A biomarker and detection kit for AFP-negative liver cancer diagnosis

Abstract

The application belongs to the field of medical biotechnology, and specifically discloses a biomarker and a detection kit for AFP-negative liver cancer diagnosis. A model constructed by combining the biomarkers is PRE (PRE=AFP-negative HCC, 3TAAbs)=1 / [1+exp(3.323-5.873*EGFR+7.415*PHF6-12.453*PTEN)], the AUC of the model for diagnosing AFP-negative liver cancer is 0.802, the sensitivity is 70.2%, and the specificity is 77.4%. The model combined with the clinical index AFP can assist the clinical diagnosis of whole liver cancer, the diagnostic sensitivity can be improved to 86.4%, and the specificity is as high as 90%, which greatly improves the detection rate of liver cancer, and has the advantages of high sensitivity, strong specificity, easy operation, low cost and the like. The application also provides a kit for AFP-negative liver cancer diagnosis, which contains reagents for detecting the above biomarkers. By detecting the expression levels of the three biomarkers in human serum, the application can effectively distinguish AFP-negative liver cancer patients from healthy people, and can be used for the auxiliary diagnosis of AFP-negative liver cancer.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of medical biotechnology, and specifically discloses a biomarker and a detection kit for AFP-negative liver cancer diagnosis. BACKGROUND

[0002] Timely and effective treatment after early diagnosis of liver cancer is the main measure to reduce its mortality. At present, China recommends that high-risk groups of liver cancer perform serum AFP and liver ultrasound examination every 6 months to achieve early screening of liver cancer, and then dynamic enhanced CT and multi-modal MRI scanning are applied to the abnormal screening subjects to make clear diagnosis. A large number of studies have confirmed that about 40% of liver cancer patients do not have elevated AFP levels. These AFP-negative liver cancer patients often have small tumor volume or are in the early stage, and the clinical symptoms are not obvious, and the imaging features of the lesion site are similar to benign nodules, so they are also not easy to be detected by imaging techniques such as ultrasound, resulting in a higher probability of missed diagnosis or misdiagnosis of AFP-negative liver cancer patients in the early screening process. Studies have found that the lower the concentration of AFP in the serum of liver cancer patients, the better the prognosis, so accurate and effective diagnosis of AFP-negative liver cancer is an important measure to improve the early diagnosis rate, 5-year survival rate, and reduce the mortality rate of liver cancer, and is also a bottleneck problem that needs to be broken through.

[0003] Serological markers are attracting attention due to their relative non-invasiveness, objectivity, economy, and practicality. Finding a serological marker that can compensate for AFP is a hot topic in liver cancer diagnosis. Many studies have shown that the serum of cancer patients contains autoantibodies that can react with a unique set of self-cell antigens, known as tumor-associated antigens (TAA). Unlike autoantibodies in autoimmune diseases, TAA autoantibodies have been detected in various tumors. Some autoantibodies exist before the clinical diagnosis of tumors for several months to several years, in addition, TAA autoantibodies as immune diagnostic markers may have greater advantages, they are more abundant and have a longer duration through the amplification of immune responses than TAA itself, and are easier to detect. Therefore, TAA autoantibodies have great potential in the early diagnosis of cancer. Therefore, it is necessary to screen markers related to AFP-negative liver cancer based on AFP-negative liver cancer, and to diagnose liver cancer in combination with the existing marker AFP to improve the early diagnosis rate of liver cancer. SUMMARY

[0004] In view of the problems and deficiencies in the prior art, the purpose of the present application is to provide a biomarker and a detection kit for AFP-negative liver cancer diagnosis.

[0005] To achieve the purpose of the application, the technical solutions adopted by the present application are as follows:

[0006] The application provides the use of a reagent for detecting a biomarker for diagnosing AFP-negative liver cancer in the preparation of a product for diagnosing AFP-negative liver cancer; the biomarker is at least one of an anti-tumor related antigen EGFR autoantibody, an anti-tumor related antigen PHF6 autoantibody, and an anti-tumor related antigen PTEN autoantibody. The expression levels of the anti-tumor related antigen EGFR autoantibody, the anti-tumor related antigen PHF6 autoantibody, and the anti-tumor related antigen PTEN autoantibody in serum of AFP-negative liver cancer patients are all higher than those in normal persons, and the difference is statistically significant.

[0007] According to the above-mentioned application, preferably, the anti-tumor related antigen EGFR autoantibody, the anti-tumor related antigen PHF6 autoantibody, and the anti-tumor related antigen PTEN autoantibody are all corresponding anti-tumor related antigen autoantibodies in serum, plasma, interstitial fluid or urine of the subject.

[0008] According to the above-mentioned application, preferably, the anti-tumor related antigen EGFR autoantibody, the anti-tumor related antigen PHF6 autoantibody, and the anti-tumor related antigen PTEN autoantibody are anti-tumor related antigen autoantibodies in serum, plasma, interstitial fluid or urine of the subject before receiving tumor treatment. More preferably, the tumor treatment is chemotherapy, radiotherapy or tumor resection.

[0009] According to the above-mentioned application, preferably, the subject is a mammal, more preferably, the subject is a primate mammal, most preferably, the subject is a human.

[0010] According to the above-mentioned application, preferably, the reagent is a reagent for detecting the biomarker in a sample by enzyme-linked immunosorbent assay, protein chip, immunoblotting or microfluidic immunoassay.

[0011] According to the above-mentioned application, preferably, the sample is serum, plasma, interstitial fluid or urine. The product is a protein chip, a kit or a preparation.

[0012] According to the application, preferably, the biomarker is a combination of anti-tumor associated antigen EGFR autoantibody, anti-tumor associated antigen PHF6 autoantibody, and anti-tumor associated antigen PTEN autoantibody; when the product is used for diagnosing AFP-negative liver cancer, the probability of predicting liver cancer is calculated by the formula: PRE = 1 / [1+exp(3.323-5.873*EGFR+7.415*PHF6-12.453*PTEN)], wherein PRE represents the probability of predicting AFP-negative liver cancer, EGFR represents the expression level of anti-tumor associated antigen EGFR autoantibody, PHF6 represents the expression level of anti-tumor associated antigen PHF6 autoantibody, PTEN represents the expression level of anti-tumor associated antigen PTEN autoantibody, and exp represents the exponential function with natural constant e as the base.

[0013] According to the application, preferably, the reagent is an antigen or an antibody for detecting the biomarker. More preferably, the reagent is an antigen for detecting the biomarker, and the antigen is at least one of EGFR protein, PHF6 protein, and PTEN protein.

[0014] The second aspect of the application provides an application of a reagent for detecting a biomarker for diagnosing liver cancer in the preparation of a product for diagnosing liver cancer; the biomarker is a combination of anti-tumor associated antigen EGFR autoantibody, anti-tumor associated antigen PHF6 autoantibody, anti-tumor associated antigen PTEN autoantibody, and AFP protein.

[0015] According to the application, preferably, the reagent is a reagent for detecting the biomarker in a sample by enzyme-linked immunosorbent assay, protein chip, immunoblotting, or microfluidic immunoassay.

[0016] According to the application, preferably, the sample is serum, plasma, interstitial fluid, or urine. The product is a protein chip, a kit, or a preparation.

[0017] According to the application, preferably, the biomarker is a combination of anti-tumor related antigen EGFR autoantibody, anti-tumor related antigen PHF6 autoantibody, anti-tumor related antigen PTEN autoantibody and AFP protein; when the product is used for diagnosing liver cancer, the probability calculation formula for predicting liver cancer is P = 1 / [1 + exp(2.487-0.256 x AFP-5.003 x PRE)], wherein P represents the probability of predicting liver cancer, AFP represents the concentration of alpha-fetoprotein in serum, PRE represents the probability of predicting AFP-negative liver cancer and PRE = 1 / [1 + exp(3.323-5.873 x EGFR + 7.415 x PHF6-12.453 x PTEN)], EGFR represents the expression amount of anti-tumor related antigen EGFR autoantibody, PHF6 represents the expression amount of anti-tumor related antigen PHF6 autoantibody, and PTEN represents the expression amount of anti-tumor related antigen PTEN autoantibody.

[0018] The third aspect of the present application provides a kit for diagnosing liver cancer, wherein the kit comprises reagents for detecting biomarkers, and the biomarkers are at least one of anti-tumor related antigen EGFR autoantibody, anti-tumor related antigen PHF6 autoantibody and anti-tumor related antigen PTEN autoantibody; or the biomarkers are a combination of anti-tumor related antigen EGFR autoantibody, anti-tumor related antigen PHF6 autoantibody, anti-tumor related antigen PTEN autoantibody and AFP protein.

[0019] According to the kit, preferably, the kit detects the biomarkers in the sample by enzyme-linked immunosorbent assay, protein chip, immunoblotting or microfluidic immunoassay.

[0020] According to the kit, preferably, when the biomarkers are a combination of anti-tumor related antigen EGFR autoantibody, anti-tumor related antigen PHF6 autoantibody and anti-tumor related antigen PTEN autoantibody, the probability calculation formula for predicting liver cancer is PRE = 1 / [1 + exp(3.323-5.873 x EGFR + 7.415 x PHF6-12.453 x PTEN)], wherein PRE represents the probability of predicting AFP-negative liver cancer, EGFR represents the expression amount of anti-tumor related antigen EGFR autoantibody, PHF6 represents the expression amount of anti-tumor related antigen PHF6 autoantibody, and PTEN represents the expression amount of anti-tumor related antigen PTEN autoantibody; and exp represents the exponential function with the natural constant e as the base.

[0021] When the biomarker is a combination of anti-tumor related antigen EGFR autoantibody, anti-tumor related antigen PHF6 autoantibody, anti-tumor related antigen PTEN autoantibody and AFP protein, the probability calculation formula of the kit for predicting liver cancer is P=1 / [1+exp(2.487-0.256*AFP-5.003*PRE)], wherein P represents the probability of predicting liver cancer, AFP represents the concentration of alpha-fetal protein in serum, PRE represents the probability of predicting AFP-negative liver cancer and PRE=1 / [1+exp(3.323-5.873*EGFR+7.415*PHF6-12.453*PTEN)], EGFR represents the expression amount of anti-tumor related antigen EGFR autoantibody, PHF6 represents the expression amount of anti-tumor related antigen PHF6 autoantibody, and PTEN represents the expression amount of anti-tumor related antigen PTEN autoantibody.

[0022] According to the kit, preferably, the kit is an ELISA detection kit.

[0023] According to the kit, preferably, the sample is serum, plasma, interstitial fluid or urine.

[0024] According to the kit, preferably, the ELISA detection kit further comprises a sample diluent, a secondary antibody, an antibody diluent, a washing solution, a color developing solution and a termination solution.

[0025] In the present application, the basic information of anti-tumor related antigens EGFR, PHF6 and PTEN is as follows:

[0026] EGFR (epidermal growth factor receptor) is an epidermal growth factor receptor. PHF6 (planthomeodomain finger protein 6) is a plant homeodomain finger protein 6. PTEN (Phosphatase and tensin homolog) is a homologous phosphatase and tensin. The sequence number of EGFR protein in Uniprot is P00533; the sequence number of PHF6 protein is Q8IWS0; and the sequence number of PTEN protein is P60484.

[0027] Compared with the prior art, the present application has the following advantages:

[0028] (1) The present application first discovers that the expression levels of the autoantibodies against the tumor-related antigens EGFR, PHF6 and PTEN in the serum of AFP-negative liver cancer patients are significantly higher than those in AFP-positive liver cancer patients and normal persons, and the difference is statistically significant. By detecting the expression levels of the autoantibodies against the tumor-related antigens EGFR, PHF6 and PTEN in human serum, AFP-negative liver cancer and normal persons can be effectively diagnosed and distinguished. It has been verified that when any one of the autoantibodies against the tumor-related antigens EGFR, PHF6 and PTEN is used alone to diagnose and distinguish AFP-negative liver cancer patients and normal persons, the AUC value of the ROC curve is more than 0.60; when multiple markers are used in combination, the AUC value of the ROC curve is closer to 1 than that of a single index, the distinguishing effect is good, and the diagnosis effect is good. Therefore, the three markers discovered by the present application for the diagnosis of AFP-negative liver cancer can be used for the auxiliary diagnosis of liver cancer.

[0029] (2) When the three markers, i.e., the autoantibodies against the tumor-related antigens EGFR, PHF6 and PTEN, are used as a combination to diagnose and distinguish AFP-negative liver cancer patients and normal persons, the AUC of the ROC curve is 0.802, the detection sensitivity is 70.2% (and the ratio of AFP-negative liver cancer patients who are correctly diagnosed as liver cancer when the three markers are used for diagnosis is 70.2%), and the specificity is 77.4% (and the ratio of healthy persons who are determined as healthy when the three markers are used for diagnosis in healthy controls is 77.4%). Therefore, the markers of the present application have high sensitivity and specificity, and effectively improve the detection rate of AFP-negative liver cancer.

[0030] (3) The present application combines the three markers, i.e., the autoantibodies against the tumor-related antigens EGFR, PHF6 and PTEN, to construct a diagnostic model, and the diagnostic model has high diagnostic value for AFP-negative liver cancer. Compared with the diagnosis of liver cancer by AFP alone (the sensitivity is 55.8%), when the diagnostic model is combined with the existing marker AFP, the sensitivity can be improved to 86.4%, and the specificity is as high as 90%, which greatly improves the detection rate of liver cancer, enables more liver cancer patients in the early stage to receive timely and effective treatment, thereby improving the 5-year survival rate of liver cancer, improving the quality of life of patients, prolonging their life, and reducing the disease burden of patients' families and even the whole society.

[0031] (4) The kit of the present application can detect the expression levels of anti-tumor related antigen EGFR autoantibody, anti-tumor related antigen PHF6 autoantibody and anti-tumor related antigen PTEN autoantibody in human serum by indirect ELISA method, so that AFP-negative liver cancer patients and healthy controls can be accurately distinguished and diagnosed, thereby providing a new reference for the diagnosis of liver cancer for clinicians.

[0032] (5) The kit of the present application uses serum as the detection sample, so that invasive diagnosis can be avoided, the risk of liver cancer can be obtained by minimally invasive serum detection, the amount of blood required is small, the pain of the person to be detected is small, the compliance is high, the operation is simple, the result detection time is short, and the kit has a broad market prospect and social benefits. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 Fig. 4 is a ROC curve diagram for diagnosing AFP-negative liver cancer using healthy people as controls, wherein the positive rates of three anti-tumor related antigen autoantibodies in AFP-negative liver cancer group, AFP-positive liver cancer group and normal control group in proteome chip detection are shown;

[0034] Figure 2 Fig. 5 is a scatter plot of the relative titers of three anti-tumor related antigen autoantibodies in serum of AFP-negative liver cancer group and normal control group (or chronic hepatitis B group) detected by ELISA in three clinical centers, wherein A is the result of Zhengzhou clinical center, B is the result of Nanchang clinical center, and C is the result of Beijing clinical center; HCC represents AFP-negative liver cancer group, HD represents normal control group, and HBV represents chronic hepatitis B group;

[0035] Figure 3 Fig. 6 is a ROC curve diagram for diagnosing AFP-negative liver cancer patients and normal people using three autoantibody single indicators in Zhengzhou, Nanchang and Beijing three clinical centers;

[0036] Figure 4 Fig. 7 is a ROC curve diagram for diagnosing AFP-negative liver cancer patients and normal people using different autoantibody combinations;

[0037] Figure 5 Fig. 8 is a nomogram of a logistic regression diagnostic model constructed by three marker combinations;

[0038] Figure 6 Fig. 9 is a ROC curve diagram of a logistic regression diagnostic model constructed by three marker combinations for diagnosing AFP-negative liver cancer patients, wherein A is a ROC curve diagram of the diagnostic model for diagnosing AFP-negative liver cancer patients and normal people, and B is a ROC curve diagram of the diagnostic model for diagnosing AFP-negative liver cancer patients and chronic hepatitis B patients; HCC represents AFP-negative liver cancer group, HD represents normal control group, and HBV represents chronic hepatitis B group;

[0039] Figure 7 The figure is the evaluation of the diagnostic value of the AFP-negative hepatocarcinoma Logistic regression diagnostic model constructed by the three marker combinations and AFP for all hepatocarcinoma. In the figure, The model represents the AFP-negative hepatocarcinoma Logistic regression diagnostic model constructed by the three marker combinations, AFP & The model represents the all hepatocarcinoma diagnostic model constructed by the AFP-negative hepatocarcinoma diagnostic model and AFP, HCC represents the hepatocarcinoma group, and NC represents the normal control group. DETAILED DESCRIPTION

[0040] The following examples are only used to further illustrate the present application. It should be noted that all the technical and scientific terms used in the present application have the same meaning as those in the technical field to which the present application belongs, unless otherwise specified. The experimental methods not specified in the following examples are conventional techniques in the technical field, or are according to the conditions suggested by the manufacturers; the reagents or instruments not specified by the manufacturers are conventional products that can be obtained from the market.

[0041] In order for those skilled in the art to more clearly understand the technical solutions of the present application, the technical solutions of the present application will be described in detail below with specific examples.

[0042] Example 1: Screening of markers for early diagnosis of AFP-negative hepatocarcinoma using human proteome chip

[0043] 1. Experimental samples:

[0044] Forty serum samples of AFP-negative hepatocarcinoma patients (AFP-negative hepatocarcinoma group), 56 serum samples of AFP-positive hepatocarcinoma patients (AFP-positive hepatocarcinoma group) and 49 serum samples of normal persons (normal control group) were collected from the sample library of the Henan Tumor Epidemiology Laboratory. Among them, the serum samples of 40 AFP-negative hepatocarcinoma patients and 56 AFP-positive hepatocarcinoma patients were derived from hepatocarcinoma patients who were diagnosed by pathology and had not received any treatment, and the serum samples of 49 normal persons were derived from healthy subjects. The inclusion criteria for healthy subjects were: no cardiovascular, respiratory, liver, kidney, gastrointestinal, endocrine, blood, mental or nervous system diseases and history of the above diseases, no acute or chronic diseases, no autoimmune diseases, and no any evidence of tumor. This study was approved by the Ethics Committee of Zhengzhou University, and all the research subjects had signed the informed consent form.

[0045] Serum collection: 5 mL peripheral blood of the research subjects in a fasting state was collected in a blood collection tube without anticoagulant, and after standing at room temperature for 1 h, it was placed in a centrifuge, set at 4°C, 3000 rpm, and centrifuged for 10 min. Then the serum on the upper layer of the blood collection tube was sucked out and aliquoted into 1.5 mL EP tubes, the sample number was marked on the top and side of the EP tube, and it was placed in a -80°C refrigerator for frozen storage, and the blood collection date and storage location were recorded. Before use, the serum was taken out and placed in a 4°C refrigerator for thawing and aliquoting to avoid repeated freezing and thawing of the serum.

[0046] 2. Human protein custom chip detection

[0047] Using HuProt TM Human protein custom chip detection was used to detect the expression levels of autoantibodies in 40 AFP-negative hepatocellular carcinoma serum samples, 56 AFP-positive hepatocellular carcinoma serum samples, and 49 normal serum samples. The HuProt TM The human protein custom chip contains 143 GST-tagged proteins or protein fragments purchased from the American CDI laboratory, and these proteins are all recombinant proteins encoded by cancer driver genes; each chip can detect 14 serum samples at the same time, and the proteins fixed on the chip interact with specific autoantibodies in the serum to bind.

[0048] (1) Experimental method:

[0049] 1) Rewarming: the HuProt TM The human protein custom chip was taken out from the -80°C refrigerator, placed in the 4°C refrigerator for 30 min, and then continued to be warmed at room temperature for 15 min; since each HuProt TM The human protein custom chip can detect 14 serum samples at the same time, so the fence containing 14 blocks was first fixed on the chip;

[0050] 2) Blocking: the warmed chip was placed in the chip incubation box with the front face upwards, 200 μL blocking solution (containing 4 μg BSA and 200 μL 1×PBST solution) was added to each block, and it was placed in a side swing shaker at 20 rpm for 3 h of blocking at room temperature;

[0051] 3) Serum sample incubation: after blocking was completed, the blocking solution was discarded, 200 μL of pre-diluted serum incubation solution (the serum sample was diluted at a ratio of 1:50 with diluent to obtain the diluted serum incubation solution; the diluent contains 1% BSA and 200 μL 1×PBST solution) was added to each block, and it was placed in a side swing shaker at 20 rpm for overnight incubation at 4°C;

[0052] 4) Washing: After the incubation is completed, the serum sample is aspirated, the fence is removed, and the chip is placed in a chip washing box containing PBST washing solution, and is washed on a horizontal shaker at 80 rpm for 3 times, 10 min each time;

[0053] 5) Secondary antibody incubation: After the washing is completed, the chip is transferred to a secondary antibody incubation box, 3 mL of secondary antibody incubation solution diluted at a ratio of 1:1000 (the secondary antibody is a fluorescently labeled anti-human IgG antibody, and the dilution solution is composed of 1 g of BSA and 100 mL of 1x PBST solution, and the secondary antibody is diluted at a ratio of 1:1000 to obtain the secondary antibody incubation solution) is added, and the chip is incubated on a side-to-side shaker at 40 rpm at room temperature for 1 h;

[0054] 6) Washing: The chip is taken out (attention should be paid to not touching or scratching the upper surface of the chip), placed in a chip washing box, and 1x PBST solution is added, and the chip is washed on a horizontal shaker at 80 rpm for 3 times, 10 min each time. After completion, the PBST washing solution is replaced with ddH2O and washed for 2 times, 10 min each time;

[0055] 7) Drying: After the washing is completed, the chip is placed in a chip drying machine for centrifugal drying (attention should be paid to not touching or scratching the surface of the chip during the process of taking the chip);

[0056] 8) Scanning: The dried chip is scanned according to the operation instructions of the LuxScan 10K microarray chip scanner instrument, and the fluorescence signal is recorded (the strength of the fluorescence signal has a positive correlation with the affinity and quantity of the corresponding autoantibody);

[0057] 9) Data extraction: The corresponding GAL file is opened, each array on the GAL file is aligned with the chip image as a whole, the automatic alignment button is clicked, the data is extracted, and the GPR is saved.

[0058] (2) Data processing:

[0059] The protein chip results are read, F532 Median refers to the median of the signal point foreground value under the 532 nm channel, and B532 Median refers to the median of the signal point background value under the 532 nm channel. In order to eliminate the non-uniformity of the signal caused by the inconsistency of the background values between different protein points in the same chip, the background normalization method is used for processing, that is, the relative expression amount of the autoantibody in 40 AFP-negative liver cancer serum samples, 56 AFP-positive liver cancer serum samples and 49 normal serum samples is displayed through the signal-noise ratio (SNR) = F532 Median / B532 Median. For any autoantibody, χ 2The differences in positive rates between groups were tested and compared, and the following screening conditions were further set: the positive rate of autoantibodies in the AFP-negative liver cancer group was higher than that in the normal control group (P<0.05), the positive rate of autoantibodies in the AFP-negative liver cancer group was higher than that in the AFP-positive liver cancer group (P<0.05), and the AUC ranking of healthy people for diagnosing AFP-negative liver cancer was in the top ten, and the intersection was screened out to meet the conditions of anti-tumor related antigen autoantibodies.

[0060] (3) Experimental results:

[0061] After screening, three anti-tumor related antigen autoantibodies were finally screened out, namely anti-tumor related antigen EGFR autoantibody, anti-tumor related antigen PHF6 autoantibody and anti-tumor related antigen PTEN autoantibody. The positive rates of the three anti-tumor related antigen autoantibodies in the AFP-negative liver cancer group, the AFP-positive liver cancer group and the normal control group, and the ROC curve of diagnosing AFP-negative liver cancer with healthy people as control are shown in Figure 1 , Figure 1 The bar chart in the first row of Figure 1 is the positive rate of the three anti-tumor related antigen autoantibodies in the AFP-positive liver cancer group, the AFP-negative liver cancer group and the normal control group.

[0062] It can be seen from Figure 1 that the three anti-tumor related antigen autoantibodies all meet the conditions that the positive rate in the AFP-negative liver cancer group is higher than that in the normal control group (P<0.05), the positive rate of autoantibodies in the AFP-negative liver cancer group is higher than that in the AFP-positive liver cancer group (P<0.05), and the difference is statistically significant. Among them, the AUC of anti-tumor related antigen EGFR autoantibody is 0.711 (0.601-0.821), the AUC of anti-tumor related antigen PHF6 autoantibody is 0.711 (0.598-0.823), and the AUC of anti-tumor related antigen PTEN autoantibody is 0.674 (0.556-0.791).

[0063] Example 2: Multicenter verification of AFP-negative liver cancer related antigens EGFR, PHF6 and PTEN autoantibodies

[0064] The expression levels of the three anti-tumor related antigen autoantibodies screened out in Example 1 were further detected in large sample population serum by indirect enzyme linked immunosorbent assay (ELISA).

[0065] 1. Experimental samples:

[0066] According to the principle of frequency matching by gender and age, 88 serum samples of AFP-negative liver cancer patients and 88 serum samples of normal controls were randomly selected from a clinical center of a third-grade hospital in Zhengzhou, 24 serum samples of AFP-negative liver cancer patients and 25 serum samples of normal controls were included under the same conditions from a clinical center of a third-grade hospital in Nanchang, and 25 serum samples of AFP-negative liver cancer patients and 13 serum samples of hepatitis B patients (unable to collect physical examinees at the same time) were included under the same conditions from a clinical center of a third-grade hospital in Beijing. The basic information of all samples is shown in Table 1. Among them, 103 serum samples of normal people (88 from Zhengzhou center and 25 from Nanchang center) were from healthy subjects. The inclusion criteria of healthy subjects were: no cardiovascular, respiratory, liver, kidney, gastrointestinal, endocrine, blood, mental, or nervous system diseases and history of the above diseases, no acute or chronic diseases, and no any evidence related to tumors. This study was approved by the Ethics Committee of Zhengzhou University, and all research objects had signed the informed consent form.

[0067] Table 1 Basic information of samples included in Zhengzhou, Nanchang and Beijing clinical centers

[0068]

[0069] 2. Experimental materials and reagents:

[0070] (1) Three tumor-related antigen proteins: EGFR protein, PHF6 protein, and PTEN protein. Among them, two recombinant proteins of EGFR and PHF6 were purchased from Wuhan Huamei Biological Engineering Co., Ltd.; the PTEN recombinant protein was obtained from the Key Laboratory of Tumor Epidemiology of Zhengzhou University. The recombinant plasmid of the protein stored in the laboratory was used to express and purify the protein by using the E. coli prokaryotic expression system and affinity chromatography method;

[0071] (2) 96-well enzyme-labeled plate (8 rows x 12 columns);

[0072] (3) Coating solution: 1000 mL coating solution containing Na2CO3 1.5 g, NaHCO3 2.9 g, sodium thiomersal 0.1 g, and 200 μL Proclin300;

[0073] (4) Blocking solution: 2% (w / v) bovine serum albumin (BSA) PBST solution;

[0074] (5) Serum sample diluent: 1% (w / v) BSA PBST buffer;

[0075] (6) Enzyme-labeled secondary antibody: horseradish peroxidase (HRP) labeled mouse anti-human immunoglobulin antibody (hereinafter referred to as HRP labeled mouse anti-human IgG antibody);

[0076] (7) Antibody diluent: PBST buffer containing 1% (w / v) BSA;

[0077] (8) Washing solution: PBST buffer;

[0078] (9) Color developing solution: The color developing solution is composed of color developing solution A and color developing solution B, wherein the color developing solution A is a 0.02% tetramethyl benzidine aqueous solution, and the color developing solution B is prepared by using deionized water to make 1 L of 37 g of sodium phosphate dibasic (Na2HPO4·12H2O), 9.2 g of citric acid and 8 mL of 0.75% hydrogen peroxide solution; when used, the color developing solution A and the color developing solution B are mixed uniformly at a ratio of 1:1 by volume, and are prepared on demand;

[0079] (10) Termination solution: 2M sulfuric acid.

[0080] 3. Experimental method:

[0081] (1) Preparation of three tumor-related antigen-coated enzyme-labeled plates

[0082] Prepare a tumor-related antigen EGFR-coated enzyme-labeled plate, a tumor-related antigen PHF6-coated enzyme-labeled plate, and a tumor-related antigen PTEN-coated enzyme-labeled plate, respectively.

[0083] Taking the preparation of the tumor-related antigen EGFR-coated enzyme-labeled plate as an example, the specific operation steps are as follows:

[0084] 1) Preparation of tumor-related antigen EGFR protein solution: The EGFR recombinant protein is configured into an EGFR protein solution with a concentration of 0.25 μg / mL using the coating solution.

[0085] 2) Coating of enzyme-labeled plate: The EGFR protein solution prepared in step 1) is added to each reaction well of the 96-well enzyme-labeled plate at a sample amount of 50 μL / well, and is coated at 4°C overnight, then the remaining coating solution is shaken off and dried.

[0086] 3) Blocking: The reaction wells of the 96-well enzyme-labeled plate after coating are added with blocking solution at a sample amount of 100 μL / well, and are blocked in a 37°C water bath for 2 h, then the blocking solution is removed, washed with washing solution (sample amount is 350 μL / well) for 3 times and dried, to obtain the tumor-related antigen EGFR-coated enzyme-labeled plate.

[0087] The operation steps for preparing the tumor-related antigen PHF6-coated enzyme plate and the tumor-related antigen PTEN-coated enzyme plate are basically the same as those for the tumor-related antigen EGFR-coated enzyme plate, except that: 1. The tumor-related antigens used in step 1) are different. When preparing the tumor-related antigen PHF6-coated enzyme plate, the tumor-related antigen used in step 1) is PHF6 recombinant protein; when preparing the tumor-related antigen PTEN-coated enzyme plate, the tumor-related antigen used in step 1) is PTEN recombinant protein. 2. When preparing the tumor-related antigen PTEN-coated enzyme plate, the PTEN recombinant protein is configured into a PTEN protein solution with a concentration of 0.125 μg / mL using the coating solution in step 1).

[0088] (2) Detection of the expression levels of three anti-tumor-related antigen autoantibodies in serum samples:

[0089] The same serum sample is used to detect the expression levels of anti-tumor-related antigen EGFR, PHF6, and PTEN autoantibodies in the serum sample by ELISA using the three tumor-related antigen-coated enzyme plates prepared above.

[0090] Taking the detection of the expression level of anti-tumor-related antigen EGFR autoantibody as an example, the specific operation steps are as follows:

[0091] 1) Serum sample incubation (primary antibody incubation):

[0092] The serum sample to be tested is diluted with serum diluent at a volume ratio of 1:100, and the diluted serum sample is added to the reaction wells of columns 1-11 of the EGFR recombinant protein-coated 96-well enzyme plate prepared in step (1) above, with a sample volume of 50 μL / well. The 1:100 diluted quality control serum is added to the first to sixth reaction wells of column 12 of the EGFR recombinant protein-coated 96-well enzyme plate, with a sample volume of 50 μL / well, and the quality control serum is used as a quality control for standardization between different enzyme plates. The antibody diluent without serum is added to the seventh to eighth reaction wells of column 12 of the EGFR recombinant protein-coated 96-well enzyme plate as a blank control, with a sample volume of 50 μL / well. Then, the 96-well enzyme plate is placed in a 37°C water bath for incubation for 1 h, and then the liquid in the reaction wells is discarded, washed with washing solution (with a sample volume of 350 μL / well) for 5 times, and tapped dry.

[0093] 2) Secondary antibody incubation:

[0094] HRP-labeled mouse anti-human IgG antibody was diluted with antibody diluent at a ratio of 1:10000 (v / v), and then the diluted HRP-labeled mouse anti-human IgG antibody was added to the corresponding reaction wells of the 96-well enzyme-labeled plate at a volume of 50 μL / well, and incubated at 37°C for 1 h, and then the liquid in the reaction wells was discarded, and the reaction wells were washed with washing solution (at a volume of 300 μL / well) for 5 times and patted dry.

[0095] 3) Color development and termination reaction:

[0096] Color developing solution A and color developing solution B were mixed at a ratio of 1:1, and then the mixed color developing solution was quickly added to the reaction wells of the 96-well enzyme-labeled plate at a volume of 100 μL / well, and color development was carried out at room temperature in the dark until the desired color was obtained, and then 50 μL of termination solution was added to each reaction well to terminate the color development reaction; within 10 min after termination, the absorbance OD 450 , OD 620 at 450 nm and 620 nm was measured using an enzyme-labeled instrument.

[0097] The specific operation steps for detecting the expression level of anti-tumor related antigen PHF6 and PTEN autoantibodies in serum samples are basically the same as those for detecting anti-tumor related antigen EGFR autoantibodies, except that in step 1), the enzyme-labeled plates used for detection are tumor related antigen PHF6 protein coated enzyme-labeled plates and tumor related antigen PTEN protein coated enzyme-labeled plates; in step 2), for the reaction wells coated with tumor related antigen PHF6, the HRP-labeled mouse anti-human IgG antibody is diluted at a ratio of 1:10000 (v / v); for the reaction wells coated with tumor related antigen PTEN, the HRP-labeled mouse anti-human IgG antibody is diluted at a ratio of 1:5000 (v / v).

[0098] 4, Data processing

[0099] The OD 450 -OD 620 values were taken as relative OD values, and the absorbance values (i.e. OD values) of the three anti-tumor related antigen autoantibodies were obtained by subtracting the absorbance values of the blank control serum samples, and the non-parametric test (Mann-Whitney U test) was used to compare the expression level differences of the three autoantibodies between the AFP-negative liver cancer group and the normal control group (or the chronic hepatitis B group).

[0100] 5, Experimental results

[0101] The relative titer scatter plots of anti-tumor related antigen EGFR, PHF6 and PTEN autoantibodies in the serum of the AFP-negative liver cancer group and the normal control group (or the chronic hepatitis B group) in the three clinical centers are as follows: Figure 2Figure 1 shows the ROC curves of the three anti-tumor antigen autoantibodies in the three clinical centers, where A is the result of the Zhengzhou clinical center, B is the result of the Nanchang clinical center, and C is the result of the Beijing clinical center; HCC represents the AFP-negative hepatocellular carcinoma group, HD represents the normal control group, and HBV represents the chronic hepatitis B group.

[0102] As shown in Figure 1, the anti-tumor antigen EGFR autoantibody, the anti-tumor antigen PHF6 autoantibody, and the anti-tumor antigen PTEN autoantibody all showed higher expression levels in the AFP-negative hepatocellular carcinoma group than in the normal control group (or the chronic hepatitis B group) in the three clinical centers. This indicates that these three anti-tumor antigen autoantibodies have good universality and robustness in multi-center verification and can be used as diagnostic markers for AFP-negative hepatocellular carcinoma. Figure 2 Example 3: Evaluation of the diagnostic ability of three anti-tumor antigen autoantibodies for AFP-negative hepatocellular carcinoma

[0103] Based on the expression levels of the anti-tumor antigen EGFR, PHF6, and PTEN autoantibodies in the serum samples from the three clinical centers in Example 2, the ROC curves for diagnosing AFP-negative hepatocellular carcinoma and normal controls were plotted using GraphPad Prism 8.0 for single anti-tumor antigen autoantibodies and combinations of multiple anti-tumor antigen autoantibodies, respectively. The diagnostic value of the three anti-tumor antigen autoantibodies for AFP-negative hepatocellular carcinoma was analyzed.

[0104] 1. The diagnostic ability of single anti-tumor antigen autoantibodies for distinguishing AFP-negative hepatocellular carcinoma from normal individuals:

[0105] Based on the expression levels of the anti-tumor antigen EGFR, PHF6, and PTEN autoantibodies in the serum samples from the three clinical centers in Example 2, the ROC curves for diagnosing AFP-negative hepatocellular carcinoma and normal controls were plotted using GraphPad Prism 8.0 for single anti-tumor antigen autoantibodies and combinations of multiple anti-tumor antigen autoantibodies, respectively. The diagnostic value of the three anti-tumor antigen autoantibodies for AFP-negative hepatocellular carcinoma was analyzed.

[0106] The ROC curves of the anti-tumor antigen EGFR autoantibody (denoted as anti-EGFR autoantibody), the anti-tumor antigen PHF6 autoantibody (denoted as anti-PHF6 autoantibody), and the anti-tumor antigen PTEN autoantibody (denoted as anti-PTEN autoantibody) for diagnosing AFP-negative hepatocellular carcinoma and normal individuals in the three clinical centers are shown in Figure 1. According to the ROC curve, the OD value with the largest Youden index was taken as the cutoff value, and the corresponding AUC, 95% confidence interval, sensitivity, and specificity were calculated.

[0107] Figure 3 The ROC curves of the anti-tumor antigen EGFR autoantibody (denoted as anti-EGFR autoantibody), the anti-tumor antigen PHF6 autoantibody (denoted as anti-PHF6 autoantibody), and the anti-tumor antigen PTEN autoantibody (denoted as anti-PTEN autoantibody) for diagnosing AFP-negative hepatocellular carcinoma and normal individuals in the three clinical centers are shown in Figure 1. According to the ROC curve, the OD value with the largest Youden index was taken as the cutoff value, and the corresponding AUC, 95% confidence interval, sensitivity, and specificity were calculated. ​

[0108] Depend on Figure 3 As can be seen, the AUC values ​​for anti-EGFR autoantibodies, anti-PHF6 autoantibodies, and anti-PTEN autoantibodies were 0.667, 0.605, and 0.745, respectively, at the Zhengzhou center; 0.799, 0.854, and 0.812, respectively, at the Nanchang center; and 0.746, 0.763, and 0.797, respectively, at the Beijing center. The AUC values ​​for these three autoantibodies were all above 0.5, with anti-PTEN autoantibodies showing the highest diagnostic value, with an AUC of 0.745 at the Zhengzhou center and a sensitivity and specificity of 78.2% and 66.1%, respectively. This suggests that all three autoantibodies have a certain reference value for diagnosing AFP-negative liver cancer. Due to the smaller sample size of the Nanchang and Beijing clinical centers, their 95% confidence intervals were larger, so the present invention primarily refers to the results from the Zhengzhou center.

[0109] 2. The ability of the combined diagnosis of two autoantibodies against tumor-associated antigens to differentiate AFP-negative HCC patients from normal subjects:

[0110] Using the expression levels of anti-EGFR autoantibodies and anti-PHF6 autoantibodies in serum samples from 112 AFP-negative HCC patients and 113 normal controls from the Zhengzhou and Nanchang centers in Example 2 as independent variables and whether or not a liver cancer event occurred as the dependent variable, logistic regression analysis was performed on the expression levels of anti-EGFR autoantibodies and anti-PHF6 autoantibodies in serum samples from the AFP-negative HCC group and the normal control group. A diagnostic model for distinguishing AFP-negative HCC patients from normal controls was constructed. The diagnostic model was: P1 (P1 = AFP-negative HCC, 2TAAbs) = 1 / [1 + exp(1.971-5.413×EGFR+0.232×PHF6)]. In this diagnostic model, exp represents an exponential function with the natural constant e as the base; P1 represents the predicted probability of the model; EGFR represents the OD value of the anti-EGFR autoantibody in the serum of the subject (measured by the absorbance value detected by the ELISA method described in Example 2); and PHF6 represents the OD value of the anti-PHF6 autoantibody in the serum of the subject. Then, the expression levels of anti-EGFR autoantibodies and anti-PHF6 autoantibodies in each serum sample are substituted into the diagnostic model to obtain the prediction probability (i.e., P1 value) of each serum sample. The ROC curve is drawn according to the prediction probability. Figure 4 The sensitivity and specificity were calculated with a predicted probability of 0.5 as the cutoff value.

[0111] Similarly, the serum samples of 112 AFP-negative HCC patients enrolled in Zhengzhou Center and Nanchang Center in Example 2 were taken as the AFP-negative HCC group, and the serum samples of 113 normal controls enrolled in Zhengzhou Center and Nanchang Center in Example 2 were taken as the normal control group. Logistic regression analysis was performed on the expression amounts of anti-EGFR autoantibody and anti-PTEN autoantibody in the serum samples of the AFP-negative HCC group and the normal control group, and a diagnostic model for distinguishing AFP-negative HCC patients from normal controls was constructed, which was P2 (P2 = AFP-negative HCC, 2TAAbs) = 1 / [1+exp(3.891-3.303×EGFR-9.366×PTEN)]. In the diagnostic model, exp represents the exponential function with the natural constant e as the base; P2 represents the prediction probability of the model, EGFR represents the OD value of anti-EGFR autoantibody in the serum of the subject (measured by the absorbance value detected by the ELISA method described in Example 2), and PTEN represents the OD value of anti-PTEN autoantibody in the serum of the subject. The expression amounts of anti-EGFR autoantibody and anti-PTEN autoantibody in each serum sample were substituted into the diagnostic model, and the prediction probability (i.e., the P2 value) of each serum sample was obtained. The ROC curve was plotted according to the prediction probability, and the ROC curve is shown in Figure 2. Figure 4 The sensitivity and specificity were calculated with the prediction probability equal to 0.5 as the cutoff value.

[0112] Similarly, the serum samples of 112 AFP-negative HCC patients enrolled in Zhengzhou Center and Nanchang Center in Example 2 were taken as the AFP-negative HCC group, and the serum samples of 113 normal controls enrolled in Zhengzhou Center and Nanchang Center in Example 2 were taken as the normal control group. Logistic regression analysis was performed on the expression amounts of anti-EGFR autoantibody and anti-PTEN autoantibody in the serum samples of the AFP-negative HCC group and the normal control group, and a diagnostic model for distinguishing AFP-negative HCC patients from normal controls was constructed, which was P2 (P2 = AFP-negative HCC, 2TAAbs) = 1 / [1+exp(3.891-3.303×EGFR-9.366×PTEN)]. In the diagnostic model, exp represents the exponential function with the natural constant e as the base; P2 represents the prediction probability of the model, EGFR represents the OD value of anti-EGFR autoantibody in the serum of the subject (measured by the absorbance value detected by the ELISA method described in Example 2), and PTEN represents the OD value of anti-PTEN autoantibody in the serum of the subject. The expression amounts of anti-EGFR autoantibody and anti-PTEN autoantibody in each serum sample were substituted into the diagnostic model, and the prediction probability (i.e., the P2 value) of each serum sample was obtained. The ROC curve was plotted according to the prediction probability, and the ROC curve is shown in Figure 2. Figure 4PHF6 & PTEN. The sensitivity and specificity were calculated with the cutoff value of the predicted probability equal to 0.5.

[0113] From Figure 4 It can be seen that the AUC of AFP-negative HCC diagnosed by EGFR and PHF6 combined diagnosis was 0.693, the cutoff value was 0.500, and the corresponding sensitivity was 55.6%, and the specificity was 70.2%. The AUC of AFP-negative HCC diagnosed by EGFR and PTEN combined diagnosis was 0.782, the cutoff value was 0.500, and the corresponding sensitivity was 66.1%, and the specificity was 71.0%. The AUC of AFP-negative HCC diagnosed by PHF6 and PTEN combined diagnosis was 0.773, the cutoff value was 0.500, and the corresponding sensitivity was 68.5%, and the specificity was 70.2%.

[0114] 3. The ability of three anti-tumor associated antigen autoantibodies combined diagnosis to distinguish AFP-negative HCC patients and normal persons:

[0115] The samples of Zhengzhou clinical center and Nanchang clinical center in Example 2 were combined to obtain 112 AFP-negative HCC patient serum samples and 113 normal control serum samples. The expression amounts of anti-EGFR autoantibody, anti-PHF6 autoantibody and anti-PTEN autoantibody in 112 AFP-negative HCC patient serum samples and 113 normal control serum samples were used as independent variables, and whether it was an AFP-negative HCC event was used as a dependent variable. Logistics regression analysis was performed on the expression amounts of anti-EGFR autoantibody, anti-PHF6 autoantibody and anti-PTEN autoantibody, and a diagnostic model for diagnosing and distinguishing AFP-negative HCC patients and normal persons was constructed. The AFP-negative HCC diagnosis model is: PRE(PRE = AFP-negative HCC, 3TAAbs) = 1 / [1+exp(3.323-5.873xEGFR+7.415xPHF6-12.453xPTEN)]. In the diagnostic model: exp represents the exponential function with natural constant e as the base; PRE represents the predicted probability of the model, EGFR represents the OD value of anti-EGFR autoantibody in the serum of the subject (measured by the ELISA method described in Example 2), PHF6 represents the OD value of anti-PHF6 autoantibody in the serum of the subject, and PTEN represents the OD value of anti-PTEN autoantibody in the serum of the subject.

[0116] The constructed diagnostic model was visually displayed by using a nomogram, as shown in Figure 5 Figure 5 ​It can be seen that the levels of three anti-tumor associated antigen autoantibodies are skewed distribution, and the OD values of each autoantibody are known. According to the figure, the probability of diagnosing liver cancer can be judged, as shown in the first row of the figure. The red dot on the Points (score) represents the OD value of the anti-EGFR autoantibody (third row), the OD value of the anti-PHF6 autoantibody (second row), and the OD value of the anti-PTEN autoantibody (fourth row). The score corresponding to the total score is shown as "Total points (fifth row)" is 127, and the corresponding prediction probability of diagnosing AFP-negative liver cancer is 0.669 (sixth row). The "mountain-shaped" graph in the figure is the probability density graph of the corresponding index part.

[0117] According to Figure 5 The obtained prediction probability is plotted as an ROC curve to evaluate the value of the diagnostic model in diagnosing and distinguishing AFP-negative liver cancer and normal people. The ROC curve is shown in part A of Figure 6 According to Figure 6 It can be seen from part A of that the area under the ROC curve of the combined diagnosis of anti-EGFR autoantibody, anti-PHF6 autoantibody and anti-PTEN autoantibody in distinguishing AFP-negative liver cancer and normal people is 0.802. When the prediction probability is equal to 0.5 as the cutoff value, the specificity is calculated to be 77.4% when the sensitivity reaches 70.2%. The calculation method of sensitivity and specificity is: sensitivity = true positive number / (true positive number + false negative number) * 100%, specificity = true negative number / (true negative number + false positive number) * 100%.

[0118] In order to facilitate comparison, the ROC curve AUC, sensitivity and specificity of the above single anti-tumor associated antigen autoantibody or multiple anti-tumor associated antigen autoantibodies in diagnosing and distinguishing AFP-negative liver cancer and normal controls are statistically analyzed, and the results are shown in Table 2. The ROC curve AUC of the single anti-tumor associated antigen autoantibody is adopted by the results of Zhengzhou Center.

[0119] Table 2 Evaluation results of three anti-tumor associated antigen autoantibodies in diagnosing and distinguishing AFP-negative liver cancer patients and normal people

[0120]

[0121] From Table 2, compared with single anti-tumor related antigen autoantibody, the AUC interval of ROC curve of EGFR autoantibody+PHF6 autoantibody combined diagnosis for distinguishing AFP-negative liver cancer patients from normal people is not obvious, even lower than PTEN autoantibody. But the AUC of ROC curve of EGFR autoantibody+PTEN autoantibody, PHF6 autoantibody+PTEN autoantibody combined diagnosis for distinguishing AFP-negative liver cancer patients from normal people is obviously higher than single anti-tumor related antigen autoantibody. When EGFR autoantibody+PHF6 autoantibody+PTEN autoantibody combined diagnosis for distinguishing AFP-negative liver cancer patients from normal people, the AUC of ROC curve reaches the maximum value 0.802. Moreover, when three anti-tumor related antigen autoantibodies combined diagnosis for AFP-negative liver cancer patients from normal people, the sensitivity is 70.2%, and the specificity of diagnosis reaches 77.4% at this time, which shows that the three anti-tumor related antigen autoantibodies combined diagnosis effect is the best. Therefore, the three anti-tumor related antigen autoantibody combined diagnosis model is preferred as the diagnosis model of AFP-negative liver cancer.

[0122] Example 4: The ability of three anti-tumor related antigen autoantibodies combined diagnosis for distinguishing AFP-negative liver cancer patients from chronic hepatitis B patients

[0123] The serum samples of 25 AFP-negative liver cancer patients in Beijing clinical center in Example 2 are taken as liver cancer group, and the serum samples of 13 chronic hepatitis B patients in Beijing clinical center in Example 2 are taken as chronic hepatitis B group. The OD values of anti-EGFR autoantibody, anti-PHF6 autoantibody and anti-PTEN autoantibody in 25 AFP-negative liver cancer patients and 13 chronic hepatitis B patients are substituted into the AFP-negative liver cancer diagnosis model PRE(PRE=AFP-negative HCC,3TAAbs)=1 / [1+exp(3.323-5.873xEGFR+7.415xPHF6-12.453xPTEN)] constructed in Example 3, so that the prediction probability of each serum sample can be obtained, and the ROC curve is drawn according to the prediction probability (as shown in Part B of Figure 6 , the value of three autoantibodies combined diagnosis for distinguishing AFP-negative liver cancer from chronic hepatitis B is verified.

[0124] From Part B of Figure 6 , the area under the ROC curve AUC of anti-EGFR autoantibody, anti-PHF6 autoantibody and anti-PTEN autoantibody combined diagnosis for distinguishing AFP-negative liver cancer from chronic hepatitis B in serum samples of Beijing clinical center is 0.730, the specificity is 69.2%, and the sensitivity is 56.0%. Therefore, the combination of anti-EGFR autoantibody, anti-PHF6 autoantibody and anti-PTEN autoantibody can be used for diagnosis and distinction of AFP-negative liver cancer and chronic hepatitis B.

[0125] Example 5: Evaluation of the ability of the three autoantibody joint diagnostic model to distinguish between liver cancer patients and normal persons in combination with AFP

[0126] 1. Experimental samples

[0127] The 308 liver cancer patients (liver cancer group) and 239 normal control serum samples included in this experiment were from the sample library of the Key Laboratory of Epidemiology of Tumors in Henan Province. See Table 3 for specific information. The 308 liver cancer patient serum samples were from patients diagnosed by pathology and not treated with any treatment, of which 137 were AFP-negative liver cancer patients. The 239 normal serum samples were from healthy subjects. The inclusion criteria for healthy subjects were: no cardiovascular, respiratory, liver, kidney, gastrointestinal, endocrine, blood, mental, or nervous system diseases or history of the above diseases; no acute or chronic diseases; and no evidence of any tumor. Moreover, there was no statistically significant difference between the 308 liver cancer patients and the 239 healthy subjects in terms of gender and age. This study was approved by the Ethics Committee of Zhengzhou University, and all research subjects had signed informed consent forms.

[0128] Table 3 Basic information of the included samples

[0129]

[0130] The absorbance values of the three autoantibodies in each experimental sample were detected using the ELISA method described in Example 2 of the present application. Then the absorbance values of the EGFR autoantibody, PHF6 autoantibody, and PTEN autoantibody in the 308 liver cancer patient serum samples (liver cancer group) and the 239 normal control serum samples (normal control group) were substituted into the AFP-negative liver cancer diagnostic model PRE (PRE = AFP-negative liver cancer, 3TAAbs) = 1 / [1+exp(3.323-5.873xEGFR+7.415xPHF6-12.453xPTEN)] constructed in Example 3 above, i.e. the prediction probability of AFP-negative liver cancer for each serum sample was obtained. Then the prediction probability PRE value of the above AFP-negative liver cancer diagnostic model and the AFP value were combined using Logistic regression to construct a full liver cancer prediction model, which was P (P = full liver cancer) = 1 / [1+exp(2.487-0.256xAFP-5.003xPRE)], where exp represents the exponential function with natural constant e as the base, AFP represents the concentration of alpha-fetoprotein in the serum, and PRE represents the prediction probability value using the AFP-negative liver cancer diagnostic model PRE. The ROC curve was plotted according to the prediction probability of full liver cancer (as shown in Figure 2), and the area under the ROC curve was 0.998, which indicated that the full liver cancer prediction model had a high prediction accuracy. Figure 7The diagnostic results of the AFP-negative liver cancer diagnosis model of the present application and AFP are shown in FIG. 2. For comparison, the diagnostic results of the AFP-negative liver cancer diagnosis model and AFP are statistically analyzed, and the results are shown in Table 4. Figure 7 Figure 7 The diagnostic results of the AFP-negative liver cancer diagnosis model and AFP are shown in FIG. 2. For comparison, the diagnostic results of the AFP-negative liver cancer diagnosis model and AFP are statistically analyzed, and the results are shown in Table 4.

[0131] Table 4 Evaluation of the AFP-negative liver cancer diagnosis model of the present application and AFP for diagnosing liver cancer

[0132]

[0133] Note: a,b,c is the Delong test result. If the superscripts of the two groups are the same, it means that there is no statistically significant difference between the two groups, otherwise, it is the opposite.

[0134] Figure 7 is the Delong test result. If the superscripts of the two groups are the same, it means that there is no statistically significant difference between the two groups, otherwise, it is the opposite.

[0135] is the Delong test result. If the superscripts of the two groups are the same, it means that there is no statistically significant difference between the two groups, otherwise, it is the opposite. Figure 7 ​As shown in Table 4, the area under the ROC curve for the AFP-negative liver cancer diagnostic model constructed using the three markers proposed in the present invention to distinguish between liver cancer patients and normal subjects is 0.820 (95% CI: 0.786-0.855). When the AFP-negative liver cancer diagnostic model proposed in the present invention is combined with AFP to distinguish between liver cancer patients and normal subjects, the area under the ROC curve can reach 0.926 (95% CI: 0.898-0.955). While maintaining a specificity of 90%, the sensitivity can be significantly improved to 86.4%, demonstrating significantly better diagnostic results than using either the AFP-negative liver cancer diagnostic model or AFP alone.

[0136] Therefore, the AFP-negative liver cancer diagnostic model proposed in the present invention has a high diagnostic value and can be used in combination with AFP for the diagnosis of all liver cancers, which can significantly improve the diagnostic value of liver cancer.

[0137] The above embodiments are specific implementation methods of the present invention, but the implementation methods of the present invention are not limited to the above embodiments. Any other combination, change, modification, substitution, and simplification that does not exceed the design concept of the present invention shall fall within the scope of protection of the present invention.

Claims

1. Use of a reagent for detecting and diagnosing a biomarker for AFP-negative liver cancer in the preparation of a product for diagnosing AFP-negative liver cancer; the biomarker is an anti-tumor-associated antigen PHF6 autoantibody, or a combination of an anti-tumor-associated antigen PHF6 autoantibody and an anti-tumor-associated antigen EGFR autoantibody, or a combination of an anti-tumor-associated antigen PHF6 autoantibody and an anti-tumor-associated antigen EGFR autoantibody and an anti-tumor-associated antigen PTEN autoantibody.

2. Use of reagents for detecting and diagnosing biomarkers of liver cancer in the preparation of products for liver cancer diagnosis; the biomarkers are a combination of anti-tumor-related antigen EGFR autoantibodies, anti-tumor-related antigen PHF6 autoantibodies, anti-tumor-related antigen PTEN autoantibodies, and AFP protein.

3. The use according to claim 1 or 2, characterized in that The reagent is a reagent for detecting the biomarker in a sample through enzyme-linked immunosorbent assay.

4. The use according to claim 1 or 2, characterized in that The reagent is a reagent for detecting the biomarker in a sample by protein chip or immunoblotting.

5. The use according to claim 1 or 2, characterized in that: The reagent is a reagent for detecting the biomarker in a sample through microfluidic immunoassay.

6. The use according to claim 3, characterized in that The sample is serum, plasma, interstitial fluid or urine; and the product is a protein chip.

7. The use according to claim 3, characterized in that The product is a test kit.

8. The use according to claim 3, characterized in that The product is a preparation.

9. The use according to claim 3, characterized in that The reagent is an antigen or antibody for detecting the biomarker.

10. The use according to claim 1, characterized in that The biomarker is a combination of anti-tumor-associated antigen EGFR autoantibody, anti-tumor-associated antigen PHF6 autoantibody, and anti-tumor-associated antigen PTEN autoantibody; When the product is used to diagnose AFP-negative liver cancer, the calculation formula for predicting the probability of AFP-negative liver cancer is: PRE = 1 / [1 + exp (3.323 - 5.873 × EGFR + 7.415 × PHF6 - 12.453 × PTEN)], where PRE represents the probability of predicting AFP-negative liver cancer, EGFR represents the expression level of anti-tumor-associated antigen EGFR autoantibodies, PHF6 represents the expression level of anti-tumor-associated antigen PHF6 autoantibodies, and PTEN represents the expression level of anti-tumor-associated antigen PTEN autoantibodies.

11. The use according to claim 2, characterized in that The biomarker is a combination of anti-tumor-associated antigen EGFR autoantibody, anti-tumor-associated antigen PHF6 autoantibody, anti-tumor-associated antigen PTEN autoantibody, and AFP protein; When the product is used to diagnose liver cancer, the formula for calculating the probability of predicting liver cancer is: P = 1 / [1 + exp (2.487 - 0.256 × AFP - 5.003 × PRE], where P represents the probability of predicting liver cancer, AFP represents the concentration of serum alpha-fetoprotein, PRE represents the probability of predicting AFP-negative liver cancer, and PRE = 1 / [1 + exp (3.323 - 5.873 × EGFR + 7.415 × PHF6 - 12.453 × PTEN)], EGFR represents the expression level of anti-tumor-associated antigen EGFR autoantibody, PHF6 represents the expression level of anti-tumor-associated antigen PHF6 autoantibody, and PTEN represents the expression level of anti-tumor-associated antigen PTEN autoantibody.

Images