Prediction or diagnosis system for identifying non-Hodgkin lymphoma and application

Abstract

The invention provides a prediction or diagnosis system for identifying non-Hodgkin lymphoma and application, and belongs to the technical field of medical diagnosis. The prediction or diagnosis system comprises a Raman spectrum detection module of a to-be-detected sample and a Raman spectrum detection module of a negative control, the Raman spectrum detection module of the sample to be detected and the Raman spectrum detection module of the negative control take Raman spectrum data for detecting serum as indexes. Research finds that healthy people and non-Hodgkin lymphoma patients are distinguished through Raman spectrum data, non-Hodgkin lymphoma subtypes can be diagnosed through Raman data in combination with multivariate analysis, non-Hodgkin lymphoma and the non-Hodgkin lymphoma subtypes can be rapidly and noninvasively predicted or diagnosed, and the non-Hodgkin lymphoma subtypes and the non-Hodgkin lymphoma subtypes can be rapidly and noninvasively predicted or diagnosed. More choices are provided for rapid detection of non-Hodgkin lymphoma and subtypes thereof, and useful guidance is provided for application to treatment of DLBCL and CLL.

Description

Technical Field

[0001] The present invention belongs to the field of medical diagnosis technology, and in particular relates to a prediction or diagnosis system and application for identifying non-Hodgkin's lymphoma. Background Art

[0002] Lymphoma is a malignant tumor originating from lymph nodes or lymphatic tissues. It can be divided into two categories according to pathological and clinical characteristics: Hodgkin's lymphoma (HL) and non-Hodgkin's lymphoma (NHL). NHL is the malignant tumor with the fastest growing incidence rate and has the highest incidence rate in the world. Therefore, accurate and rapid prediction or diagnosis of lymphoma, especially non-Hodgkin's lymphoma, is of great significance for better and more effective treatment plans.

[0003] Among them, diffuse large B-cell lymphoma (DLBCL) and chronic lymphocytic leukemia (CLL) are two different subtypes of non-Hodgkin's lymphoma. DLBCL is the most common type of NHL, originating from mature B cells, characterized by rapid growth, usually manifested as lymphadenopathy, and may also invade other organs. In contrast, CLL mainly affects the elderly population, usually progresses slowly, and may be asymptomatic in the early stages. CLL originates from the clonal expansion of abnormal B cells that co-express B cell markers such as CD19 and CD20, and is characterized by the accumulation of these B cells in the blood, bone marrow, and lymph nodes. CLL is an indolent lymphoma that usually does not have a very aggressive course and has a relatively long overall survival. No significant improvement is observed with early treatment, so conservative observation therapy is usually adopted.

[0004] Although both DLBCL and CLL involve abnormal growth of B cells, there is no direct causal relationship between them. They are two diseases that develop independently, each with its own risk factors, genetic background and clinical manifestations. However, in some cases, one disease may transform into another. For example, some CLL patients may develop Richter transformation, a rare and more aggressive form of DLBCL. Richter transformation usually manifests as rapidly progressive DLBCL, whose biological behavior and treatment response are similar to those of primary DLBCL. The cause of Richter transformation is not fully understood, but may be related to genetic variations, immune escape and microenvironmental factors. Therefore, the development of a diagnostic marker that helps to identify DLBCL and CLL for the prevention, prediction, diagnosis of non-Hodgkin's lymphoma or differentiation between DLBCL and CLL can further understand the differences and transformation phenomena between the two diseases, which is of great significance for optimizing treatment options and improving patient prognosis.

[0005] In the field of scientific research, Raman spectroscopy has been used as a rapid, non-invasive and label-free method to analyze and classify cells with location specificity. The advantages of using Raman spectroscopy for diagnosis have been demonstrated: it can potentially replace traditional diagnostic methods based on visual inspection of histological samples. The technology has successfully demonstrated early detection of cancer, differentiation of tumor grades, and provides cellular information from different living cells. However, there is no research and application of Raman spectroscopy for serum analysis of non-Hodgkin's lymphoma, nor has there been any report on the identification of serological indicators for DLBCL and CLL. Summary of the invention

[0006] In view of this, the object of the present invention is to provide a prediction or diagnosis system and application for identifying non-Hodgkin's lymphoma.

[0007] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0008] The present invention provides a prediction or diagnosis system for identifying non-Hodgkin's lymphoma, the prediction or diagnosis system comprising a Raman spectrum detection module for a sample to be tested and a Raman spectrum detection module for a negative control; the Raman spectrum detection module for the sample to be tested and the Raman spectrum detection module for the negative control both use Raman spectrum data of serum as an indicator; the Raman spectrum data are Raman peaks 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm representing proteins -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 intensity level.

[0009] Preferably, the negative control is healthy human serum.

[0010] Preferably, the prediction or diagnosis system further comprises a conclusion output module, and when the Raman spectrum data of the negative control is significantly higher than the Raman spectrum data of the sample to be tested, it is predicted or diagnosed as non-Hodgkin's lymphoma.

[0011] Preferably, the non-Hodgkin's lymphoma is diffuse large B-cell lymphoma and chronic lymphocytic leukemia.

[0012] The present invention provides a method for detecting serum proteins based on Raman spectroscopy, wherein the Raman peaks 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 The application of the intensity level as a biomarker in the preparation of a product for predicting or diagnosing non-Hodgkin's lymphoma.

[0013] Preferably, the biomarkers in the serum of the test sample are detected and compared with the peak intensity values ​​of the negative control to predict or diagnose non-Hodgkin's lymphoma;

[0014] Preferably, the non-Hodgkin's lymphoma is diffuse large B-cell lymphoma and chronic lymphocytic leukemia.

[0015] The present invention also provides a method for detecting serum proteins based on Raman spectroscopy, wherein the Raman peaks 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 The application of the intensity level as a biomarker in the preparation of a predictive or diagnostic non-Hodgkin's lymphoma typing product.

[0016] Preferably, the non-Hodgkin's lymphoma is classified as diffuse large B-cell lymphoma or chronic lymphocytic leukemia.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] The present invention provides a prediction or diagnosis system and application for identifying non-Hodgkin's lymphoma. The present invention has found that healthy people and non-Hodgkin's lymphoma patients can be distinguished by Raman spectral data, and non-Hodgkin's lymphoma subtypes can be diagnosed by combining Raman data with multivariate analysis. The present invention can quickly and non-invasively predict or diagnose non-Hodgkin's lymphoma and its subtypes, providing more options for rapid detection of non-Hodgkin's lymphoma and its subtypes, and also providing useful guidance for the treatment of DLBCL and CLL. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The average serum Raman spectra of Control, CLL and DLBCL (from bottom to top);

[0020] Figure 2 A is the arrangement diagram of the OPLS-DA model for identification of control, CLL and DLBCL groups; B is the cluster analysis diagram of the OPLS-DA model for identification of control, CLL and DLBCL; C is the ROC curve of the OPLS-DA model for identification of control, CLL and DLBCL groups, AUC(control)=1, AUC(CLL)=1, AUC(DLBCL)=1; D is the arrangement diagram of the OPLS-DA model for identification of control and CLL groups; E is the cluster analysis diagram of the OPLS-DA model for identification of control and CLL groups; F is the ROC curve of the OPLS-DA model for identification of control and CLL groups, AUC(control )=1, AUC(CLL)=1; G is the arrangement diagram of the OPLS-DA model for identification of control and DLBCL groups; H is the cluster analysis diagram of the OPLS-DA model for identification of control and DLBCL groups; I is the ROC curve of the OPLS-DA model for identification of control and DLBCL groups, AUC(control)=1, AUC(DLBCL)=1; J is the arrangement diagram of the OPLS-DA model for identification of CLL and DLBCL groups; K is the cluster analysis diagram of the OPLS-DA model for identification of CLL and DLBCL groups; L is the ROC curve of the OPLS-DA model for identification of CLL and DLBCL groups, AUC(CLL)=1, AUC(DLBCL)=1;

[0021] Figure 3: OPLS-DA identification of serum samples of Control, CLL and DLBCL; A is the identification score map of Control, CLL and DLBCL through OPLS-DA drawn by Hotelling's 95% confidence ellipse; B is the identification load line map of Control, CLL and DLBCL; C is the identification V+S map of Control, CLL and DLBCL; D is the identification score map of Control and CLL through OPLS-DA drawn by Hotelling's 95% confidence ellipse; E is the identification load line map of Control and CLL; F is the identification V+S map of Control and CLL; G is the identification score map of Control and DLBCL through OPLS-DA drawn by Hotelling's 95% confidence ellipse; H is the identification load line map of Control and DLBCL; I is the identification V+S map of Control and DLBCL; J is the identification score map of CLL and DLBCL through OPLS-DA drawn by Hotelling's 95% confidence ellipse; K is the identification load line map of CLL and DLBCL; L is the identification V+S map of CLL and DLBCL;

[0022] Figure 4 A in the figure shows the statistical results of the Raman peaks of nucleic acids, lipids, carotenoids and collagen, which are potential biomarkers of control vs CLL vs DLBCL; B shows the statistical results of the Raman peaks of proteins and carbohydrates, which are potential biomarkers of control vs CLL vs DLBCL;

[0023] Figure 5A is a permutation diagram of the OPLS-DA model for distinguishing CLL and DLBCL groups based on biochemical indicator data; B is a cluster analysis diagram of the OPLS-DA model for distinguishing CLL and DLBCL groups based on biochemical indicator data; C is a ROC curve of the OPLS-DA model for distinguishing CLL and DLBCL groups based on biochemical indicator data, AUC (CLL) = 1, AUC (DLBCL) = 1; D is a score diagram of the CLL and DLBCL groups distinguished by OPLS-DA based on biochemical indicator data, drawn with Hotelling's 95% confidence ellipse; E is a load diagram of the CLL and DLBCL groups distinguished based on biochemical indicator data line graph; F is the V+S diagram for distinguishing CLL and DLBCL groups based on biochemical index data; G is the ALB comparison result of CLL and DLBCL groups; H is the TG comparison result of CLL and DLBCL groups; I is the HDL comparison result of CLL and DLBCL groups; J is the GGT comparison result of CLL and DLBCL groups; K is the ALP comparison result of CLL and DLBCL groups; L is the ALT comparison result of CLL and DLBCL groups; M is the TT comparison result of CLL and DLBCL groups; N is the WBC comparison result of CLL and DLBCL groups; O is the NEUT% comparison result of CLL and DLBCL groups. DETAILED DESCRIPTION

[0024] The present invention provides a prediction or diagnosis system for identifying non-Hodgkin's lymphoma, the prediction or diagnosis system comprising a Raman spectrum detection module for a sample to be tested and a Raman spectrum detection module for a negative control; the Raman spectrum detection module for the sample to be tested and the Raman spectrum detection module for the negative control both use Raman spectrum data of serum as an indicator; the Raman spectrum data are Raman peaks 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm representing proteins -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 intensity level.

[0025] In the above prediction or diagnosis system, the negative control is healthy human serum. The prediction or diagnosis system also includes a data processing device, which includes a data input module, a data recording module, a data comparison module and a conclusion output module. The data input module is configured to input the Raman spectrum data of the sample to be tested and the negative control; the data recording module is configured to store the Raman spectrum data of the sample to be tested and the negative control; the data comparison module is configured to receive the Raman spectrum data of the sample to be tested and the negative control sent by the data input module, and then compare the Raman spectrum data. The conclusion output module receives the comparison result sent by the data comparison module and judges the comparison result according to a predetermined judgment condition. The judgment condition is that when the Raman spectrum data of the negative control is significantly higher than the Raman spectrum data of the sample to be tested, it is predicted or diagnosed as non-Hodgkin's lymphoma, and when the Raman spectrum data of the negative control is not significantly higher than the Raman spectrum data of the sample to be tested, it is not predicted or diagnosed as non-Hodgkin's lymphoma.

[0026] The present invention found that the loading plot of the three models of Control vs CLL, Control vs DLBCL, and CLL vs DLBCL showed that the peak positions of the representative proteins of the control group (621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1 ), peaks representing nucleic acids (726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 ), peaks representing carotenoids (957, 1155, 1162 cm -1 ), representing sugars (842cm -1 ) and collagen (1345cm -1 ) and the peaks representing lipids (957, 1078, 1119, 1285, 1299, 1437, 1443, 1446 cm -1 ) was significantly higher than that of the CLL group and the DLBCL group. Based on this, the non-Hodgkin's lymphoma is diffuse large B-cell lymphoma and chronic lymphocytic leukemia.

[0027] The present invention provides a method for detecting serum proteins based on Raman spectroscopy, wherein the Raman peaks 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1, Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 The application of the intensity level as a biomarker in the preparation of a product for predicting or diagnosing non-Hodgkin's lymphoma.

[0028] In the present invention, the biomarkers in the serum of the sample to be tested are detected and compared with the peak intensity values ​​of the negative control to predict or diagnose non-Hodgkin's lymphoma, and the prediction or diagnosis is made when the Raman spectrum data in the serum of the sample to be tested is significantly higher than that of the negative control. The product includes a kit or a reagent.

[0029] The present invention also provides a method for detecting serum proteins based on Raman spectroscopy, wherein the Raman peaks 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 The application of the intensity level as a biomarker in the preparation of a predictive or diagnostic non-Hodgkin's lymphoma typing product.

[0030] In the above-mentioned non-Hodgkin's lymphoma typing, the present invention uses Raman spectroscopy to detect the Raman peaks of serum representative proteins 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1, Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 The intensity level is compared by multivariate statistical methods, and further, the multivariate statistical method is orthogonal-partial least squares discriminant analysis (OPLS-DA), so as to achieve the purpose of non-Hodgkin's lymphoma typing. As a preferred embodiment, the non-Hodgkin's lymphoma is classified as diffuse large B-cell lymphoma or chronic lymphocytic leukemia. The product includes a kit or a reagent. The present invention utilizes the above-mentioned Raman spectral data through OPLS-DA to accurately diagnose diffuse large B-cell lymphoma or chronic lymphocytic leukemia, and can also provide new ideas and strategies for the treatment of the disease, which is ultimately expected to improve the treatment effect and quality of life of patients.

[0031] The Raman spectrum detection of the patient serum of the present invention can obtain statistical differences, indicating that the Raman detection of serum is more sensitive, can capture subtle differences that cannot be captured by peripheral blood detection, and is more helpful in screening key genes.

[0032] In the present invention, unless otherwise specified, all raw material components are commercially available products well known to those skilled in the art.

[0033] The technical solutions provided by the present invention are described in detail below in conjunction with the embodiments, but they should not be construed as limiting the protection scope of the present invention.

[0034] Example 1

[0035] 1. Sample collection

[0036] All cases underwent bone marrow puncture and underwent bone marrow cell morphology, flow cytometry, histocytochemistry, chromosomes, fusion genes and electron microscopy examinations. The clinical diagnosis was mainly based on cell morphology and histocytochemistry, combined with other examinations, and confirmed according to FAB.

[0037] Serum biochemical analysis study recruited patients from the Blood Disease Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, Chinese Academy of Medical Sciences) as research subjects. 56 volunteers were divided into 18 DLBCL groups (10 males and 8 females) and 38 CLL groups (25 males and 13 females). The age range of the DLBCL group was 25 years old to 71 years old, and the age range of the CLL group was 35 years old to 76 years old. Raman spectroscopy study selected serum from CLL and DLBCL patients in the Blood Disease Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, Chinese Academy of Medical Sciences) in 2024, mainly patients who were hospitalized in our hospital from January 2024 to April 2024, including 5 males and 4 females, with a minimum age of 31 years old and a maximum age of 73 years old, including 4 CLL patients, 5 DLBCL patients, and 5 control groups (healthy people). This study was approved by the Ethics Committee of the Blood Disease Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, Chinese Academy of Medical Sciences) (KT2020016-EC-2). All patients underwent routine serum biochemical tests. Serum biochemical test data were obtained from the Clinical Testing Center of the Hematology Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, Chinese Academy of Medical Sciences).

[0038] 2. General clinical data collection

[0039] The subjects' detailed medical history, clinical data and blood-related indices were collected. Serum was collected after fasting for 10 hours, and peripheral blood biochemical indices were measured using an automatic biochemical analyzer. Biochemical indices included white blood cell (WBC), platelet (PLT), hemoglobin (HGB), neutrophil percentage (NEU%), neutrophil absolute value (NEU#), lymphocyte percentage (LYMPH%), lymphocyte absolute value (LYMPH#), total protein (TP), albumin (ALB), globulin (GLB), alanine transaminase (ALT), aspartate transaminase (AST), alkaline phosphatase (ALP), glutamyl transpeptidase (GGT), total bile acid (TBA), and serum albumin (SLA). creatinine (CREA), uric acid (UA), lactate dehydrogenase (LDH), creatine kinase (CK), creatine kinase isoenzyme (CK-MB), a-hydroxybutyrate dehydrogenase (a-HBDH), amylase (AMY), potassium (K), sodium (Na), chloride (Cl), calcium (Ca), phosphorus (P), magnesium (Mg), carbon dioxide combining power (CO2 combining power, CO2CP), glucose (GLU), triglyceride (TG), total cholesterol (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL),LDL), folic acid (FA), vitamin B12 (Vitamin B12, B12), ferritin (F), erythropoietin (TPO), iron (Iron), unsaturated iron (UIBC), total iron binding force (TIBC), iron saturation (ISAT), prothrombin time (PT), international standardization ratio (INR), partial thromboplastin time (APTT), thrombin time (TT), fibrinogen (FIB), determination of antithrombin III activity (ATI II), fibrinogen decomposition products (FDB), D-dimer quantification (DD), immunoglobulin G (Immunoglobulin G, IgG), Immunoglobulin A (ImmunoglobulinA, IgA), Immunoglobulin M (Ig), Complement C3 (complement C3, C3), Complement C4 (complement C4, C4), C-reactive protein (CRP), Rheumatoid factor, Anti-Streptolysin O (Anti-Streptolysin O, ASO). ,

[0040] 3. Raman spectroscopy

[0041] Drop 5 μL of serum on a calcium fluoride slide (or quartz slide) and use a confocal Raman spectrometer XploRARaman microscope to measure the amount. Select 785nm laser as the excitation light, the output power is 10mW, the objective lens is 40 times, and the specimen is fixed on the XYZ three-dimensional platform. The shooting process uses a ×400.6NA Nikon lens, and the spot size range of about 2×2μm on the sample receives the irradiation of the laser beam with an output power of 10mW, and the measurement range is 600~1800cm -1 , 6 sites were measured in each group with a resolution of 1cm -1At the same time, the Raman spectrum of the quartz glass slide was measured as the background. Labspec6 software was used for data processing such as smoothing, background removal and baseline correction. All spectra were analyzed at their respective 1450cm -1 The Raman peaks were intensity normalized to the internal standard.

[0042] 4. Analysis of clinical data and Raman spectroscopy data and establishment of identification model

[0043] SIMCA14.1 software was used to perform OPLS-DA analysis on the clinical data (biochemical indicators) of the DLBCL and CLL groups and the Raman spectroscopy data of the serum of healthy controls, DLBCL and CLL patients. The goodness of fit parameter R was used. 2 and Q 2 The performance of the OPLS model was evaluated separately. Under the null hypothesis, the model was resampled 200 times by random changes in the y matrix for model validation. Cluster analysis and ROC curve analysis were performed. In order to find statistically significant Raman peaks in the classification model as potential biomarkers, V+S analysis was used. Peaks with Variable Importance (VIP)>0.5 and correlation coefficients (or distances from the center in the V+S graph) in the top range of the equivalence column were selected as potential biomarkers. The potential biomarkers obtained were subjected to a significance test, and potential biomarkers with P<0.05 were considered statistically significant. Origin software was used for relevant data processing.

[0044] 5. Statistical analysis

[0045] SPSS27.0 and GraphpadPrism 9.0 were used for data analysis and graphics generation. The statistical spectral data were independent of each other. Normally distributed data were expressed as M±SE (mean±standard error of the mean). Homogeneous variance groups were analyzed using one-way analysis of variance (ANOVE) and Bonferroni's multiple comparison test. Heterogeneous variance groups were analyzed using Welch's one-way analysis of variance (Welch'sANOVE) and Tamhane's T2 multiple comparison test. Non-normally distributed data were expressed as m(Q1-Q3) (median (25-75 percentile) and compared using KruskalWallis test and Dunn's multiple comparison test.

[0046] Clinical data were as described above. Normally distributed data were analyzed using unpaired t-test and Welch's unpaired t-test between equal variance groups and unequal variance groups, respectively. Non-normally distributed data were compared using the Mann-Whitney U test. P < 0.05 was the significance threshold. Frequency data were compared using the chi-square test and Fisher's exact test.

[0047] 6. Results

[0048] 6.1 Raman spectroscopy analysis of serum in CLL, DLBCL and control groups

[0049] In this study, 25 serum Raman characteristic spectra of CLL patients, 29 serum Raman characteristic spectra of DLBCL patients, and 30 serum Raman characteristic spectra of the control group were obtained. Table 1 shows the peak positions of serum Raman spectra. -1 The Raman spectra of the serum of the CLL group, DLBCL group and control group can be seen in the range of 1.34°, with similar morphology, which can well reflect the content and composition of the substances in the serum of the CLL group, DLBCL group and control group (see Figure 1 ). Based on the above spectral graphs, it is impossible to identify the sample differences among the CLL group, DLBCL group and control group of patients, so it is necessary to use multivariate statistical methods to establish a classification model.

[0050] Table 1 Peak positions of serum Raman spectra

[0051]

[0052]

[0053] 6.2 Establishment of a method for distinguishing CLL, DLBCL and control groups based on Raman spectroscopy

[0054] The Raman characteristic spectra obtained from the serum samples of the control group (healthy people), CLL and DLBCL patients were used to form three sets of data. The sample data were subjected to supervised orthogonal partial least squares discriminant analysis (OPLS-DA) for detailed analysis and comparison. Permutation analysis showed that Q 2 The intercept on the Y axis is negative, indicating that the OPLS-DA model is valid and not overfitting (see Figure 2 The cluster analysis under the OPLS-DA model distinguished the Raman spectra of serum samples from the CLL group, DLBCL group, and control group with 100% accuracy (see Figure 2 B, E, H, K in the figure). On this basis, the ROC curve was used to evaluate the reliability of the identification method. The closer the area under the curve (AUC) is to 1, the higher the reliability of the identification method. The ROC curve shows that in the control vs CLL vs DLBCL model, AUC(control)=1, AUC(CLL)=1, AUC(DLBCL)=1 ( Figure 2 C in the control vs CLL model, AUC(control)=1, AUC(CLL)=1( Figure 2In the control vs DLBCL model, AUC(control)=1, AUC(DLBCL)=1( Figure 2 In the CLLvsDLBCL model, AUC(CLL)=1, AUC(DLBCL)=1( Figure 2 L in the figure indicates that the discriminant analysis results are highly reliable.

[0055] 6.3 Identification of Potential Raman Biomarkers

[0056] To further investigate the relationship between serum samples of CLL group, DLBCL group and control group and screen potential biomarkers, OPLS-DA model was used for further analysis. Figure 3 A in the figure is OPLS-DA scoreplot, and the horizontal axis represents the score value of the main component in the orthogonal signal filtering (OSC) process, so the difference between the groups can be seen from the direction of the horizontal axis. The vertical axis represents the score value of the orthogonal component in the OSC process, so the difference between the samples in the group can be seen from the direction of the vertical axis. In the figure, the three groups of serum samples of CLL group, DLBCL group and control group are well distinguished. The control group is located on the positive half axis of X, and the CLL and DLBCL groups are located on the negative half axis of X, reflecting that the control group and the disease group (non-Hodgkin's lymphoma) are distinguished. This distinction result shows that the use of supervised OPLS-DA method can better identify the serum samples of healthy people and non-Hodgkin's lymphoma patients, laying the foundation for analyzing the material metabolism characteristics of the three groups.

[0057] Figure 3 B in the figure is OPLS-DA Loading plot, which is used to preliminarily screen the Raman peaks that contribute to the control vs CLL vs DLBCL classification model. The red, blue, green, purple, orange and black peak numbers in the figure are related to nucleic acids, proteins, carotenoids, lipids, carbohydrates and collagen, respectively. There is a correlation between the loading plot and the score plot in representing the relationship between the substance content of the two groups of patients, that is, the content of the substance represented by the peak located on the positive semi-axis of the vertical coordinate of the loading plot is relatively higher in the group on the positive semi-axis of the horizontal coordinate of the score plot, and the negative semi-axis of the vertical coordinate of the loading plot and the negative semi-axis of the horizontal coordinate of the score plot also have a similar corresponding relationship. It can be observed from the figure that the peaks representing proteins in the control group (621, 1603cm -1 ), peaks representing nucleic acids (726, 1579 cm -1 ), the peak representing carotenoids (1155cm -1 ), representing sugars (842cm -1) and the peak representing lipids (1446cm -1 ) was significantly higher than that in the CLL group and DLBCL group.

[0058] Figure 3 The C in the figure is OPLS-DAV+S plot, which integrates parameters such as VIP and correlation coefficient that characterize the contribution of the peak position to the classification model. Each point in the figure represents a peak position. When the color of the point is closer to red, it means that the correlation coefficient value of this peak position is higher, and the contribution of this peak position to the classification model is relatively greater. On the contrary, the closer the color of the point is to blue, the lower the correlation coefficient value of this peak position is, and the contribution of this peak position to the classification model is relatively smaller. In addition, when the point representing the peak position is farther away from the center in the V+S plot, it means that the contribution of this peak position to the classification model is relatively greater. In order to further study the peak positions that contribute to the identification model, the V+S plot was used to further screen potential biomarkers, and the characteristic peak positions were tested for significance in subsequent analysis to determine the peak positions that can effectively identify the CLL group, DLBCL group and control group as potential biomarkers ( Figure 3 C in the figure). The VIP value corresponding to the Raman peak in the classification model is the main indicator for evaluating the contribution of this peak to the classification model. V+Splot can provide a list of Raman peaks in descending order of VIP values. Peaks with VIP>0.5 and biological significance are screened as potential markers.

[0059] Based on the control vs CLL vs DLBCL model, the samples of the control group were combined with those of the CLL group and the DLBCL group for OPLS-DA analysis ( Figure 3 D~L).

[0060] Figure 3 D, G, and J are the OPLS-DA score plots of the three models: Control vs CLL, Control vs DLBCL, and CLL vs DLBCL. The two groups of samples in the three figures are located on the positive and negative half axes of the X-axis, respectively. The samples are clearly clustered in the scatter plots, reflecting that OPLS-DA can better extract the difference information in the spectrum. The established identification method can identify the differences in metabolic components of serum samples. The three models can well identify the CLL group and DLBCL group samples in the model.

[0061] Figure 3E, H, and K in the figure are the loading plots of the three models of Control vs CLL, Control vs DLBCL, and CLL vs DLBCL, respectively. The results show that the peak positions of the proteins represented by the control group (621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1 ), peaks representing nucleic acids (726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 ), peaks representing carotenoids (957, 1155, 1162 cm -1 ), representing sugars (842cm -1 ) and collagen (1345cm -1 ) and the peaks representing lipids (957, 1078, 1119, 1285, 1299, 1437, 1443, 1446 cm -1 ) was significantly higher than that in the CLL group and DLBCL group.

[0062] Figure 3 F, I, and L in the figure are V+S plots of the three models of Control vs CLL, Control vs DLBCL, and CLL vs DLBCL, respectively. The V+S plot can provide a list of Raman peaks in descending order of VIP values, which provides the main basis for determining potential biomarkers in the Control, CLL, and DLBCL models. Peaks with VIP>0.5 and biological significance are screened as potential markers. In the four models of control vs CLL vs DLBCL, Control vs CLL, Control vs DLBCL, and CLL vs DLBCL, VIP (VIP>0.5), correlation coefficient, load, and distance from the center in the V+S plot were comprehensively considered, and important Raman peaks that affect sample classification were found. In the subsequent biomarker validation process, Raman peaks with no statistical difference were excluded from the biomarker range.

[0063] 6.4 Validation of Raman Biomarkers

[0064] Potential biomarkers derived from the CLL vs DLBCL model were validated and statistical analysis was performed on all Raman data of the diseases involved in the model ( Figure 4Statistical analysis showed that the peak intensities of the representative proteins in the Control group were higher than those in the CLL group and the DLBCL group, including 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, and 1647 cm -1 There was a significant statistical difference compared with the CLL group and the DLBCL group (P<0.05). The peak intensity of the nucleic acid in the control group was higher than that in the CLL group and the DLBCL group, including 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 There were significant statistical differences compared with the CLL group and DLBCL group (P<0.05). The peak positions of carotenoids in the Control group (957, 1155, 1162 cm -1 ) was significantly higher than that in the CLL group and DLBCL group (P=0.000).

[0065] The control group represents sugars (842 cm -1 ) and collagen (1345cm -1 The peak intensities of the control group (957, 1078, 1119, 1285, 1299, 1437, 1443, 1446 cm) were significantly higher than those of the CLL group and the DLBCL group (P<0.001). -1 ) was significantly higher than that in the CLL group and DLBCL group (P<0.05).

[0066] 7. Establishment of a method for differentiating DLBCL and CLL groups based on multi-parameter analysis

[0067] In order to establish a method for distinguishing DLBCL and CLL groups based on multi-parameter analysis, the biochemical indicators of DLBCL and CLL groups constituted a set of data. The supervised orthogonal partial least squares discriminant analysis (OPLS-DA) was applied to the sample data using SIMCA-P software and a detailed analysis and comparison was carried out. The effectiveness of the supervised OPLS-DA model established based on multi-parameter analysis can be verified by permutation plots ( Figure 5 A in the figure), cluster analysis diagram ( Figure 5 B), and ROC plots( Figure 5 In order to determine whether the OPLS-DA model is valid, permutation is used to determine whether the model is valid. Permutation analysis shows that Q 2 The intercept on the Y axis is negative, indicating that the OPLS-DA model is valid and not overfitting (see Figure 5 The cluster analysis results show that the model has a good effect in distinguishing the samples in each group (see Figure 5 B in Figure 1). On this basis, ROC plots were used to evaluate the authenticity of the identification method. The closer the area under the curve (AUC) is to 1, the higher the authenticity of the identification method. The ROC curve shows that in the DLBCL vs CLL model, AUC (DLBCL) = 1 and AUC (CLL) = 1 (see Figure 1). Figure 5 C) in the figure indicates that the discriminant analysis results are highly accurate.

[0068] 8. Identification and validation of potential biochemical markers

[0069] In order to further study the relationship between the relevant indicators of DLBCL and CLL groups and B-cell blood tumors and screen potential biomarkers, the OPLS-DA model was used for further analysis. Figure 5 D in the figure is OPLS-DA scoreplot, the horizontal axis represents the score value of the main component in the OSC process, and the vertical axis represents the score value of the orthogonal component in the OSC process. The two groups of samples in the figure are clearly clustered, reflecting that the DLBCL and CLL groups are distinguished. This distinction result shows that the use of the supervised OPLS-DA method can well identify the relevant indicators of the DLBCL and CLL groups, laying the foundation for analyzing the relevant indicators of the two groups. The two groups of samples in the figure are located on the positive and negative half axes of the X-axis, respectively. The samples are clearly clustered in the scatter plot, reflecting that OPLS-DA can better extract the difference information in the relevant indicators, and the established identification method can identify the differences in the relevant indicators of different groups of samples. Figure 5 E in the figure is OPLS-DA loading plot, which is used to preliminarily screen the indicators that contribute to the identification model of DLBCL and CLL groups. Loading plot and score plot are correlated when expressing the relationship between sample-related indicators, that is, the indicators on the positive semi-axis of the loading plot ordinate are relatively higher in the population on the positive semi-axis of the score plot abscissa, and the negative semi-axis of the loading plot ordinate and the negative semi-axis of the score plot abscissa also have a similar corresponding relationship. Figure 5The F in the figure is OPLS-DAVIPplot, which reflects the relationship between the peak VIP value and the correlation coefficient in the model. The closer the peak in the figure is to red, the greater the correlation coefficient of this peak in the identification model. V+S plot integrates parameters such as VIP and correlation coefficient that characterize the contribution of indicators to the classification model. Each point in the figure represents an indicator. When the color of the point is closer to red, it means that the correlation coefficient value of this indicator is higher, and the contribution of this indicator to the classification model is relatively greater. On the contrary, the closer the color of the point is to blue, the lower the correlation coefficient value of this indicator is, and the contribution of this indicator to the classification model is relatively smaller. In addition, when the point representing the indicator is farther away from the center in the V+S plot, it means that the contribution of this indicator to the classification model is relatively greater. In order to further study the indicators that contribute to the identification model, V+S plot was used to further screen potential biomarkers, and the characteristic peaks were tested for significance in subsequent analysis to determine the indicators that can effectively identify the DLBCL and CLL groups as potential biomarkers ( Figure 5 F in the figure). The VIP value corresponding to the relevant index in the classification model is the main indicator for evaluating the contribution of this index to the classification model. V+S plot can provide a list of blood indexes in descending order of VIP values. Peaks with VIP>0.5 and biological significance are screened as potential markers.

[0070] Statistical analysis of peripheral blood biochemical data suggests (see Figure 5 In the G~O), ALB, TT and WBC of CLL group were significantly higher than those of DLBCL group (P<0.05), and TG, HDL, GGT, ALP, ALT and NEUT% of CLL group were significantly lower than those of DLBCL group (P<0.05).

[0071] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A predictive or diagnostic system for identifying non-Hodgkin's lymphoma, characterized in that The prediction or diagnosis system comprises a Raman spectrum detection module for the sample to be tested and a Raman spectrum detection module for the negative control; the Raman spectrum detection module for the sample to be tested and the Raman spectrum detection module for the negative control both use the Raman spectrum data of the test serum as an indicator; the Raman spectrum data are Raman peaks 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm representing proteins -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 intensity level.

2. The prediction or diagnosis model system according to claim 1, characterized in that: The negative control was healthy human serum.

3. The prediction or diagnosis system according to claim 1, characterized in that The prediction or diagnosis system also includes a conclusion output module. When the Raman spectrum data of the negative control is significantly higher than the Raman spectrum data of the sample to be tested, it is predicted or diagnosed as non-Hodgkin's lymphoma.

4. The prediction or diagnosis system according to claim 1, characterized in that The non-Hodgkin's lymphoma is diffuse large B-cell lymphoma and chronic lymphocytic leukemia.

5. Raman peaks of serum representative proteins detected by Raman spectroscopy: 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 The application of the intensity level as a biomarker in the preparation of a product for predicting or diagnosing non-Hodgkin's lymphoma.

6. The use according to claim 5, characterized in that The biomarkers in the serum of the test sample are detected and compared with the peak intensity values ​​of the negative control to predict or diagnose non-Hodgkin's lymphoma.

7. The use according to claim 5, characterized in that The non-Hodgkin's lymphoma is diffuse large B-cell lymphoma and chronic lymphocytic leukemia.

8. Raman peaks of serum representative proteins detected by Raman spectroscopy: 621, 643, 848, 853, 869, 935, 1003, 1031, 1221, 1230, 1260, 1344, 1443, 1446, 1548, 1579, 1603, 1647 cm -1 , Raman peaks representing nucleic acids: 726, 781, 786, 1078, 1190, 1415, 1573, 1579 cm -1 , Raman peaks representing carotenoids: 957, 1155, 1162 cm -1 , Raman peak representing sugars 842cm -1 , Raman peak representing collagen 1345cm -1 and the Raman peaks representing lipids at 957, 1078, 1119, 1285, 1299, 1437, 1443, and 1446 cm -1 The application of the intensity level as a biomarker in the preparation of a predictive or diagnostic non-Hodgkin's lymphoma typing product.

9. The use according to claim 8, characterized in that: The non-Hodgkin's lymphoma is classified as diffuse large B-cell lymphoma or chronic lymphocytic leukemia.

Images