Marker set for predicting postoperative complication risk of patient with thymoma and application of marker set

Abstract

The invention belongs to the technical field of biology, and particularly relates to a marker set for postoperative complication risk of a patient with thymoma and application of the marker set. It is found that the risk value of postoperative complications of a patient suffering from thymoma can be predicted through index values of Masaoka staging, MG, intraoperative bleeding amount, FEV1% pred and albumin expression level (Alb). The marker set is established to measure the postoperative complication risk value of the patient, the marker set is convenient to sample, the result can be quickly obtained, and the accuracy is high. Meanwhile, a postoperative complication prediction model based on perioperative period data is constructed, operation and anesthesia schemes are accurately adjusted, and perioperative period management of the patient is optimized. Through the model, a doctor identifies a high-risk patient before an operation and formulates a personalized operation and postoperative monitoring nursing plan, so that the occurrence rate of complications and the related death risk are effectively reduced, and the safety, the overall treatment effect and the life quality of the patient of the thymoma operation are improved; the pain of a patient in an operation is relieved, the medical cost of the patient is reduced, and the compliance of the patient is improved; the clinical application significance is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of biology (biomarkers), and particularly relates to a biomarker set for the risk of postoperative complications in thymoma patients and its application. Background Art

[0002] The thymus is an important immune organ in the human body, located in the upper part of the anterior mediastinum behind the manubrium sterni. Its main function is to participate in the development of the immune system. A mediastinal tumor derived from thymic epithelium is called thymoma (Ettinger DS, Riely GJ, Akerley W, et al. Thymomas and thymic carcinomas: Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2013. 11(5):562 - 76.). Thymoma is closely related to autoimmune disorders, and its unique disease characteristics include: associated paraneoplastic syndrome; occult invasive growth; the location of the tumor body is adjacent to important organs in the human body. The treatment of thymoma needs to formulate an individualized treatment plan according to the specific situation of the patient. The gold standard treatment for surgery is median sternotomy, which destroys the integrity of the bony structure of the human chest and has a relatively large trauma. Due to various factors, the postoperative mortality rate is relatively high.

[0003] For patients with thymoma who can be surgically resected, the preferred treatment method is surgery (Ma Xianyou, Xia Hongwei, Li Qingzhi, et al. Clinical analysis of surgical treatment of thymoma [J]. Journal of Qiqihar Medical College, 2008.). Surgical resection can effectively remove tumors, relieve symptoms, reduce recurrence and metastasis, especially prolong the long-term survival of tumor patients (Kamel MK, Stiles BM, Ghaly G, et al. Predictors of Pleural Implants in Patients With Thymic Tumors. Ann Thorac Surg. 2016. 102(5):1647-1652.). However, surgical resection also brings postoperative complications that cannot be ignored. According to the literature, the incidence of postoperative complications of thymoma is as high as 25-48%, which is the main cause of high postoperative mortality (Kocaman G, Kayi Cangir A. Early stage thymoma and the surgical extent paradigm. Updates Surg. 2024.). Complications that may occur after thymoma surgery include myasthenic crisis, pulmonary infection, respiratory failure, etc. The occurrence of pulmonary infection is related to various factors, including preoperative pulmonary dysfunction, smoking history, preoperative diabetes mellitus, etc. (Ma Zhenfei. Analysis of related factors and prevention of pulmonary infection after thymoma surgery [D]. Zhengzhou University, 2016.). A small number of patients cannot be weaned from long-term continuous invasive ventilator support treatment, which becomes a complex and difficult problem in the diagnosis and treatment process of thoracic surgery. At present, there is no specific and detailed study on the occurrence factors of postoperative complications of thymoma in the literature.

[0004] Postoperative complications are related to a variety of factors, including inadequate preoperative preparation, surgical trauma, inappropriate dosage of anticholinesterase drugs, infection (especially lung infection), etc. The prognosis of thymoma patients is affected by a variety of factors, including tumor stage, pathological type, whether accompanied by myasthenia gravis, etc. (Qian Kunjie, Sun Qingchao, Li Desheng, et al. Analysis of long-term postoperative efficacy and influencing factors in patients with thymoma combined with myasthenia gravis [J]. Chinese Journal of Clinical Thoracic and Cardiovascular Surgery, 2016.). Masaoka staging is an important indicator for evaluating the prognosis of thymoma. The survival rate of patients in the early stage (stage I and II) is higher, while the survival rate of patients in the late stage (stage III and IV) is lower (Wu Kailiang, Jiang Guoliang, Mao Jingfang, et al. Analysis of comprehensive treatment results and prognostic factors of 283 cases of thymoma [J]. Chinese Journal of Oncology, 2008.). In addition, the pathological type of thymoma is also an important factor affecting prognosis (Liu Lu, Yang Na, Liu Zhi, et al. Discussion on the clinical pathological characteristics of 37 cases of thymoma in the new version of the WHO classification [J]. Modern Clinical Medicine, 2020.). The prevention and management of postoperative complications in patients with thymoma require comprehensive consideration of multiple factors, including preoperative preparation, choice of surgical method, postoperative nursing intervention, and effective respiratory management. In short, the treatment of thymoma patients needs to comprehensively consider factors such as the stage of the tumor, pathological type, and whether there is myasthenia gravis, adopt an individualized treatment plan, and improve the safety of surgery in order to achieve the best treatment effect and prognosis. However, there is currently no systematic data to support this.

[0005] Therefore, there is an urgent need to determine indicators for predicting the risk of postoperative complications in patients, so as to develop personalized preoperative intervention treatments for thymoma patients with different characteristics, improve surgical safety, improve prognosis, and further enhance the diagnosis and treatment level of thoracic surgery. Summary of the invention

[0006] The present invention found that Masaoka staging, MG, intraoperative blood loss, FEV 1 %pred and albumin (Alb) have an impact on the risk of postoperative complications in thymoma patients, and the probability of postoperative short-term complications can be predicted by each index value. Based on this, the present invention is completed.

[0007] In a first aspect, the present invention provides a marker set for predicting the risk of postoperative complications in patients with thymoma, wherein the marker set is selected from Masaoka-Koga stage, myasthenia gravis (MG), intraoperative blood loss, forced expiratory volume in one second as a percentage of predicted value (FEV 1 When the markers meet at least one of the following conditions, the thymoma patient has a risk of complications after surgery:

[0008] Masaoka-Koga is stage II;

[0009] The Masaoka-Koga stage is III;

[0010] The Masaoka-Koga stage is IV;

[0011] FEV 1 %pred ≤ 85.1%;

[0012] Diagnosed with MG;

[0013] The intraoperative blood loss ≥ 85 ml;

[0014] The albumin expression level ≤ 42.15 g / dl.

[0015] Furthermore, when the Masaoka-Koga stage of the thymoma patient is II, the risk of postoperative complications for this patient is 1.58 times that of a patient with Masaoka-Koga stage I.

[0016] Furthermore, when the Masaoka-Koga stage of the thymoma patient is III, the risk of postoperative complications for this patient is 5.49 times that of a patient with Masaoka-Koga stage I.

[0017] Furthermore, when the Masaoka-Koga stage of the thymoma patient is IV-a, the risk of postoperative complications for this patient is 9.66 times that of a patient with Masaoka-Koga stage I.

[0018] Furthermore, when the FEV of the thymoma patient 1 %pred ≥ 85.1%, the risk of postoperative complications for this patient is 3.56 times that of a patient with FEV 1 %pred < 85.1%.

[0019] Furthermore, when the thymoma patient has concomitant MG, the risk of postoperative complications for this patient is 7.43 times that of a patient without concomitant MG.

[0020] Furthermore, when the preoperative serum albumin expression level of the thymoma patient > 42.15 g / dl, the risk of postoperative complications for this patient is 4.61 times that of a patient with an albumin expression level ≤ 42.15 g / dl.

[0021] Furthermore, when the intraoperative blood loss of the thymoma patient > 85 mL, the risk of postoperative complications for this patient is 3.38 times that of a patient with an intraoperative blood loss ≤ 85 mL.

[0022] In a second aspect, the present invention provides an application of the marker set as described in the first aspect in the preparation of a reagent for predicting the risk of postoperative complications in thymoma patients, and the reagent is a reagent capable of detecting the marker indexes as described in the first aspect.

[0023] Furthermore, when the Masaoka-Koga stage of the thymoma patient is stage II, the risk of postoperative complications for this patient is 1.58 times that of a patient with Masaoka-Koga stage I.

[0024] Furthermore, when the Masaoka-Koga stage of the thymoma patient is stage III, the risk of postoperative complications for this patient is 5.49 times that of a patient with Masaoka-Koga stage I.

[0025] Furthermore, when the Masaoka-Koga stage of the thymoma patient is stage IV-a, the risk of postoperative complications for this patient is 9.66 times that of a patient with Masaoka-Koga stage I.

[0026] Furthermore, when the FEV 1 %pred ≥ 85.1% in the thymoma patient, the risk of postoperative complications for this patient is 3.56 times that of a patient with FEV 1 %pred < 85.1%.

[0027] Furthermore, when the thymoma patient has concomitant MG, the risk of postoperative complications for this patient is 7.43 times that of a patient without concomitant MG.

[0028] Furthermore, when the preoperative serum albumin expression level in the thymoma patient > 42.15 g / dl, the risk of postoperative complications for this patient is 4.61 times that of a patient with an albumin expression level ≤ 42.15 g / dl.

[0029] Furthermore, when the intraoperative blood loss in the thymoma patient > 85 mL, the risk of postoperative complications for this patient is 3.38 times that of a patient with intraoperative blood loss ≤ 85 mL.

[0030] In a third aspect, the present invention provides a kit for predicting the risk of postoperative complications in thymoma patients, and the kit contains the reagents as described in the second aspect.

[0031] Furthermore, when the Masaoka-Koga stage of the thymoma patient is stage II, the risk of postoperative complications for this patient is 1.58 times that of a patient with Masaoka-Koga stage I.

[0032] Furthermore, when the Masaoka-Koga stage of the thymoma patient is stage III, the risk of postoperative complications for this patient is 5.49 times that of a patient with Masaoka-Koga stage I.

[0033] Furthermore, when the Masaoka-Koga stage of the thymoma patient is stage IV-a, the risk of postoperative complications for this patient is 9.66 times that of a patient with Masaoka-Koga stage I.

[0034] Further, when the FEV of the thymoma patient 1 %pred ≥ 85.1%, the risk of postoperative complications for this patient is 3.56 times that of a patient with FEV 1 %pred < 85.1%.

[0035] Further, when the thymoma patient has concomitant MG, the risk of postoperative complications for this patient is 7.43 times that of a patient without concomitant MG.

[0036] Further, when the preoperative serum albumin expression level of the thymoma patient > 42.15 g / dl, the risk of postoperative complications for this patient is 4.61 times that of a patient with an albumin expression level ≤ 42.15 g / dl.

[0037] Further, when the intraoperative blood loss of the thymoma patient > 85 mL, the risk of postoperative complications for this patient is 3.38 times that of a patient with an intraoperative blood loss ≤ 85 mL.

[0038] Fourthly, the present invention provides a system for predicting the risk of postoperative complications in thymoma patients. Through this system, the following modules can be obtained:

[0039] (1) Data collection module, configured to select and input the basic information of thymoma patients: n1 (Masaoka-Koga), n2 (FEV 1 %pred), n3 (MG), n4 (albumin expression level), and n5 intraoperative blood loss data;

[0040] (2) Risk assessment operation module, which processes the data collected above:

[0041] When n1 = Masaoka-Koga stage II, P1 = 1.58;

[0042] When n1 = Masaoka-Koga stage III, P1 = 5.49

[0043] When n1 = Masaoka-Koga stage IV, P1 = 9.66

[0044] When n2 > 85.1, P2 = 3.56

[0045] When n3 = MG, P3 = 7.43;

[0046] When n4 > 42.15, P4 = 4.61;

[0047] When n5 > 85, P5 = 3.38;

[0048] The risk of postoperative complications P for thymoma patients = P1 + P2 + P3 + P4 + P5;

[0049] (3) Risk assessment report display module, presenting the P value obtained by the risk assessment operation module on the display screen.

[0050] Further, when the Masaoka-Koga stage of the thymoma patient is II, the risk of postoperative complications for this patient is P1 = 1.58 times that of patients with Masaoka-Koga stage I.

[0051] Further, when the Masaoka-Koga stage of the thymoma patient is III, the risk of postoperative complications for this patient is P1 = 5.49 times that of patients with Masaoka-Koga stage I.

[0052] Further, when the Masaoka-Koga stage of the thymoma patient is IV-a or b, the risk of postoperative complications for this patient is P1 = 9.66 times that of patients with Masaoka-Koga stage I.

[0053] Further, when the FEV 1 %pred ≥ 85.1% for the thymoma patient, the risk of postoperative complications for this patient is P2 = 3.56 times that of patients with FEV 1 %pred < 85.1%.

[0054] Further, when the thymoma patient has concomitant MG, the risk of postoperative complications for this patient is P3 = 7.43 times that of patients without concomitant MG.

[0055] Further, when the albumin expression level of the thymoma patient > 42.15 g / dl, the risk of postoperative complications for this patient is P4 = 4.61 times that of patients with albumin expression level ≤ 42.15 g / dl.

[0056] Further, when the intraoperative blood loss of the thymoma patient > 85 mL, the risk of postoperative complications for this patient is P5 = 3.38 times that of patients with intraoperative blood loss ≤ 85 mL.

[0057] Fifth aspect, the present invention provides the application of the system as described in the fourth aspect in the collection of postoperative complication information of thymoma patients.

[0058] Beneficial effects

[0059] A set of markers for predicting the risk of postoperative complications in thymoma patients is screened out. The risk value of postoperative complications of patients is obtained by measuring the marker set. This marker set has the advantages of convenient sampling, can obtain results quickly, and has high accuracy. At the same time, a postoperative complication prediction model based on perioperative data is constructed to precisely and individually adjust the surgical and anesthesia plans and optimize the perioperative management of patients. Through this model, doctors can identify high-risk patients before surgery, formulate more personalized surgical and postoperative monitoring and nursing plans, thereby effectively reducing the incidence of complications and related death risks, improving the safety, overall treatment effect of thymoma surgery and the quality of life of patients; reducing the pain of patients during surgery, reducing medical expenses, and improving patient compliance. This prediction model, kit, and system have production prospects and are of great significance in clinical applications. Description of the Drawings

[0060] Figure 1 Predicted values, AUC values of influencing factors, and data preprocessing results.

[0061] Note: (A). ROC data of the predictive variable at the optimal cut-off point; (B). Correlation matrix in the multivariable regression model, (C). Variance inflation factor (VIF) analysis.

[0062] Figure 2 Nomination map for recent complications of patients after thymoma surgery.

[0063] Note: FEV 1 %pred: Percentage of forced expiratory volume in the first second to the predicted value.

[0064] Figure 3 ROC curves for predicting recent postoperative complications of thymoma patients in the model cohort and validation cohort.

[0065] Figure 4 Validation and evaluation of the prediction model.

[0066] Note: Calibration curves (A) of the modeling cohort and (B) of the validation cohort for predicting recent complications of patients after thymoma surgery; the X-axis is the complication probability predicted by the nomination map, and the Y-axis is the actual overall survival rate; the long dashed line (ideal) represents the ideal graph, that is, the predicted result is exactly the same as the actual result; the blue dashed line (bias correction) represents the bootstrap corrected performance, and the orange solid line (apparent) represents the apparent accuracy; both the apparent line and the bias correction line roughly follow the ideal line downwards, indicating that the scores calculated by the nomination map accurately represent the actual prediction results of recent complications in the modeling cohort and validation cohort; decision curve analysis (DCA) of the nomination map for the modeling cohort (C) and validation cohort (D); the X-axis represents the threshold probability, and the Y-axis represents the net benefit. Detailed Implementation Manner

[0067] The following further describes the specific embodiments of the present invention. It should be noted here that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation to the present invention. In addition, the technical features involved in the following described embodiments can be combined with each other as long as they do not conflict with each other.

[0068] Unless otherwise specified, the experimental methods in the following examples are all conventional methods. Unless otherwise specified, the test materials used in the following examples can all be obtained through conventional commercial channels.

[0069] Term Explanation

[0070] Forced expiratory volume in one second (FEV 1 ): It refers to the fastest expiratory volume within 1 second after maximal inspiration to the TLC position, abbreviated as one-second volume, and is often used to judge the degree of impairment of lung ventilation function.

[0071] Diffusing capacity of the lung for carbon monoxide (DLCO): Also known as the transfer factor of the lung for carbon monoxide (TLCO), it is an index to measure the transfer of gas in the inhaled air to red blood cells. Clinically, the single-breath method is mostly used to measure DLCO of the lung to reflect the lung diffusion function.

[0072] The Masaoka stage is an important indicator for evaluating the prognosis of thymoma. The survival rate of patients in the early stage (stage I and II) is relatively high, while the survival rate of patients in the late stage (stage III and IV) is relatively low.

[0073] The Akaike information criterion (AIC) is a standard for evaluating the complexity of a statistical model and measuring the goodness of fit of a statistical model to data. It was founded and developed by the Japanese statistician Hirotugu Akaike. The Akaike information criterion is based on the concept of information entropy.

[0074] The univariate Logistic regression method is a statistical analysis method used to handle binary classification problems. It converts the continuous predicted value into a probability value between 0 and 1 by inputting the result of linear regression into the Logistic function (also known as the Sigmoid function), so as to determine the probability of each result occurring. This method is particularly applicable to many fields such as medicine, social science, and marketing analysis. For example, in medical research, it is often used to predict the incidence probability of a certain disease. In this application, variables with a p-value < 0.1 are considered relevant to the research outcome.

[0075] Backward stepwise Logistic regression method: A method used in statistical analysis for model selection and simplification. This method starts with a complete model that includes all independent variables and gradually eliminates those variables that contribute the least to the model until all remaining variables are statistically significant.

[0076] Modeling Queue and Validation Queue: In this application, they are channels for processing training data and validation data. The main purpose is to efficiently provide data to the model for training and validation. To ensure the comparability and consistency of the model results in the training and validation phases, the model and its parameters need to be kept consistent: avoiding inequality between training and validation, ensuring parameter sharing, guaranteeing the continuity of weights and biases, preventing validation bias, and connecting with the inference phase.

[0077] Example 1 Inclusion and exclusion criteria and data collection

[0078] Through the hospital's electronic medical record system, the clinical data of 253 patients who underwent thymoma resection in the Department of Thoracic Surgery of our hospital from June 2006 to July 2024 were retrospectively collected for analysis. The design of this study conforms to the Helsinki Declaration of Research Ethics and has obtained the approval of the Ethics Committee of Beijing Chest Hospital, Capital Medical University to waive the written informed consent form (Ethical number: Clinical Research 2019(85)).

[0079] Inclusion criteria:

[0080] The postoperative pathological morphology conforms to thymoma; there is no respiratory failure in the preoperative arterial blood gas analysis. For those with underlying diseases, the diseases are under stable control; the case data are relatively complete.

[0081] Exclusion criteria:

[0082] There is chronic heart failure; chronic renal failure; and concomitant malignancies in other parts.

[0083] The definition of postoperative complications in this study is:

[0084] During the period from after thymoma surgery to the patient's discharge, the occurrence of: death within 30 days after surgery, respiratory failure, secondary infection of the lungs or other organs, and heart, liver, and kidney insufficiency are regarded as the occurrence of recent postoperative complications (187 cases in the non-complication group and 66 cases in the complication group).

[0085] Baseline data collection:

[0086] General clinical data of the patients were collected, including gender, age, body mass index (BMI), smoking status, and comorbidities (including hypertension, coronary heart disease, chronic lung disease, and diabetes). The size of the tumor (cm) was measured based on the maximum diameter of the tumor during the operation. According to Masaoka-Koga, thymomas were classified into stage I, stage II, stage III, and part of stage IV. According to the WHO pathological classification, they were divided into A, AB, B1, B2, and B3. According to the Osserman MG staging system, they were divided into stage 0 (without MG) and ≥ stage 1 (with MG).

[0087] Preoperative auxiliary examinations:

[0088] The pulmonary function measuring instrument was Master Screen-IOS and Master Screen-PFT produced by Jaeger Company, Germany. The one-breath method was used to measure the pulmonary diffusion function. The pulmonary function indexes included the percentage of the forced expiratory volume in the first second accounting for the predicted value (FEV 1 %pred), and the percentage of the carbon monoxide diffusion capacity accounting for the predicted value (DLCO%pred). Laboratory examinations included hemoglobin, serum albumin, and C-reactive protein. Chest CT and whether the surrounding organs, tissues, and blood vessels were invaded were also examined.

[0089] Surgical-related information:

[0090] Common surgical approaches included left or right video-assisted thoracoscopy, median sternotomy, left or right thoracotomy. The intraoperative blood loss and operation duration were recorded. Whether the phrenic nerve was invaded by the tumor and resected during the operation; whether the pericardium, part of the lung lobe, lymph, and adjacent adipose tissue were invaded by the tumor were also examined.

[0091] Example 2 Collection of sample data before and after surgery

[0092] A total of 253 patients were included, and they were followed up before and after surgery. All the data in Example 1 were classified and statistically analyzed.

[0093] A. Statistical methods

[0094] Measurement data were expressed as mean ± standard deviation, and the t-test was used for comparison between groups; measurement data with non-normal distribution were expressed as median percentage numbers (25%, 75%), and the Z-test was used for comparison between groups; the data of count data were expressed as constituent ratios or rates (%), and the χ 2 test was used for comparison between groups of data, and the baseline data were statistically described. The patients included in the study were randomly divided into a modeling cohort and a validation cohort at a ratio of 7:3. The mean and median were respectively selected as the cut-off values for normal and non-normal distribution continuous variables, and they were converted into binary variable data to calculate the differences between the training set (i.e., the modeling cohort) and the validation set (i.e., the validation cohort).

[0095] To convert continuous variables into categorical variables, we adopted the AUC method to determine the cut-off of variables (Figure 1A), excluding only the two variables of age and BMI. For the age variable, the median was selected as the dividing criterion, while for the BMI variable, it was divided according to the internationally accepted weight classification criteria: BMI < 18.5 kg / ㎡ was defined as underweight, 18.5 kg / ㎡ ≤ BMI < 24.0 kg / ㎡ was defined as normal weight, 24.0 kg / ㎡ ≤ BMI < 28.0 kg / ㎡ was defined as overweight, and BMI ≥ 28.0 kg / ㎡ was defined as obesity. After completing the conversion of all variable types, we further evaluated whether there were significant differences among these categorical variables between the training set (i.e., the modeling cohort) and the validation set (i.e., the validation cohort).

[0096] The risk factors for postoperative complications of thymoma were identified using the Logistic regression method. In the univariate Logistic regression analysis, variables with a p-value < 0.1 were considered related to the study outcome and were further analyzed using the 'backward stepwise' Logistic regression method. A nomogram was developed using a regression model with the minimum Akaike information criterion (AIC) (Song X, Zhao Q, Zhang H, et al. Development and Validation of a Preoperative CT-Based Nomogram to Differentiate Invasive from Non-Invasive Pulmonary Adenocarcinoma in Solitary Pulmonary Nodules. Cancer Manag Res. 2022. 14:1195-1208.). Multicollinearity was evaluated by correlation test and variance inflation factor (VIF, Figure 1B,C) (Cheng J, Sun J, Yao K, Xu M, Cao Y. A variable selection method based on mutual information and variance inflation factor. Spectrochim Acta A Mol Biomol Spectrosc. 2022. 268:120652.). The discrimination of the evaluation model was evaluated using the receiver operating characteristic curve (ROC) in the modeling and validation cohorts, and the calibration of the constructed prediction model in the modeling and validation cohorts was evaluated using the calibration curve and Hosmer-Lemeshow goodness-of-fit test (HLGT) (Wang Z, Liu Q. Development of novel prediction models for nodal and distant metastasis in G1 and G2 colorectal neuroendocrine tumors. Int J Colorectal Dis. 2023. 38(1):37.). The decision curve analysis (DCA) method was used to evaluate the clinical benefit of the nomogram. P < 0.05 was considered statistically significant.All statistical analyses were performed in R software V.4.2.0. The analyses involved: "mass" (for performing stepwise regression); "rms" (for constructing nomograms); "car" (for calculating the variance inflation factor VIF); "pROC" (for plotting ROC curves); "rms" (for calculating and plotting calibration curves and performing the Hosmer-Lemeshow goodness-of-fit test); "rmda" (for decision curve analysis).

[0097] B. Grouping of postoperative complications

[0098] A total of 253 patients were included, 66 with postoperative complications and 187 without postoperative complications.

[0099] From the time the patient left the operating room after the operation until the patient was discharged, the following occurred: death within 30 days after the operation, myasthenic crisis, respiratory muscle paralysis, respiratory failure, secondary infection of the lungs or other organs, and heart, liver, and kidney insufficiency were regarded as postoperative complications in the patient. There were a total of 66 cases, and the specific conditions are as follows:

[0100] There were 2 deaths during hospitalization and within 30 days after discharge, 33 cases of respiratory failure, 26 cases of postoperative infection, 3 cases of heart failure, 1 case of liver insufficiency, and 1 case of kidney insufficiency.

[0101] Among the patients who developed respiratory failure after the operation, approximately 64.7% (22 / 33 cases) were patients with severe respiratory failure, and the duration of mechanical ventilation was 12 - 252 hours.

[0102] C. Test results

[0103] As shown in Table 1, there were differences in Masaoka-Koga (P < 0.001); Pathological Type (P < 0.001); Myasthenia Gravis (MG) (P < 0.001); surgical approach (P < 0.001); intraoperative blood loss (BloodLossVolume) (P < 0.001); and the clearing range between the two groups (P < 0.001).

[0104] Table 1. Demographics and clinical characteristics of thymoma surgery patients

[0105]

[0106]

[0107]

[0108] Note: Intraoperative resection and clearing range #: 0. Tumor body; 1. Tumor body + fat; 2. Tumor body + fat + part of the lung, part of the pericardium, pleura Abbreviation: FEV 1 %pred: Percentage of forced expiratory volume in the first second as a percentage of the predicted value; DLCO%pred: Percentage of carbon monoxide diffusing capacity as a percentage of the predicted value.

[0109] Example 3 Create a cohort

[0110] As shown in Table 2, among 253 patients, 177 patients entered the modeling cohort and 76 patients entered the validation cohort.

[0111] A. Statistical methods

[0112] All data in Example 1 were statistically analyzed. Measurement data were expressed as mean ± standard deviation, and t-tests were used for between-group comparisons; measurement data with non-normal distributions were expressed as median percentages (25%, 75%), and Z-tests were used for between-group comparisons; count data were expressed as constituent ratios or rates (%), and χ 2 tests were used for between-group comparison of data, and baseline data were statistically described. The patients included in the study were randomly divided into a modeling cohort and a validation cohort at a ratio of 7:3. The mean and median were respectively selected as the cut-off values for normal and non-normal distribution continuous variables, and they were converted into binary variable data, and the differences in this data between the training set and the validation set were calculated.

[0113] To convert continuous variables into categorical variables, the AUC (Area Under the Curve) method was used to determine the optimal cut-off values (cutoff values) for variables other than age and BMI (Figure 1A). For the age variable, the median was selected as the division criterion, while the BMI variable was divided according to the internationally accepted weight classification standard: BMI < 18.5 kg / m 2 was underweight, 18.5 kg / ㎡ ≤ BMI < 24.0 kg / m 2 was normal weight, 24.0 kg / ㎡ ≤ BMI < 28.0 kg / m 2 was overweight, BMI ≥ 28.0 kg / m 2 was obese. After completing this conversion of variable types, χ 2 tests were further used to evaluate whether there were significant differences in these categorical variables between the training set (i.e., the modeling cohort) and the validation set (i.e., the validation cohort).

[0114] B. Test results

[0115] The incidence rate of postoperative complications in the modeling cohort was 25.42% (45 / 177), and that in the validation cohort was 27.63% (21 / 76). After testing, there was no significant statistical difference in the clinical characteristic factors between the modeling cohort and the validation cohort, indicating reasonable data splitting.

[0116] Table 2. Basic clinical data of thymoma surgery patients in the comparison of modeling and validation cohorts

[0117]

[0118]

[0119] Abbreviation: FEV 1 %pred: Percentage of forced expiratory volume in the first second accounting for the predicted value; DLCO%pred: Percentage of diffusing capacity of carbon monoxide accounting for the predicted value.

[0120] Establishment of the prediction system model in Example 4

[0121] In the univariate regression analysis of the modeling cohort, multiple significantly correlated risk factors for postoperative complications were identified.

[0122] A. Analysis method

[0123] The Logistic regression method was used to identify the risk factors of postoperative complications in thymoma. In the univariate Logistic regression analysis, variables with a p-value < 0.1 were considered related to the research outcome and were further analyzed using the 'backward stepwise' Logistic regression method.

[0124] A regression model with the minimum Akaike information criterion (AIC) was used to develop a nomogram (Song X, Zhao Q, Zhang H, et al. Development and Validation ofa Preoperative CT-Based Nomogram to Differentiate Invasive from Non-Invasive Pulmonary Adenocarcinoma in Solitary Pulmonary Nodules. Cancer Manag Res. 2022. 14:1195-1208.).

[0125] Through correlation test and variance inflation factor (VIF, such as Figure 1Evaluate multicollinearity as shown (Cheng J, Sun J, Yao K, Xu M, Cao Y. A variable selection method based on mutual information and variance inflation factor. Spectrochim Acta A Mol Biomol Spectrosc. 2022;268:120652.).

[0126] Apply the receiver operating characteristic curve (ROC) to evaluate the discrimination of the model in the modeling and validation cohorts respectively, and use the calibration curve and Hosmer-Lemeshow goodness-of-fit test (HLGT) to evaluate the calibration of the constructed prediction model in the modeling and validation cohorts (Wang Z, Liu Q. Development of novel prediction models for nodal and distant metastasis in G1 and G2 colorectal neuroendocrine tumors. Int J Colorectal Dis. 2023;38(1):37.).

[0127] Use the decision curve analysis (DCA) method to evaluate the clinical benefit of the nomogram. P < 0.05 is considered statistically significant. All statistical analyses were performed in R software, including: "mass" (for performing stepwise regression); "rms" (for constructing the nomogram); "car" (for calculating the variance inflation factor VIF); "pROC" (for plotting the ROC curve); "rms" (for calculating and plotting the calibration curve and performing the Hosmer-Lemeshow goodness-of-fit test); "rmda" (for decision curve analysis).

[0128] B. Test Results

[0129] As Figure 1 shown, the Hb (g / L) threshold was 138.5, the AUC value was 0.553, 95% CI: 0.470 - 0.636, and the Youden index was 0.14349; the Alb (g / L) threshold was 42.15, the AUC value was 0.596, 95% CI: 0.517 - 0.676, and the Youden index was 0.17469; the CRP (mg / L) threshold was 2.51, the AUC value was 0.576, 95% CI: 0.495 - 0.658, and the Youden index was 0.21301; FEV 1The %pred threshold was 85.1, the AUC value was 0.637, 95% CI: 0.553 - 0.721, and the Youden index was 0.27897; the DLCO%pred threshold was 72.2, the AUC value was 0.602, 95% CI: 0.519 - 0.686, and the Youden index was 0.20945; the Maximum.Diameter.of.Tumor (cm) threshold was 4.55, the AUC value was 0.611, 95% CI: 0.534 - 0.688, and the Youden index was 0.22371; the Operation.Time (min) threshold was 157.5, the AUC value was 0.634, 95% CI: 0.555 - 0.713, and the Youden index was 0.20321; the Blood.Loss.Volume (mL) threshold was 85, the AUC value was 0.691, 95% CI: 0.616 - 0.765, and the Youden index was 0.33333.

[0130] As Figure 2 shown, it is a nomogram for the short-term complications of patients after thymoma surgery. After further screening of the short-term complications after thymoma surgery; through detailed univariate regression analysis of the modeling cohort, significant associations were found between Masaoka-Koga, FEV 1 %pred, MG, intraoperative blood loss, and albumin expression levels and the risk of postoperative complications; therefore, a risk prediction system model for postoperative complications in thymoma patients was constructed through the above indicators.

[0131] Table 3. Logistic regression analysis of risk factors for short-term complications after thymoma surgery (modeling cohort)

[0132]

[0133]

[0134]

[0135] Note: Variables with p < 0.1 in univariate logistic regression were included in multivariate stepwise backward logistic regression. The resulting model was fitted using the variables selected by this stepwise method. Surgical resection extent: 0. Tumor; 1. Tumor + fat; 2. Tumor + fat + lung, pericardium, pleura, etc.

[0136] Example 5 Validation of the system model

[0137] The system model constructed in Example 4 was used to evaluate the risk of postoperative complications in patients in the training cohort, and the analysis method was as shown in Examples 3 and 4. The selected indicators were: Masaoka-Koga, FEV 1%pred, MG, intraoperative blood loss, and albumin expression level.

[0138] As Figure 3 shown, the area under the ROC curve of the prediction model was 0.84 (0.78 - 0.91); the area under the ROC curve of the validation model was 0.76 (0.63 - 0.89); both had high accuracy.

[0139] As Figure 4 shown, the calibration curves for predicting short-term complications in patients after thymoma resection were located in the modeling cohort (A) and the validation cohort (B), respectively; the X-axis was the complication probability predicted by the nomogram, and the Y-axis was the actual overall survival rate. The long dashed line (ideal) represented the ideal nomogram, that is, the predicted results were completely consistent with the actual results. The blue dashed line (bias correction) indicated the bootstrap correction performance of the nomogram, and the orange solid line (apparent) indicated the apparent accuracy of the nomogram. Both the apparent line and the bias correction line generally descended along the ideal line, indicating that the scores calculated by the nomogram accurately represented the actual predicted results of short-term complications in the modeling cohort and the validation cohort. Decision curve analysis (DCA) of the nomograms for the modeling cohort (C) and the validation cohort (D); the X-axis represented the threshold probability, and the Y-axis represented the net benefit.

[0140] The evaluation of risk factors for postoperative complications was as follows:

[0141] When the Masaoka-Koga stage of the thymoma patient was stage II, the risk of postoperative complications in this patient was compared with that in the Masaoka-Koga stage I group, OR: 1.58 (0.33 - 7.49), that is, the probability of postoperative complications in this group of patients was 1.58 times that of the Masaoka-Koga stage I group.

[0142] When the Masaoka-Koga stage of the thymoma patient was stage III, the risk of postoperative complications in this patient was compared with that in the Masaoka-Koga stage I group, OR: 5.49 (1.16 – 25.96), that is, the probability of postoperative complications in this group of patients was 5.49 times that of the Masaoka-Koga stage I group.

[0143] When the Masaoka-Koga stage of the thymoma patient was stage IV-a or b, the risk of postoperative complications in this patient was compared with that in the Masaoka-Koga stage I group, OR: 9.66 (0.99 – 93.98), that is, the probability of postoperative complications in this group of patients was 9.66 times that of the Masaoka-Koga stage I group.

[0144] When the FEV 1 %pred ≥ 85.1% in the thymoma patient, the risk of postoperative complications in this patient was compared with that in the FEV 1For patients in the <85.1% predicted group, the risk odds ratio (OR) was 4.61 (1.71–12.45), indicating that such patients had a 3.56-fold higher probability of developing postoperative complications compared to patients in the ≤85.1% predicted group. 1 For patients in the ≤85.1% predicted group, the probability of developing postoperative complications was 3.56 times higher.

[0145] When the thymoma patient had concomitant MG, the OR value for the risk of postoperative complications was 7.43, indicating that such patients had a 7.43-fold higher probability of developing postoperative complications compared to patients without MG.

[0146] When the albumin expression level in the thymoma patient was >42.15 g / dl, the risk of postoperative complications for such patients was 4.61 times that of patients with an albumin expression level ≤42.15 g / dl.

[0147] When the intraoperative blood loss in the thymoma patient was >85 mL, the OR value for the risk of postoperative complications was 3.38, indicating that such patients had a 3.38-fold higher risk of postoperative complications compared to patients with an intraoperative blood loss ≤85 mL.

[0148] The risk value of postoperative complications in patients was determined through a biomarker set, which had the advantages of convenient sampling, rapid result acquisition, and high accuracy. At the same time, a postoperative complication prediction model based on perioperative data was constructed to precisely and individually adjust the surgical and anesthesia plans and optimize the perioperative management of patients. Through this model, doctors can identify high-risk patients before surgery, formulate more personalized surgical and postoperative monitoring and nursing plans, thereby effectively reducing the incidence of complications and related death risks, improving the safety, overall treatment effect of thymoma surgery, and the quality of life of patients; alleviating the pain of patients during surgery, reducing medical costs, and enhancing patient compliance.

Claims

1. A marker set for predicting the risk of postoperative complications in patients undergoing thymoma surgery, wherein the marker set is selected from Masaoka-Koga stage, myasthenia gravis (MG), intraoperative blood loss, forced expiratory volume in the first second as a percentage of predicted value (FEV1%pred) and albumin expression level (Alb) When the markers meet at least one of the following conditions, the thymoma patient has a risk of complications after surgery: Masaoka-Koga is stage II; Masaoka-Koga is stage III; Masaoka-Koga is stage IV; FEV1%pred≤85.1%; Diagnosis of MG; Intraoperative blood loss ≥85ml; Albumin expression level ≤42.15g / dl.

2. A marker set for predicting the risk of postoperative complications in thymoma patients as claimed in claim 1, when the Masaoka-Koga of the thymoma patient is stage II, the risk of postoperative complications of the patient is 1.58 times that of the patient in stage I; when the Masaoka-Koga of the thymoma patient is stage III, the risk of postoperative complications of the patient is 5.49 times that of the patient in stage I; when the Masaoka-Koga of the thymoma patient is stage IV-a, the risk of postoperative complications of the patient is 9.66 times that of the patient in stage I; when the FEV1%pred of the thymoma patient is ≥85.1%, the probability of postoperative complications in the patient is 3.56 times that of the patient in stage I; When the thymoma patient has MG, the risk of postoperative complications of the patient is 7.43 times that of the patient without MG; when the albumin expression level of the thymoma patient is >42.15g / dl, the risk of postoperative complications of the patient is 4.61 times that of the patient with albumin expression level ≤42.15g / dl; when the intraoperative bleeding volume of the thymoma patient is >85mL, the risk of postoperative complications of the patient is 3.38 times that of the patient with intraoperative bleeding volume ≤85mL.

3. Use of the marker set according to claim 1 in the preparation of a reagent for predicting the risk of postoperative complications in patients with thymoma, wherein the reagent is a reagent that can detect the marker index according to claim 1.

4. Use of the marker set as claimed in claim 3 in the preparation of a reagent for predicting the risk of postoperative complications in patients with thymoma, wherein in the marker set, when the Masaoka-Koga of the thymoma patient is stage II, the risk of postoperative complications of the patient is 1.58 times that of the patient with Masaoka-Koga stage I; when the Masaoka-Koga of the thymoma patient is stage III, the risk of postoperative complications of the patient is 5.49 times that of the patient with Masaoka-Koga stage I; when the Masaoka-Koga of the thymoma patient is stage IV-a, the risk of postoperative complications of the patient is Masaoka-Koga stage IV-b. The risk of postoperative complications in the thymoma patient is 9.66 times that of the patient in stage I; when the FEV1%pred of the thymoma patient is ≥85.1%, the probability of postoperative complications in the patient is 3.56 times that of the patient with FEV1%pred<85.1%; when the thymoma patient has MG, the risk of postoperative complications in the patient is 7.43 times that of the patient without MG; when the albumin expression level of the thymoma patient is >42.15g / dl, the risk of postoperative complications in the patient is 4.61 times that of the patient with albumin expression level ≤42.15g / dl; when the intraoperative blood loss of the thymoma patient is >85mL, the risk of postoperative complications in the patient is 3.38 times that of the patient with intraoperative blood loss ≤85mL.

5. A kit for predicting the risk of postoperative complications in patients with thymoma, the kit comprising the reagent as claimed in claim 2.

6. A kit for predicting the risk of postoperative complications in patients with thymoma as claimed in claim 5, wherein the reagent detects the marker set as claimed in claim 1, wherein, in the marker set, when the Masaoka-Koga of the thymoma patient is stage II, the risk of postoperative complications of the patient is 1.58 times that of the patient with Masaoka-Koga stage I; when the Masaoka-Koga of the thymoma patient is stage III, the risk of postoperative complications of the patient is 5.49 times that of the patient with Masaoka-Koga stage I; when the Masaoka-Koga of the thymoma patient is stage IV-a, the risk of postoperative complications of the patient is Masaoka-Koga stage IV-b. The risk of postoperative complications in the thymoma patient is 9.66 times that of the patient in stage I; when the FEV1%pred of the thymoma patient is ≥85.1%, the probability of postoperative complications in the patient is 3.56 times that of the patient with FEV1%pred<85.1%; when the thymoma patient has MG, the risk of postoperative complications in the patient is 7.43 times that of the patient without MG; when the albumin expression level of the thymoma patient is >42.15g / dl, the risk of postoperative complications in the patient is 4.61 times that of the patient with albumin expression level ≤42.15g / dl; when the intraoperative blood loss of the thymoma patient is >85mL, the risk of postoperative complications in the patient is 3.38 times that of the patient with intraoperative blood loss ≤85mL.

7. A system for predicting the risk of postoperative complications in patients with thymoma, wherein the system can be obtained. The system includes the following modules: (1) Data collection module, which is configured for the selection and input of basic information of thymoma patients: n1 (Masaoka-Koga), n2 (FEV1% pred), n3 (MG), n4 (albumin expression level), and n5 intraoperative blood loss; (2) Risk assessment calculation module, which processes the data collected above: When n1=Masaoka-Koga stage II, P1=1.58; When n1=Masaoka-Koga stage III, P1=5.49 When n1=Masaoka-Koga stage IV, P1=9.66 When n2=FEV1%pred>85.1, P2=3.56 When n3 = MG, P3 = 7.43; When n4=albumin>42.15, P4=4.61; When n5 intraoperative blood loss>85, P5=3.38; Thymoma patients have a risk of complications after surgery P = P1 + P2 + P3 + P4 + P5; (3) A risk assessment report display module displays the P value obtained by the risk assessment calculation module on a display screen.

8. A system for predicting the risk of postoperative complications in patients with thymoma as claimed in claim 7, wherein when the data is processed, when the Masaoka-Koga of the thymoma patient is stage II, the risk of postoperative complications in the patient is P1=1.58 times that of the patient with Masaoka-Koga stage I; when the Masaoka-Koga of the thymoma patient is stage III, the risk of postoperative complications in the patient is P1=5.49 times that of the patient with Masaoka-Koga stage I; when the Masaoka-Koga of the thymoma patient is stage IV-a or b, the risk of postoperative complications in the patient is P1=5.6 times that of the patient with Masaoka-Koga stage II. P1=9.66 times for stage I patients; when the thymoma patient's FEV1%pred≥85.1%, the patient's postoperative complications are P2=3.56 times that of patients with FEV1%pred<85.1%; when the thymoma patient has MG, the patient's postoperative complication risk is P3=7.43 times that of patients without MG; when the thymoma patient has an albumin expression level>42.15g / dl, the patient's postoperative complication risk is P4=4.61 times that of patients with an albumin expression level≤42.15g / dl; when the thymoma patient has an intraoperative blood loss>85mL, the patient's postoperative complication risk is P5=3.38 times that of patients with an intraoperative blood loss≤85mL.

Images