A biomarker and its application in the diagnosis of lung squamous cell carcinoma
By detecting myeloid cell markers in the tumor and interstitial regions of lung squamous cell carcinoma, especially CD163 and S100A12, a prognostic detection and diagnostic model of lung squamous cell carcinoma independent of clinical characteristics was constructed, solving the problem that the existing technology did not introduce tumor infiltration. The problem of sexual myeloid cells has achieved good prognostic detection and diagnostic effects.
Patent Information
- Application Number
- CN202210802337.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-07-07
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2042-07-07
AI Technical Summary
The existing lung squamous cell prognostic risk prediction model has not introduced tumor-invasive myeloid cells, and its application in lung squamous cell prognostic risk prediction has not been discussed, resulting in a lack of effective targeted therapy and a lower five-year survival rate.
By detecting the expression levels of myeloid cell markers in tumor areas and interstitial areas, biomarkers such as CD163 and S100A12 were screened out to build a prognostic detection and diagnostic model for lung squamous cell carcinoma independent of clinical characteristics, and use these markers for prediction and diagnosis. .
It realizes prognostic detection and diagnosis of lung squamous cell carcinoma with convenient operation. The established model has good AUC values in the training set and verification set, and can be effectively applied in medical technology research, providing new survival of myeloid-based cells. Risk prediction model.
Smart Images

Figure CN115308418B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of bioengineering, and in particular to a biomarker and its application in the diagnosis of lung squamous cell carcinoma. Background Art
[0002] Lung cancer is one of the malignant tumors with the highest morbidity and mortality in the world. Non-small cell lung cancer (NSCLC) is the main subtype of lung cancer, accounting for about 80% of all lung cancers, and squamous cell carcinoma (LUSC) is a common tissue type. The treatment of LUSC relies on surgical intervention, lacks effective targeted drugs, and has a poor prognosis, with a low five-year survival rate. Therefore, building a high-performance LUSC prognosis prediction model is necessary for selecting better treatment methods and improving prognosis.
[0003] Tumor microenvironment (TME) refers to the cellular environment in which tumors or cancer stem cells exist, which contains immune cells, blood vessels, extracellular matrix, fibroblasts, lymphocytes, bone marrow-derived inflammatory cells, and signaling molecules. TME affects the poor prognosis and survival outcomes of cancer patients by affecting programmed death ligand 1 or programmed death receptor 1. Monitoring the components of TME can complement the evaluation of patient treatment response by histopathology and molecular biomarkers. At present, people have a relatively broad understanding of the role of microenvironment, however, how to translate it into clinical application remains unclear. Therefore, effective and simple characterization of the local TME of individual LUSC patients can predict survival outcomes and related risk stratification and clinical management. Immune cells of various types and sources are important components of TME and play a key role in regulating anti-tumor immune responses. At present, most studies focus on T cells, and only a few studies have begun to focus on immune cells other than T cells, among which the role of tumor-infiltrating myeloid cells (TIMs) in tumors has attracted attention. For example, patent CN202111442433.9 discloses a biomarker for detecting lung cancer and lung cancer prognosis, and the biomarker includes any combination of SLITRK6, NIPAL4, DPPA2, ID1, STK24, ARL2BP, MYEOV, MME, CRYBB1, RETN, LRIF1, EPG5, COX16, PTCD1, C8B, and UFD1L. Patent CN202110729775.2 discloses another biomarker for predicting lung cancer prognosis, and the marker includes NLN, BMP5, CBFA2T3, and DSG2.
[0004] The existing prognostic risk prediction model for lung squamous cell carcinoma does not introduce tumor-infiltrating myeloid cells, and it is not disclosed whether tumor-infiltrating myeloid cells can be used to predict the prognostic risk of lung squamous cell carcinoma. Therefore, it is urgent to provide a new biomarker based on tumor-infiltrating myeloid cell markers and its application in the prognosis prediction and diagnosis of lung squamous cell carcinoma. Summary of the invention
[0005] In view of the above-mentioned deficiencies, the present invention provides a biomarker and its application in the diagnosis of squamous cell carcinoma of the lung. The present invention detects the expression levels of myeloid cell markers in the tumor area and interstitial area of the sample to be tested, and screens out biomarkers that can be used for prognosis detection and / or diagnosis of squamous cell carcinoma of the lung, wherein the biomarkers include CD163 and S100A12. Compared with the prior art, the use of the biomarkers described in the present invention for prognosis detection and / or diagnosis of squamous cell carcinoma of the lung is easy to operate and can be effectively applied to actual medical technology research. The prognosis model established by the present invention has a good AUC in both the training set and the validation set, and is a model independent of clinical characteristics.
[0006] In order to achieve the above-mentioned object of the invention, the technical solution of the present invention is as follows:
[0007] In one aspect, the present invention provides a biomarker for prognosis detection and / or diagnosis of squamous cell lung carcinoma, wherein the biomarker comprises CD1A, CD14, CD123, CD68, CD163, CD74, S100A12 and / or XCR1.
[0008] Specifically, the biomarkers include CD163 and / or S100A12.
[0009] On the other hand, the present invention provides the use of the above biomarkers in the preparation of lung squamous cell carcinoma prognosis detection and / or diagnosis products.
[0010] Specifically, the biomarkers include CD1A, CD14, CD123, CD68, CD163, CD74, S100A12 and / or XCR1.
[0011] More specifically, the biomarkers include CD163 and / or S100A12.
[0012] Specifically, in the product, the prognosis prediction and / or diagnosis model is: R=C1×CD163T+C2×CD163S+C3×S100A12T+1.53.
[0013] More specifically, the C1, C2, and C3 are 0.02205, -0.00489, and 0.00396, respectively.
[0014] More specifically, the T is the tumor region, and the S is the interstitial region.
[0015] In yet another aspect, the present invention provides a product for prognosis detection and / or diagnosis of squamous cell lung carcinoma, wherein the product comprises the above-mentioned biomarker.
[0016] As an example of the product, the product may be a device for predicting the prognosis of lung squamous cell carcinoma, the device comprising a predictive prognosis analysis unit, and the predictive prognosis analysis unit predicts the prognosis of lung squamous cell carcinoma patients using the lung squamous cell carcinoma prognosis model described above.
[0017] Specifically, the device further comprises an immunohistochemical staining unit, which detects the expression level of molecular markers.
[0018] Specifically, the device also includes a display unit, which displays the prognosis prediction results of patients with lung squamous cell carcinoma.
[0019] Specifically, the device further includes an evaluation result sending unit, which sends the prognosis prediction result obtained by the prediction analysis unit to the display unit.
[0020] As an example of the product, the product may be an electronic device, comprising: a memory for non-temporarily storing computer-readable instructions; and a processor for executing the computer-readable instructions, wherein when the computer-readable instructions are executed by the processor, the aforementioned construction method is executed or the following steps are executed: a risk score is calculated using the aforementioned risk score model according to the aforementioned myeloid cell marker expression level for constructing the risk score model in the sample of a patient with squamous cell carcinoma of the lung.
[0021] As an example of the product, the product may be a storage medium that non-temporarily stores computer-readable instructions, wherein when the non-temporary computer-readable instructions are executed by a computer, instructions for the construction method described above or instructions for the following steps are executed: a risk score is calculated using the risk score model described above based on the expression levels of myeloid cell markers used to construct the risk score model described above in samples of patients with squamous cell carcinoma of the lung.
[0022] In another aspect, the present invention provides a method for establishing a prognosis detection and / or diagnosis model for lung squamous cell carcinoma, the method comprising the following steps:
[0023] (1) Select myeloid cell markers;
[0024] (2) The collected specimens were randomly divided into a training set and a test set. In the training set, the tumor and stromal regions of the collected LUSC tissues were subjected to standard immunohistochemical staining procedures for myeloid markers and the degree of myeloid cell infiltration in each region was calculated;
[0025] (3) By performing LASSO regression on the immunohistochemistry results in the training set, setting the random seed number and defining a 10-fold cross-validation model and plotting the variation graph of the coefficient of variation, the corresponding variable at lambda.1se was selected according to the size of the coefficient of variation;
[0026] (4) Calculate the coefficients of the screened variables through COX regression to determine the prognostic model;
[0027] (5) The Cox regression model established above was verified using the data from the test set. If the C-index > 0.5, the model was considered to have statistical significance in predicting prognosis;
[0028] (6) Determine the prognostic risk of patients with squamous cell lung cancer based on the model.
[0029] Specifically, the myeloid cell markers described in step (1) include CD1A, CD14, CD123, CD68, CD163, CD74, S100A12 and / or XCR1, which include myeloid cell markers including monocytes, macrophages, dendritic cells and neutrophils.
[0030] Specifically, the training set in step (2) accounts for 50% of the total data set, and the test set accounts for 50% of the total data set.
[0031] Specifically, the random seed number described in step (3) is 123.
[0032] Specifically, the variables described in step (4) include CD163T, CD163S, and S100A12T.
[0033] Specifically, the formula of the prognostic prediction model is risk score R=C1×CD163T+C2×CD163S+C3×S100A12T+1.53.
[0034] More specifically, the C1, C2, and C3 are 0.02205, -0.00489, and 0.00396, respectively; T is the tumor area, and S is the interstitial area.
[0035] Further specifically, the risk score of each subject in the training data set is calculated based on the risk scoring model; the goodness of fit of the risk scoring model is evaluated using the time-dependent receiver operating characteristic curve analysis of the training data set; the grouping cutoff value is determined based on the time-dependent receiver operating characteristic curve analysis of the training data set, and the subjects in the training data set are divided into a high-risk group and a low-risk group according to the grouping cutoff value; and the Kaplan-Meier curve of the training data set is used to evaluate whether there is a significant difference in survival between the high-risk group and the low-risk group.
[0036] Further specifically, the risk score of each subject in the test data set is calculated based on the risk scoring model; the goodness of fit of the risk scoring model is evaluated using the time-dependent receiver operating characteristic curve analysis of the test data set; the grouping cutoff value is determined based on the time-dependent receiver operating characteristic curve analysis of the test data set, and the subjects in the test data set are divided into a high-risk group and a low-risk group according to the grouping cutoff value; and the Kaplan-Meier curve of the test data set is used to evaluate whether there is a significant difference in survival between the high-risk group and the low-risk group.
[0037] Terms and Definitions:
[0038] The CD163 described in the present invention is a marker of M2 tumor-associated macrophages.
[0039] The S100A12 described in the present invention is a marker for neutrophils.
[0040] The "sample" described in the present invention is limited to lung squamous cell carcinoma tumor tissue.
[0041] Compared with the prior art, the present invention has the following positive and beneficial effects:
[0042] (1) Currently, most models for lung cancer survival prediction are based on lymphocytes, especially T cells, and there is a lack of research on the predictive function of immune cells other than lymphocytes. In addition, most lung cancer prognosis studies are based on non-small cell lung cancer, and there is a lack of research on prediction models related to squamous cell lung cancer.
[0043] The present invention has discovered a survival risk prediction model for squamous cell lung cancer based on the expression of myeloid cells, which solves the problem that previous models only focus on T cells and overall lung cancer. At the same time, the myeloid cells proposed in the present invention are a novel cell line with a huge impact on tumors, which can predict the survival of patients with squamous cell lung cancer independently or in combination with other risk factors.
[0044] (2) The biomarkers described in the present invention are used for prognosis detection and / or diagnosis of lung squamous cell carcinoma, which is easy to operate and can be effectively applied to actual medical technology research. The prognosis model established by the present invention has a good AUC in both the training set and the validation set, and is a model independent of clinical characteristics. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 This is the ROC curve of the training set.
[0046] Figure 2 This is the survival curve of the training set.
[0047] Figure 3 This is the ROC curve of the test set.
[0048] Figure 4 This is the survival curve of the test set. DETAILED DESCRIPTION
[0049] The present invention will be further described in detail below in conjunction with specific examples. The following examples are not intended to limit the present invention, but are only intended to illustrate the present invention. The experimental methods used in the following examples are generally conventional, unless otherwise specified, and the materials, reagents, etc. used in the following examples are commercially available, unless otherwise specified.
[0050] Example 1. Risk analysis model construction
[0051] 1. Study population
[0052] All patients must meet the following criteria: histological diagnosis of LUSC; no anticancer treatment before surgical resection; no diagnosis or history of any other concurrent malignant tumors; follow-up data are available. The selected LUSC patients were randomly divided into training and test sets. The characteristics of the training and test set populations are shown in Table 1.
[0053] Table 1: Baseline data of the study population
[0054]
[0055]
[0056] 2. Immunohistochemistry (IHC) staining
[0057] Standard immunohistochemical staining procedures were performed on LUSC samples. Formalin-fixed and paraffin-embedded tumor tissues were cut into 4-μm-thick tumor sections, which were subsequently dewaxed in xylene and rehydrated in ethanol. Antigen retrieval took 10 min in a Tris (pH 6.0) pressure cooker. Endogenous peroxidase was blocked with 3% H2O2 for 15 min. Primary antibodies were incubated on tissue sections overnight at 4°C. After washing three times in phosphate-buffered saline, sections were incubated with secondary antibodies for 1 h at room temperature and washed three times. After color development with 3,3-diaminobenzidine at a dilution of 1:200, hematoxylin was restained, and sections were dehydrated in ethanol and xylene.
[0058] 3. Determination of myeloid cell marker expression
[0059] 1) IHC-stained tissues were imaged using the Vectra 2.0 automated quantitative pathology imaging system with a 10X field of view, and the average counts of 2 fields of view were used for statistical analysis; 2) Image analysis was performed using InForm Cell Analysis software (PerkinElmer) to separate diaminobenzidine and hematoxylin signals. 3) Pathologists manually animated training areas of tumor, stroma, or blank areas to train machine learning tissue segmentation, which was then applied to all tissue images; 4) After tissue segmentation, the images were color deconstructed and the number of cells was estimated using the Count Objects module of InForm Cell Analysis.
[0060] 4. Model construction
[0061] Model construction was performed using the “rms”, “glmnet”, “survminer”, and “survival” packages in R software. A penalized Cox proportional hazards model with adaptive least absolute shrinkage and selection operator (LASSO) was used to screen variables using a 1-se criterion. Then, a scoring algorithm based on three bone marrow features was used to construct a penalty coefficient for predicting one-year survival in the training cohort. The final model was determined as follows: risk score R = 0.02205 × CD163T - 0.00489 ×
[0062] CD163S+0.00396×S100A12T+1.53.
[0063] The specific commands are as follows:
[0064]
[0065] 5. Model performance verification
[0066] In order to evaluate the accuracy of the prognostic model composed of three myeloid cell markers in predicting the prognosis of lung squamous cell carcinoma, we first performed 1-year and 5-year receiver operating characteristic (ROC) curve analysis in the training set and compared the respective AUC values. The results showed that the AUCs of 1 year and 5 years in the training set were 0.86 (95% CI: 0.64-0.96) and 0.65 (95% CI: 0.54-0.73), respectively. Figure 1 ). The AUC value showed that the prognostic model composed of three myeloid cell markers had good discriminatory performance for the prognosis of patients with lung squamous cell carcinoma. The results of survival analysis (KM) showed that the survival time of patients in the high-risk group was significantly shorter than that in the low-risk group ( Figure 2 ). Secondly, the same formula was used in the test set to calculate the risk score of each sample, and ROC and KM curve analysis were performed. The results showed the same trend as the training set ( Figure 3 and Figure 4 ). These results indicate that the risk score calculated based on the prognostic risk signature composed of three myeloid cell markers can better predict the prognosis of patients with lung squamous cell carcinoma.
[0067] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.
Claims
1. Application of the biomarkers CD163 and S100A12 in the preparation of a prognosis detection and / or prognosis diagnosis product for lung squamous cell carcinoma, wherein the prognosis model is: R=C1×CD163T+C2×CD163S+C3×S100A12T+1.53; T is the tumor area, and S is the interstitial area.
2. The use according to claim 1, characterized in that: The C1, C2 and C3 are 0.02205, -0.00489 and 0.00396 respectively.
3. A lung squamous cell carcinoma prognosis detection and / or prognosis diagnosis product, characterized in that: The product includes reagents for detecting biomarkers CD163 and S100A12; in the product, the prognostic model is: R=C1×CD163T+C2×CD163S+C3×S100A12T+1.53; T is the tumor area and S is the interstitial area.
4. The product according to claim 3, characterized in that: The C1, C2 and C3 are 0.02205, -0.00489 and 0.00396 respectively.
5. A method for prognosis detection and / or establishment of a prognostic diagnosis model for lung squamous cell carcinoma, characterized in that: The method comprises the following steps: (1) Select myeloid cell markers; (2) The collected specimens were randomly divided into a training set and a test set. In the training set, the tumor and stromal regions of the collected LUSC tissues were subjected to standard immunohistochemical staining procedures for myeloid markers and the degree of myeloid cell infiltration in each region was calculated; (3) By performing LASSO regression on the immunohistochemistry results in the training set, setting the random seed number and defining a 10-fold cross-validation model and plotting the variation graph of the coefficient of variation, the corresponding variable at lambda.1se was selected according to the size of the coefficient of variation; (4) Calculating the coefficients of the screened variables by COX regression to determine the prognostic model, wherein the variables include CD163T, CD163S and S100A12T, and the formula of the prognostic model is risk score R = C1 × CD163T + C2 × CD163S + C3 × S100A12T + 1.53, where T represents the tumor area and S represents the stromal area; (5) Use the data from the test set to verify the Cox regression model established above. If the C-index > 0.5, the model is considered to have statistical significance in predicting prognosis; (6) Determine the prognostic risk of patients with squamous cell lung cancer based on the model.
6. The method according to claim 5, characterized in that: The myeloid cell markers described in step (1) include CD1A, CD14, CD123, CD68, CD163, CD74, S100A12 and / or XCR1.
7. The method according to claim 6, characterized in that: The myeloid cell markers described in step (1) are CD163 and / or S100A12.
8. The method according to claim 5, characterized in that: The C1, C2, and C3 described in step (4) are 0.02205, -0.00489, and 0.00396, respectively.
