Blood metabolism marker combination for nasopharynx cancer diagnosis or auxiliary diagnosis and application thereof

Abstract

The invention provides a blood metabolism marker combination for nasopharynx cancer diagnosis or auxiliary diagnosis and application thereof, and belongs to the technical field of in vitro diagnosis. By analyzing metabonomics data of blood samples of healthy people and nasopharyngeal carcinoma patients, a specific blood metabolism marker combination related to nasopharyngeal carcinoma is found, nasopharyngeal carcinoma can be accurately diagnosed, an innovative and efficient evaluation tool is provided for clinicians, the clinicians are helped to optimize treatment schemes in time, and the treatment efficiency of the clinicians is improved. Therefore, the survival rate and life quality of the patient are remarkably improved, and a brand new possibility is developed for precise medical treatment of nasopharynx cancer. According to the scheme, the blank of a nasopharynx cancer precise diagnosis and treatment system in the metabolism dimension is filled up in a breakthrough mode, and a diagnosis and treatment mode is promoted to step towards a new stage of molecule-metabolism double-engine driving from traditional anatomical image driving.

Description

Technical Field

[0001] The present invention belongs to the technical field of in vitro diagnosis, and in particular relates to a blood metabolic marker combination for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma and its application. Background Art

[0002] Nasopharyngeal carcinoma (NPC), a malignant tumor of the head and neck with a distinct geographical distribution, has attracted considerable attention for its epidemiological characteristics and pathogenic mechanisms. The biological pathogenesis of NPC involves multiple interactions involving persistent Epstein-Barr virus (EBV) infection, genetic susceptibility, dietary exposure to nitrosamines, and environmental factors such as smoking and alcohol consumption.

[0003] Although diagnostic technologies such as imaging omics and liquid biopsy, as well as treatment methods such as intensity-modulated radiotherapy and immune targeted therapy have made significant progress in recent years, due to the hidden anatomical location of the nasopharynx and the lack of specificity of early symptoms, in clinical practice, about three-quarters of patients have already progressed to the locally advanced stage (stage III-IV) when diagnosed. Even with comprehensive treatment, the 5-year survival rate of such patients is still difficult to break through the bottleneck of 80%, and advanced cases face severe challenges such as treatment resistance and distant metastasis. Given that early diagnosis of nasopharyngeal carcinoma is of great significance to improving patient survival and quality of life, there is an urgent need to explore early diagnostic markers that are both sensitive and specific, which is of extremely important practical significance. Summary of the Invention

[0004] The purpose of the present invention is to provide a blood metabolic marker combination for the diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma and its application. The present invention provides a non-invasive, non-invasive, and highly accurate blood metabolic marker combination that can be used for the diagnosis and screening of nasopharyngeal carcinoma.

[0005] The present invention provides a blood metabolic marker combination for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma, comprising: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5 and lysophosphatidylcholine 18:3.

[0006] Preferably, the blood metabolic marker combination includes: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, hemolytic phosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1 and phosphatidylinositol 38:6.

[0007] Preferably, the blood metabolic marker combination includes: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, hemolytic phosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1, phosphatidylinositol 38:6, phosphatidylcholine 32:0, ceramide d40:1 and phosphatidylethanolamine 36:2.

[0008] The present invention also provides the use of the blood metabolic marker combination described in the above scheme as a diagnostic marker in the preparation of products for diagnosing or assisting in the diagnosis of nasopharyngeal carcinoma.

[0009] Preferably, the product includes a reagent or kit for detecting the combination of blood metabolic markers described in the above scheme.

[0010] Preferably, the reagent or kit for detecting the combination of blood metabolic markers described in the above scheme includes a liquid chromatography-mass spectrometry detection reagent.

[0011] The present invention also provides a method for screening a combination of blood metabolic markers for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma, comprising the following steps:

[0012] Plasma samples were collected from healthy individuals and nasopharyngeal cancer patients;

[0013] Separation of the organic and aqueous phases from the plasma;

[0014] Liquid chromatography-mass spectrometry was used to detect metabolites in the aqueous phase and the organic phase, respectively, to obtain metabolome detection data;

[0015] Preprocessing the metabolomics detection data and identifying metabolites to obtain metabolomics data;

[0016] Lasso regression was used to screen out metabolic markers for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma from the metabolomics data.

[0017] Preferably, the criteria for metabolite identification are: the retention time difference is within 0.1 min, and the metabolite molecular weight error is less than 10 ppm.

[0018] The present invention also provides a method for constructing a diagnostic model for nasopharyngeal carcinoma, comprising the following steps:

[0019] Multivariate ROC curve analysis was performed on the metabolic markers screened by the screening method described in the above scheme.

[0020] Preferably, the multivariate ROC curve analysis includes: selecting 3 / 4 samples from the modeling group data as a training set, the remaining 1 / 4 samples as a test set, and using a support vector machine to randomly iterate 1000 times, and constructing a diagnostic model for nasopharyngeal carcinoma by statistically calculating the average value of the final model accuracy.

[0021] The present invention provides a blood metabolic marker combination for the diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma, comprising: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5 and lysophosphatidylcholine 18:3; or comprising: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, lysophosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1 and phosphatidylinositol 38:6; or comprising: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, lysophosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1, phosphatidylinositol 38:6, phosphatidylcholine 32:0, ceramide d40:1 and phosphatidylethanolamine 36:2. By analyzing the metabolomics data of blood samples from healthy people and nasopharyngeal carcinoma patients, this invention has found a combination of 4, 7, or 10 specific blood metabolic markers related to nasopharyngeal carcinoma, which can accurately diagnose nasopharyngeal carcinoma. It provides clinicians with an innovative and efficient assessment tool to help them optimize treatment plans in a timely manner, thereby significantly improving patients' survival rate and quality of life, and opening up new possibilities for precision medicine for nasopharyngeal carcinoma. The solution of this invention has made a breakthrough in filling the gap in the metabolic dimension of the precision diagnosis and treatment system for nasopharyngeal carcinoma, and has promoted the diagnosis and treatment model from the traditional "anatomical imaging driven" to a new stage of "molecular-metabolic dual engine driven". BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 The results of multivariate ROC curve analysis of 10 metabolic markers in the modeling group;

[0024] Figure 2 Figure 3 is the multivariate ROC curve analysis results of 10 metabolic markers in the validation group. DETAILED DESCRIPTION

[0025] The present invention provides a blood metabolic marker combination for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma, comprising: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5 and lysophosphatidylcholine 18:3.

[0026] As an embodiment, the blood metabolic marker combination consists of: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, and lysophosphatidylcholine 18:3. In the present invention, the combination of four metabolic markers has an AUC of 0.982 (sensitivity = 1, specificity = 0.917, accuracy = 0.953, precision = 0.905). The model performance indicators of this marker combination are all maintained at a high level, indicating that the model has stable discriminative ability under different marker combinations and exhibits high clinical application value.

[0027] As an embodiment, the blood metabolic marker combination includes: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, hemolytic phosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1 and phosphatidylinositol 38:6; further, the blood metabolic marker combination consists of: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, hemolytic phosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1 and phosphatidylinositol 38:6. In the present invention, the combination of 7 metabolic markers has an AUC of 0.989 (sensitivity = 0.947, specificity = 0.917, accuracy = 0.93, precision = 0.9). The model performance index of this marker combination is maintained at a high level, indicating that the model has stable discrimination ability under different marker combinations and shows high clinical application value.

[0028] As an embodiment, the blood metabolic marker combination includes: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, hemolytic phosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1, phosphatidylinositol 38:6, phosphatidylcholine 32:0, ceramide d40:1 and phosphatidylethanolamine 36:2; further, the blood metabolic marker combination consists of: β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5 (" 53" indicates that the fatty acid chain contains 53 carbon atoms, and "5" indicates that the fatty acid chain contains 5 double bonds), lysophosphatidylcholine 18:3 ("18" indicates that the fatty acid chain contains 18 carbon atoms, and "3" indicates that the fatty acid chain contains 3 double bonds), triglyceride 58:5 ("58" indicates that the fatty acid chain contains 58 carbon atoms, and "5" indicates that the fatty acid chain contains 5 double bonds), phosphatidylinositol 34:1 ("34" indicates that the total number of carbon atoms in the two fatty acid chains is 34, and "1" indicates that the total number of double bonds in the two fatty acid chains is 1), phosphatidylinositol 38:6 ("38" indicates that the total number of carbon atoms in the two fatty acid chains is 38, and "6" indicates that the total number of double bonds is 6), phosphatidylcholine 32:0 ("32" indicates that the total number of carbon atoms in the two fatty acid chains is 32, and "0" indicates that the two fatty acid chains do not contain double bonds), ceramide d40:1 and phosphatidylethanolamine 36:2 ("36" indicates The total number of carbon atoms in the two fatty acid chains is 36, and "2" indicates that the total number of double bonds in the two fatty acid chains is 2); the ceramide d40:1 is ceramide d40:1(d16:1 / 24:0) (d40:1 indicates that the sum of the total number of carbon atoms of sphingosine and fatty acids is 40, and the total number of double bonds is 1; d16:1 / 24:0 indicates that sphingosine has 16 carbon atoms and 1 double bond, and the fatty acid has 24 carbon atoms and 0 double bonds). In the present invention, the diagnostic model constructed based on the combination of the 10 blood metabolic markers had an AUC of 0.993 (sensitivity = 0.947, specificity = 1, accuracy = 0.977, precision = 1), and the threshold was set at 0.7049, which showed excellent discriminative ability in distinguishing different groups.

[0029] The present invention also provides the use of the blood metabolic marker combination described in the above scheme as a diagnostic marker in the preparation of products for diagnosing or assisting in the diagnosis of nasopharyngeal carcinoma.

[0030] The present invention adopts a non-invasive detection method, which only requires the collection of blood samples. It is safe, reliable and has high patient compliance. Secondly, the blood metabolic marker combination of the present invention has high sensitivity and high specificity, which can accurately capture subtle changes in metabolites, thereby achieving accurate prediction and evaluation. Third, it has high-throughput analysis capabilities and can detect multiple metabolites at the same time, which is suitable for efficient processing of large-scale clinical samples. Fourth, it is cost-effective, effectively reduces resource consumption, and significantly improves detection efficiency. Fifth, it has the advantage of early diagnosis and can accurately evaluate the early stages of diseases such as cancer. In addition, the application of the present invention can continuously track the progression of the disease and the treatment effect by dynamically monitoring the changing trends of metabolic markers, providing an important basis for timely adjustment of treatment strategies and reducing the risk of recurrence. Overall, the application of the present invention overcomes the invasiveness and limitations of traditional biopsy, realizes real-time monitoring of metabolite changes, and provides a key reference for clinicians to formulate and optimize treatment plans. It not only opens up new possibilities for disease prognosis prediction and health management, but is also expected to play a core role in future clinical practice, helping patients obtain better treatment effects and a higher quality of life.

[0031] As an embodiment, the product includes a reagent or kit for detecting the combination of blood metabolic markers described in the above scheme.

[0032] As an embodiment, the reagent or kit for detecting the combination of blood metabolic markers described in the above scheme includes a liquid chromatography-mass spectrometry detection reagent.

[0033] The present invention also provides a method for screening a combination of blood metabolic markers for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma, comprising the following steps:

[0034] Plasma samples were collected from healthy individuals and nasopharyngeal cancer patients;

[0035] Separation of the organic and aqueous phases from the plasma;

[0036] Liquid chromatography-mass spectrometry was used to detect metabolites in the aqueous phase and the organic phase, respectively, to obtain metabolome detection data;

[0037] Preprocessing the metabolomics detection data and identifying metabolites to obtain metabolomics data;

[0038] Lasso regression was used to screen out metabolic markers for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma from the metabolomics data.

[0039] In the present invention, the step of separating the organic phase and the aqueous phase from the plasma preferably includes: mixing the plasma with a solvent, extracting the mixture, and obtaining an extract; adding a mixture of methanol and water to the extract, centrifuging, and separating the upper organic phase from the lower aqueous phase. In a specific embodiment of the present invention, the volume ratio of the plasma to the solvent is 1:10; the solvent is a mixture of methyl tert-butyl ether and methanol; and the volume ratio of the methyl tert-butyl ether to methanol is 3:1. In a specific embodiment of the present invention, based on a volume of 100 μL of plasma, the volume of the methanol and water mixture added to the extract is 500 μL; the volume ratio of the methanol to water is 3:1; and the centrifugation speed is 12,700 rpm, and the time is 5 minutes.

[0040] As an embodiment, the organic phase is prepared using Waters ACQUITY BEH C8 1.7 μm 2.1 × 100 mm column, the aqueous phase was Waters ACQUITY Small molecule separation was performed using an HSS T3 1.8 μm 2.1 × 100 mm column. Liquid chromatography and mass spectrometry analyses were performed using an ACQUITY UPLC I-Class liquid chromatography system (Waters) and a Q-Exactive mass spectrometry system (Thermo Fisher Scientific).

[0041] As an embodiment, the mobile phase parameters of the organic phase test liquid are as follows: mobile phase A is an aqueous solution containing 0.1% acetic acid and 0.1% ammonium acetate; mobile phase B is a mixed solution of acetonitrile and isopropanol containing 0.1% acetic acid and 0.1% ammonium acetate (the volume ratio of acetonitrile to isopropanol is 7:3, the mass content of acetic acid is 0.1%, and the mass content of ammonium acetate is 0.1%), and the separation elution gradient is as follows: 0-12 minutes is 55%-89% mobile phase B, and 12-19.5 minutes is 100% mobile phase B.

[0042] As an embodiment, the mobile phase parameters of the aqueous phase test liquid are as follows: mobile phase A is an aqueous solution containing 0.1% formic acid, mobile phase B is an acetonitrile solution containing 0.1% formic acid, and the flow rate is set to 0.4 mL / min. The separation elution gradient is as follows: 0 min, 1% mobile phase B; 13 min, 70% mobile phase B; 13.01, 99% mobile phase B; 18 min, maintain 99% mobile phase B; 18.01 min, return to 1% mobile phase B; 22 min, maintain 1% mobile phase B. The sample injection volume is 3 μL, and the autosampler temperature is set to 10°C;

[0043] As an embodiment, the mass spectrometry parameters are as follows: Analysis of polar metabolites uses a full scan (Full scan) combined with data-dependent acquisition (DDA) to obtain spectral information of the primary mass spectrum (MS1) and the secondary mass spectrum (MS2). The mass spectrum range of the full scan is set to 100–1500 Da, while the scanning range of the secondary scan mode (Full MS / dd-MS2) is divided into three intervals: 100-300 Da, 300-710 Da, and 700-1500 Da. The mass spectrometer used in the experiment is an Orbitrap high-resolution mass spectrometer equipped with an electrospray ionization source (Electrospray Ionization, ESI), and data is collected in both positive and negative ionization modes. Specific parameters were set as follows: Automatic Gain Control (AGC) was set to 3E+6, Maximum ion injection time (IT) was 200 ms, and the full scan resolution was 70,000 FWHM (Full Width at Half Maximum) at 200 m / z. In Full MS / dd-MS2 mode, the MS resolution was adjusted to 17,500, the quadrupole isolation window width was 1.5 m / z, the AGC value was set to 1E+5, the maximum ion injection time was 50 ms, and the relative collision energy (HCD) was 30 eV. Furthermore, the ion spray voltage was 3,500 V in positive mode and 3,000 V in negative mode. The nebulizer gas pressure was 20 psi, the sheath gas temperature was 400°C, and the sheath gas flow rate was 10 L / min.

[0044] As an embodiment, preprocessing the metabolome detection data includes: extracting detection peaks from all mass spectra and removing background noise through baseline correction while retaining the original signal peaks; converting the original data into discrete center data; then, accurately matching the peaks in a single sample with the retention times in the chromatogram; in the data set, further removing isotope peaks to obtain the final mass spectrum matrix data; in order to reduce the differences in metabolite concentrations between different samples and make the data distribution more symmetrical, a normalization autoencoder (NormAE) is used to normalize the data.

[0045] As an embodiment, the metabolite identification includes: by performing software analysis on the raw data, spectral information of the compound parent ion and its secondary fragment ions can be obtained, including the mass-to-charge ratio (m / z) of the primary mass spectrum and the fragment characteristics of the secondary ions; by matching this information with the primary and secondary metabolite spectral data in the database, qualitative analysis of the metabolites can be achieved; commonly used metabolite databases include: HMDB (www.hmdb.ca), PubChem (https: / / pubchem.ncbi.nlm.nih.gov), MassBank (http: / / www.massbank.jp), North American MassBank (https: / / massbank.us) and LipidMaps (www.lipidmaps.org); metabolites preliminarily identified based on the relevant databases can be finally confirmed by the retention time, MS1 and MS2 mass spectral information when the standard is separated on the same chromatographic column and under the same mass spectrometry conditions.

[0046] As an embodiment, the standard for metabolite identification is: the retention time difference is within 0.1 min, and the metabolite molecular weight error is less than 10 ppm.

[0047] The present invention also provides a method for constructing a diagnostic model for nasopharyngeal carcinoma, comprising the following steps:

[0048] Multivariate ROC curve analysis was performed on the metabolic markers screened by the screening method described in the above scheme.

[0049] As an embodiment, the multivariate ROC curve analysis includes: selecting 3 / 4 samples from the modeling group data as a training set, the remaining 1 / 4 samples as a test set, and using a support vector machine to randomly iterate 1000 times, and constructing a diagnostic model for nasopharyngeal carcinoma by statistically calculating the average value of the final model accuracy.

[0050] To further illustrate the present invention, a blood metabolic marker combination for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma and its application provided by the present invention are described in detail below in conjunction with the accompanying drawings and examples, but they should not be understood as limiting the scope of protection of the present invention.

[0051] Example 1

[0052] 1. Subject conditions and sample collection

[0053] The inclusion and exclusion criteria for patients with nasopharyngeal carcinoma were as follows:

[0054] 1) Inclusion criteria:

[0055] (1) All subjects signed written informed consent before participating in the study;

[0056] (2) Males and females aged 18 years and above;

[0057] (3) Patients diagnosed with NPC by biopsy or postoperative pathology, or by a clinician based on comprehensive evaluation;

[0058] 2) Exclusion criteria:

[0059] (1) Pregnancy or lactation;

[0060] (2) Emergency or emergency treatment required;

[0061] (3) history of blood transfusion within 7 days before sampling;

[0062] (4) People who have received organ transplantation or non-autologous (allogeneic) bone marrow or stem cell transplantation;

[0063] (5) history of malignant tumor within 5 years or any anti-tumor treatment before sampling;

[0064] (6) Multiple primary malignant tumors are present simultaneously.

[0065] 3) Subjects' conditions:

[0066] This study included plasma samples from 121 healthy controls (HC) and 101 nasopharyngeal carcinoma (NPC) subjects. The modeling group included 95 plasma samples from the HC group and 77 plasma samples from the NPC group, while the validation group included 26 plasma samples from the HC group and 24 plasma samples from the NPC group (Table 1).

[0067] Table 1 Subject information

[0068]

[0069]

[0070] 2. Detection, identification and metabolic marker screening of small molecule metabolites in plasma samples

[0071] (1) Reagents:

[0072] Methanol, acetonitrile, water, acetic acid, isopropanol, and methyl tert-butyl ether of mass spectrometry grade, and formic acid and ammonium acetate of HPLC grade were purchased from Sigma-Aldrich, USA.

[0073] (2) Sample preparation:

[0074] Take 100 μL of plasma, add 1000 μL of a pre-cooled mixture of methyl tert-butyl ether and methanol (volume ratio 3:1), and vortex to mix to obtain a sample extract. Then, add 500 μL of a mixture of methanol and water (volume ratio 3:1) to the sample extract, sonicate, let stand, vortex, and centrifuge to separate the layers.

[0075] Organic phase: After sample separation, transfer 500 μL of the upper layer to a centrifuge tube as the organic phase. After drying the organic phase, add 200 μL of a mixture of acetonitrile and isopropanol (3:1 volume ratio) and incubate at room temperature for 15 minutes. After incubation, vortex to mix, sonicate for 5 minutes, and centrifuge at 12,000 rpm for 5 minutes at room temperature. Finally, transfer 180 μL of the supernatant to a 2 mL glass vial as the organic phase for LC-MS analysis.

[0076] Aqueous phase: After the sample is separated, remove 400 μL of the lower aqueous phase to a centrifuge tube and add 1100 μL of ice methanol to precipitate the protein. After the protein is completely precipitated, centrifuge and transfer 1000 μL of the supernatant to a new centrifuge tube and dry at room temperature overnight. After drying, add 200 μL of water to the centrifuge tube and incubate at room temperature for 15 minutes. After the incubation is completed, vortex to mix, ultrasonically treat for 5 minutes, and centrifuge at 12000 rpm for 5 minutes at room temperature. Finally, transfer 180 μL of the supernatant to a 2 mL glass injection vial as an aqueous phase for LC-MS detection;

[0077] (3) Small molecule metabolite detection:

[0078] The organic phase was prepared using Waters ACQUITY BEH C8 1.7 μm 2.1 × 100 mm column, the aqueous phase was Waters ACQUITY An HSS T3 1.8 μm 2.1 × 100 mm column was used for small molecule separation. Liquid chromatography and mass spectrometry analyses were performed using an ACQUITY UPLC I-Class liquid chromatography system (Waters) and a Q-Exactive mass spectrometry system (Thermo Fisher Scientific);

[0079] The mobile phase parameters of the organic phase to be tested are as follows:

[0080] Mobile phase A is an aqueous solution containing 0.1% acetic acid and 0.1% ammonium acetate; mobile phase B is a mixed solution of acetonitrile and isopropanol containing 0.1% acetic acid and 0.1% ammonium acetate (the volume ratio of acetonitrile to isopropanol is 7:3, the mass content of acetic acid is 0.1%, and the mass content of ammonium acetate is 0.1%). The separation elution gradient is as follows: 55%-89% mobile phase B from 0 to 12 minutes, and 100% mobile phase B from 12 to 19.5 minutes.

[0081] The mobile phase parameters of the aqueous solution to be tested are as follows:

[0082] Mobile phase A was an aqueous solution containing 0.1% formic acid, and mobile phase B was an acetonitrile solution containing 0.1% formic acid. The flow rate was set at 0.4 mL / min. The separation elution gradient was as follows: 0 min, 1% mobile phase B; 13 min, 70% mobile phase B; 13.01 min, 99% mobile phase B; 18 min, maintain 99% mobile phase B; 18.01 min, return to 1% mobile phase B; 22 min, maintain 1% mobile phase B. The sample injection volume was 3 μL, and the autosampler temperature was set to 10°C.

[0083] The mass spectrometry parameters are as follows:

[0084] Polar metabolites were analyzed using a full scan method combined with data-dependent acquisition (DDA) to acquire spectral information from both the primary mass spectrometer (MS1) and the secondary mass spectrometer (MS2). The full scan mass spectrum range was set to 100–1500 Da, while the secondary scan mode (Full MS / dd-MS2) scan range was divided into three intervals: 100–300 Da, 300–710 Da, and 700–1500 Da. The mass spectrometer used in the experiment was an Orbitrap high-resolution mass spectrometer equipped with an electrospray ionization (ESI) source, and data were acquired in both positive and negative ionization modes.

[0085] The specific parameters were set as follows: Automatic Gain Control (AGC) was set to 3E+6, the maximum ion injection time (Maximum IT) was 200ms, and the resolution of the first full scan was 70,000FWHM (FullWidth at HalfMaximum) (@200m / z). In the second scan mode (Full MS / dd-MS2), the resolution of the secondary mass spectrometer was adjusted to 17,500, the quadrupole isolation window width was 1.5m / z, the AGC value was set to 1E+5, the maximum ion injection time was 50ms, and the relative collision energy (Higher-energy Collisional Dissociation, HCD) was 30eV. In addition, the ion spray voltage was 3,500V in positive mode and 3,000V in negative mode; the nebulizer gas pressure was 20psi, the sheath gas temperature was 400℃, and the sheath gas flow rate was 10L / min;

[0086] (4) Metabolomics data processing

[0087] First, the detection peaks were extracted from all mass spectra, and background noise was removed by baseline correction while retaining the original signal peaks. The raw data were converted into discrete center data. Then, the peaks in individual samples were accurately matched with the retention times in the chromatogram. In the data set, the isotopic peaks were further removed to obtain the final mass spectrum matrix data. To reduce the differences in metabolite concentrations between different samples and make the data distribution more symmetrical, the normalization encoder (NormAE) was used to normalize the data.

[0088] (5) Identification of metabolites

[0089] By performing software analysis on the raw data, the spectral information of the compound parent ion and its secondary fragment ion can be obtained, including the mass-to-charge ratio (m / z) of the primary mass spectrum and the fragment characteristics of the secondary ion. By matching this information with the primary and secondary metabolite spectral data in the database, qualitative analysis of metabolites can be achieved. Commonly used metabolite databases include: HMDB (www.hmdb.ca), PubChem (https: / / pubchem.ncbi.nlm.nih.gov), MassBank (http: / / www.massbank.jp), North American MassBank (https: / / massbank.us) and LipidMaps (www.lipidmaps.org). The metabolites preliminarily identified based on the relevant database can be finally confirmed by the retention time, MS1 and MS2 mass spectrometry information when the standard is separated on the same chromatographic column and under the same mass spectrometry conditions. The criteria for metabolite identification are: the retention time difference is within 0.1min, and the metabolite molecular weight error is less than 10ppm;

[0090] (6) Metabolomics data analysis

[0091] 1) Screening of metabolic markers for differentiating patients with nasopharyngeal carcinoma

[0092] The metabolomics data of the HC and NPC groups were analyzed using the Least Absolute Shrinkage and Selection Operator (Lasso) regression feature screening method. Ultimately, 10 metabolites with significant differences between the two groups were screened out (Table 2). These metabolites can serve as key metabolic markers to distinguish between the HC and NPC groups.

[0093] Table 2 Ten metabolic markers used to distinguish HC and NPC groups

[0094] Serial number Metabolic Markers-English Metabolic Markers-Chinese 1 beta-Hydroxycapricacid β-Hydroxydecanoic acid 2 Pyrophosphate pyrophosphate 3 TAG53:5 Triglycerides 53:5 4 LysoPC18:3 Lysophosphatidylcholine 18:3 5 TAG58:5 Triglycerides 58:5 6 PI 34:1 Phosphatidylinositol 34:1 7 PI 38:6 Phosphatidylinositol 38:6 8 PC32:0 Phosphatidylcholine 32:0 9 Cerd40:1(d16:1 / 24:0) Ceramide d40:1 (d16:1 / 24:0) 10 PE36:2 Phosphatidylethanolamine 36:2

[0095] 2) Build a diagnostic model for differentiating nasopharyngeal carcinoma patients

[0096] To evaluate the discriminative ability of the selected metabolite markers in differentiating NPC groups, a multivariate receiver operating characteristic (ROC) analysis was performed on these 10 metabolite markers. Specifically, 3 / 4 of the samples from the modeling group were randomly selected as the training set to build and optimize the machine learning classification model; the remaining 1 / 4 of the samples were used as the test set to evaluate the model's classification performance. In addition, a support vector machine (SVM) algorithm was employed, and 1000 randomized recurrent cross-validation cycles were performed to further enhance the model's stability and reliability. Ultimately, by calculating the average model accuracy, a diagnostic model was successfully constructed that effectively distinguished HC from NPC groups.

[0097] The ROC curve is a tool used to analyze the relationship between model sensitivity (Sensitivity) and specificity (Specificity). It is plotted with sensitivity as the vertical axis (Y-axis) and 1-specificity as the horizontal axis (X-axis). Model performance is usually evaluated by the area under the curve (AUC): when the AUC value is greater than 0.5, the closer the AUC is to 1, the better the model performance and the stronger the diagnostic ability; conversely, if the AUC is lower than 0.5, it indicates that the model's diagnostic ability is weak. In addition, the ROC classification diagnostic model system not only covers the basic ROC curve and AUC indicators, but also involves multiple key parameters such as sensitivity, specificity, accuracy, and precision.

[0098] The sensitivity is:

[0099]

[0100] Specificity is:

[0101]

[0102] The accuracy is:

[0103]

[0104] The accuracy is:

[0105]

[0106] in,

[0107] TP (True Positive): True positive, the number of samples that are actually positive examples that are correctly predicted as positive examples;

[0108] TN (Ture Negative): True negative, the number of samples that are actually negative examples but are correctly predicted as negative examples;

[0109] FP (False Positive): False positive, the number of samples that are actually negative examples but are mistakenly predicted as positive examples;

[0110] FN (False Negative): False negatives, the number of samples that are actually positive examples but are mistakenly predicted as negative examples;

[0111] The results are as follows Figure 1 As shown, the diagnostic model constructed based on the combination of 10 metabolic markers had an AUC of 0.993 (sensitivity = 0.947, specificity = 1, accuracy = 0.977, precision = 1), which showed that the model had excellent discriminative ability in distinguishing different groups.

[0112] In addition, in the modeling group, the diagnostic efficacy of different metabolite marker combinations was analyzed by ROC curve analysis, and the following two metabolite marker combinations were used: 1) a combination of 7 metabolite markers: including β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, hemolytic phosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1, and phosphatidylinositol 38:6; 2) a combination of 4 metabolite markers: including β-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, and hemolytic phosphatidylcholine 18:3.

[0113] The results showed that the diagnostic models constructed based on different metabolite marker combinations all showed high efficacy. Among them, the AUC for the combination of seven metabolite markers was 0.989 (sensitivity = 0.947, specificity = 0.917, accuracy = 0.93, precision = 0.9); and the AUC for the combination of four metabolite markers was 0.982 (sensitivity = 1, specificity = 0.917, accuracy = 0.953, precision = 0.905). The model efficacy indicators of these two marker combinations remained at a high level, indicating that the model has stable discriminative ability under different marker combinations and exhibits high clinical application value.

[0114] 3) Validation of diagnostic models for distinguishing HC from NPC groups

[0115] In order to further verify the effectiveness of the diagnostic model for distinguishing nasopharyngeal carcinoma based on the modeling group data, the validation group data was used to test the above model. Specifically, the independent validation performance of the model on unknown data sets other than the modeling group data set was evaluated through multivariate ROC curve analysis. The validation group samples were substituted into the diagnostic model constructed by the modeling group, and the corresponding probability values ​​(Probability) were calculated based on the detection data of 10 key metabolite markers in each sample. The model will output the corresponding probability value, and use the probability value of each sample as the threshold to generate a set of confusion matrices (including true positive, true negative, false positive and false negative). Based on the confusion matrix, sensitivity and specificity can be further calculated. In the ROC analysis diagram with sensitivity (sensitivity) as the vertical coordinate and 1-specificity (1-specificity) as the horizontal coordinate, each threshold corresponds to a point. Similarly, when the probability value of each sample is used as the threshold, multiple different points can be obtained in the ROC analysis diagram. Connecting these points can draw the ROC curve (such as Figure 2 (As shown). On the ROC curve, the point with the best sensitivity and specificity was selected, corresponding to a threshold of 0.7049. Calculated using the same method, the thresholds for the diagnostic models of the seven-metabolite combination and the four-metabolite combination were 0.4458 and 0.4035, respectively.

[0116] As shown in the confusion matrix results in Table 3, in the diagnostic model constructed based on the above 10 metabolite markers, when the threshold was set to 0.7049, 23 of the 24 HC group samples were accurately identified, and 1 was misclassified as a NPC group sample; in the NPC group, all 26 samples were correctly identified. Based on the confusion matrix data, the ROC analysis results of the model in the validation group were further calculated, as shown below. Figure 2 The results showed that AUC = 0.989 (sensitivity = 0.958, specificity = 1, accuracy = 0.98, precision = 1). The above results indicate that the diagnostic model constructed for distinguishing HC and NPC groups also exhibited excellent discriminant efficacy in the validation group.

[0117] Table 3 Confusion matrix of diagnostic models used to distinguish HC and NPC groups

[0118] NPC HC 24 NPC group 23(TP) 1(FN) 26 cases in HC group 0(FP) 26(TN)

[0119] In addition, multivariate receiver operating characteristic (ROC) curve analysis was performed in the validation cohort for the diagnostic models based on different metabolite marker combinations. The results showed that the AUC value for the seven-marker combination in the validation cohort was 0.979 (sensitivity = 0.875, specificity = 0.962, accuracy = 0.92, precision = 0.955); and the AUC value for the four-marker combination was 0.959 (sensitivity = 0.833, specificity = 0.923, accuracy = 0.88, precision = 0.909). These results demonstrate that the constructed diagnostic model for distinguishing HC from NPC also exhibited good discriminative performance in the validation cohort.

[0120] In summary, this study, based on plasma samples, applied advanced metabolomics techniques to accurately detect small molecule metabolites in plasma, successfully identifying a panel of specific metabolite markers that effectively differentiated healthy subjects from those with NPC. Furthermore, using machine learning algorithms, we constructed an efficient and accurate diagnostic model. This model effectively identified patients with NPC, providing valuable insights into prognosis and personalized treatment.

[0121] Although the above embodiment provides a detailed description of the present invention, it is only a part of the embodiments of the present invention, not all of the embodiments. People can also obtain other embodiments based on this embodiment without creativity, and these embodiments all fall within the scope of protection of the present invention.

Claims

1. A blood metabolic marker combination for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma, characterized in that: include: Beta-hydroxydecanoic acid, pyrophosphate, triglycerides 53:5, and lysophosphatidylcholine 18:

3.

2. The blood metabolic marker combination according to claim 1, characterized in that: include: Beta-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, lysophosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1, and phosphatidylinositol 38:

6.

3. The blood metabolic marker combination according to claim 1, characterized in that: include: Beta-hydroxydecanoic acid, pyrophosphate, triglyceride 53:5, lysophosphatidylcholine 18:3, triglyceride 58:5, phosphatidylinositol 34:1, phosphatidylinositol 38:6, phosphatidylcholine 32:0, ceramide d40:1, and phosphatidylethanolamine 36:

2.

4. Use of the blood metabolic marker combination according to any one of claims 1 to 3 as a diagnostic marker in the preparation of a product for the diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma.

5. The use according to claim 4, characterized in that The product includes a reagent or a kit for detecting the blood metabolic marker combination according to any one of claims 1 to 3.

6. The use according to claim 5, characterized in that The reagent or kit for detecting the blood metabolic marker combination according to any one of claims 1 to 3 comprises a liquid chromatography-mass spectrometry detection reagent.

7. A method for screening a combination of blood metabolic markers for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma, characterized in that: The following steps are involved: Plasma samples were collected from healthy individuals and nasopharyngeal cancer patients; Separation of the organic and aqueous phases from the plasma; Liquid chromatography-mass spectrometry was used to detect metabolites in the aqueous phase and the organic phase, respectively, to obtain metabolome detection data; Preprocessing the metabolomics detection data and identifying metabolites to obtain metabolomics data; Lasso regression was used to screen out metabolic markers for diagnosis or auxiliary diagnosis of nasopharyngeal carcinoma from the metabolomics data.

8. The screening method according to claim 7, characterized in that The criteria for metabolite identification are: retention time difference within 0.1 min and metabolite molecular weight error less than 10 ppm.

9. A method for constructing a diagnostic model for nasopharyngeal carcinoma, characterized in that: The following steps are involved: Multivariate ROC curve analysis was performed on the metabolic markers screened by the screening method according to claim 7 or 8.

10. The method according to claim 9, characterized in that The multivariate ROC curve analysis includes: selecting 3 / 4 samples from the modeling group data as a training set, and the remaining 1 / 4 samples as a test set; using a support vector machine to randomly iterate 1000 times; and constructing a diagnostic model for nasopharyngeal carcinoma by calculating the average value of the final model accuracy.

Images