and the liver cancer group and the healthy group respectively.

From fig. 11 to 20, it can be seen that the distribution sensitivity of the individual tumor autoantibodies in the liver cancer patients is low due to the difference in the immune system of the liver cancer patients and the diversity of the tumor generation mechanism. Statistical analysis of the level distribution of autoantibodies in liver cancer group and healthy control group was performed using Mann-WHITNEY TEST, and it was found that the level distribution of anti-TRIM 21, AIF and PIM1 antibodies in validation queue liver cancer and healthy control group were significantly different (p < 0.05), and other antibody molecules also had an upward trend in liver cancer group.

Example 3 screening of autoantibody biomarker combinations related to liver cancer

According to the detection condition of single candidate autoantibodies in the training queue crowd, candidate antibodies with the specificity being more than 96% are selected according to the embodiment 1, on the premise of ensuring high specificity, the single positive contribution of the antibodies is combined (namely, the candidate molecules with high overlapping positive detection rate are excluded), so that a detection model covers more liver cancer patients to the maximum extent, different autoantibody combinations are formed, and the detection is carried out in the training queue by using corresponding detection reagents, and the results are shown in Table 6.

TABLE 6 sensitivity and specificity of autoantibody combinations

As can be seen from Table 6, in comparison with groups 8, 9 and 10, increasing the detection of anti-AIF and anti-IMPDH1 in group 8 did not increase the number of detections in the liver cancer patient population, and therefore they did not contribute to the molecular combination, so they were excluded from the liver cancer autoantibody detection combination.

Based on the principle that the sum of the sensitivity and the specificity of different molecular combinations is maximum, the Anti-HNRPA1+ Anti-SCYL & lt 3 & gt + Anti-TRIM21+ Anti-PIM1+ Anti-CEA + Anti-SSX2+ Anti-CAGE + Anti-CTAG2 is selected as the optimal molecular combination, wherein the sensitivity is 60.00% and the specificity is 90.32%.

Example 4 analysis of the working characteristics (ROC) of subjects in training cohorts with the autoantibody combinations of the invention

This example further analyzed 10 individual autoantibody molecules and various combinations using ROC curves, and in a training cohort (example 1), screening ability against liver cancer patients was analyzed for healthy and liver cancer patients, and the results are shown in table 7.

Table 7 comparison of liver cancer prediction Properties of different autoantibody molecules and combinations (training set)

As can be seen from Table 7, the AUC values of the combined models are significantly improved, wherein the autoantibody combination (Anti-HNRPA1+anti-SCYL 3+anti-TRIM21+anti-PIM1+anti-CEA+anti-SSX2+anti-CAGE+an ti-CTAG 2) of group 17 has better sensitivity and specificity, higher about index and highest AUC value, which is consistent with the autoantibody combination result obtained by the screening of example 3.

This example continues to analyze the screening ability of the autoantibody combinations of the invention (group 17) in a training cohort for liver cancer patients using ROC curves. As shown in FIG. 21, the sensitivity of the combination of molecules reached 59.26% at about the maximum value of the dengue index (under ideal conditions) in the control of healthy subjects, at which time the specificity was 93.22% and the area under the curve was 0.8280.

Example 5 analysis of subject working characteristics (ROC) of the autoantibody combinations of the invention in validation cohorts this example further uses ROC curves to analyze the screening ability of the antibody combinations of the invention (Anti-hnrpa1+anti-SCYL 3+anti-TRIM21+anti-PIM1+anti-cea+anti-ssx2+anti-cage+an ti-CTAG 2) in validation cohorts (example 2) for liver cancer patients.

As shown in FIG. 22, the sensitivity of the molecular combination of the present invention reached 58.02% at about the maximum value of the dengue index (under ideal conditions) in the control of healthy physical examination population, at which time the specificity was 98.80% and the area under the curve was 0.8054.

Example 6 analysis of the antibody detection model of the present invention for different types of liver cancer detection Capacity

In this example, the subjects with known liver cancer onset tissues were classified into hepatocellular carcinoma (28 cases) and non-hepatocellular carcinoma (11 cases). For different types of patients, their serum test data were analyzed. The ability to detect hepatocellular carcinoma and non-hepatocellular carcinoma was analyzed using an antibody combination (Anti-HNRPA1+anti-SCYL 3+anti-TRIM 21+anti-PIM1+anti-CEA+anti-SSX2+anti-CAGE+anti-CTAG 2). The results are shown in FIGS. 23 and 24.

As shown in FIG. 23, the detection sensitivity of the detection model of the present invention to hepatocellular carcinoma was 60.00% at the maximum about dengue index, at which time the specificity was 96.30%, and the area under the curve was 0.8489; as shown in FIG. 24, the sensitivity for detection of non-hepatocellular carcinoma was 72.73%, the specificity at this time was 90.00%, the area under the curve was 0.8182, and the sensitivity was higher.

From the ROC analysis results, the detection model has similar detection capability (area under curve) for different types of liver cancer, and has no obvious preference for detecting liver cancer at different positions.

Example 7 screening of patients with liver cancer by combining clinically common tumor antigen marker AFP with autoantibody combination

The biomarker related to liver cancer clinically adopted at present is AFP, and the embodiment detects the tumor antigen marker AFP for patient groups in a verification queue. And then combining an antigen marker AFP with the established antibody combination (Anti-HNRPA1+anti-SCYL 3+anti-TRIM 21+anti-PIM1+anti-CEA+anti-SSX2+anti-CAGE+anti-CTAG 2) to establish a liver cancer detection model of the autoantibody and antigen collocation. As shown in Table 8, the sensitivity to liver cancer screening was further improved after AFP (threshold 25 ng/mL) was combined with the antibody.

Table 8 sensitivity of AFP-autoantibody combination to screening for liver cancer

In this example, ROC analysis using AFP alone as a liver cancer marker was continued, and the result (fig. 25) showed that the sensitivity was 67.86% at the maximum about the mount index (ideal state), the specificity was 94.87%, and the area under the curve was 0.8636.

In this example, further ROC analysis was performed on the antigen-antibody combination (Anti-HNRPA1+anti-SCYL 3 +anti-TRIM21+anti-PIM1+anti-CEA+anti-SSX2+anti-CAGE+anti-CTAG 2+AFP) in the validation queue, and the ROC curve is shown in FIG. 26. In the case of detection of antigen binding to autoantibodies, the sensitivity of the molecule combination of the invention reaches 88.57% at about the maximum value of the dengue index (under ideal conditions) in comparison with healthy physical examination population, the specificity is 97.67% at this time, and the area under the curve is 0.9389. Therefore, the screening capability of the liver cancer detection model combined by the autoantibody combination and the AFP on the patients is further improved.

Although the present invention is disclosed above, the present invention is not limited thereto. Various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention, and the scope of the invention should be assessed accordingly to that of the appended claims.

Claims (7)

1. A biomarker combination for predicting whether an individual is liver cancer, comprising the following autoantibodies:

anti-HNRPA1, anti-SCYL, anti-Trim21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE and anti-CTAG2.

2. The biomarker combination according to claim 1, further comprising AFP.

3. A kit for predicting whether an individual is liver cancer comprising a detection reagent comprising a biomarker combination according to claim 1 or 2.

4. A system for predicting whether an individual is liver cancer, the system comprising a data analysis module; the data analysis module is used for analyzing the detection condition of the biomarker, wherein the biomarker is the biomarker combination as claimed in claim 1 or 2.

5. The system of claim 4, wherein the data analysis module analyzes the data by: detecting whether a biomarker in a body fluid sample of an individual is positive; and comparing whether the biomarker is positive or not through a cutoff value, wherein the biomarker is positive or not, and the biomarker is negative or not.

6. The system of claim 5, wherein the analysis method of the data analysis module further comprises: when one or more biomarkers in the biomarker combination are positive, the biomarker combination is positive, and the individual is predicted to be a liver cancer patient; when all the biomarkers in the biomarker combination are negative, the biomarker combination is negative, and the individual is predicted to be a healthy population or a benign disease population.

7. Use of a biomarker in the preparation of a reagent for predicting whether an individual has liver cancer, wherein the biomarker is an autoantibody comprising: anti-HNRPA1, anti-SCYL, anti-Trim21, anti-PIM1, anti-CEA, anti-SSX2, anti-CAGE and anti-CTAG2.