A system and method for long-term prognosis prediction of liver disease patients

Abstract

The application discloses a kind of long-term prognosis prediction system and method of liver disease patient, comprising the following steps, step S1, the clinical data of cirrhosis acute decompensation and acute-on-chronic liver failure patient are collected, including anthropometry data, vital signs, laboratory data and the albumin binding capacity obtained by specific detection;Step S2, the competitive risk model for predicting the death of cirrhosis acute decompensation and acute-on-chronic liver failure patient after 1 year of admission is constructed, the prediction model is constructed using competitive risk model, and dynamic nomogram visualization prediction model is made into convenient prediction tool;Step S3, the prediction performance of model is evaluated by receiver operating characteristic curve.The albumin binding capacity and the long-term prognosis of decompensated cirrhosis and acute-on-chronic liver failure patient establish index relationship with great clinical significance, which can further optimize clinical management and guide drug treatment.

Description

Technical Field

[0001] This invention relates to the field of liver disease prediction, and more particularly to a system and method for predicting the long-term prognosis of liver disease patients. Background Technology

[0002] Acute decompensation (AD) is the acute onset of complications in patients with decompensated cirrhosis, including hepatic encephalopathy, ascites, esophageal and gastric variceal bleeding, and bacterial infection. Simultaneously, factors such as acute exacerbations of viral hepatitis, excessive alcohol consumption, use of hepatotoxic drugs, bacterial or fungal infections, and acute gastrointestinal bleeding can all exacerbate acute decompensated cirrhosis, leading to a rapid deterioration of liver function and ultimately liver and / or extrahepatic organ failure, known as acute-on-chronic liver failure (ACLF). The occurrence of complications significantly increases the mortality risk in patients with decompensated cirrhosis, and the risk is further amplified in patients with ACLF, with a 90-day mortality rate as high as 20%–30%. Currently, there is no specific treatment; liver transplantation is the only effective treatment that can improve survival rates. Patients with acute decompensated cirrhosis and ACLF can survive for a relatively long time with timely treatment, but because more than half of these patients lack identifiable predisposing factors or triggering events, they still face a high risk of death without transplantation. Therefore, there is an urgent clinical need for prognostic factors that can predict the long-term survival of patients with decompensated cirrhosis and acute-on-chronic liver failure, which would be helpful for high-risk groups and disease management.

[0003] Human serum albumin (HSA) is the most abundant protein in human blood plasma, accounting for approximately 50% of total plasma protein. HSA is not only a major component in maintaining plasma colloid osmotic pressure, but also plays a role in binding and transporting substances, antioxidation, anticoagulation and antithrombosis, regulating immune function, maintaining capillary integrity, promoting positive cardiac inotropic effects, and neuroprotection. Binding ability is the primary function of albumin molecules; albumin can bind to numerous endogenous and exogenous ligands thanks to its unique molecular structure. HSA exhibits a globular, heart-shaped conformation and contains three homologous domains, denoted as I (1-195), II (196-383), and III (384-585). Each domain consists of two subdomains, A and B, each composed of a single α-helix, mediating binding to various endogenous and exogenous ligands. Two important drug-binding sites, Sudlow I and Sudlow II, are located in subdomains IIA and IIIA, respectively. The former mediates the binding of drugs such as warfarin and phenylbutazone, while the latter has a preferential affinity for ibuprofen. In end-stage liver disease, albumin-binding capacity has been found to be reduced. This may be due to the damage to the molecular structure of circulating human serum albumin caused by severe systemic inflammation and oxidative stress in cirrhosis, thereby affecting the domains of the albumin molecule. Impaired albumin binding function includes not only impaired binding to drugs but also impaired binding to bilirubin, fatty acids, metal ions, etc. Meanwhile, albumin binding capacity is also associated with short-term clinical outcomes in patients with cirrhosis and acute-on-chronic liver failure; patients with lower albumin binding capacity have higher mortality rates at 28 and 90 days post-admission. However, the potential value of albumin binding capacity for the long-term prognosis of patients with acute decompensated cirrhosis and acute-on-chronic liver failure remains to be confirmed. Exploring the relationship between albumin binding capacity and the long-term prognosis of acute decompensated cirrhosis and acute-on-chronic liver failure has significant clinical implications, as it can reflect the degree of albumin molecule damage and changes in the body's homeostasis. Furthermore, it can be developed as a novel clinical biomarker to further optimize clinical management and guide drug therapy. Summary of the Invention

[0004] The purpose of this invention is to provide a system and method for predicting the long-term prognosis of patients with liver disease, so as to solve the problems mentioned in the background art.

[0005] To achieve the above-mentioned objectives, one aspect of the present invention provides a long-term prognosis prediction system for patients with liver disease. This system, targeting patients with decompensated cirrhosis and acute-on-chronic liver failure, includes a data collection module, a model building module, and an assessment and prediction module, wherein:

[0006] The data collection module is used to collect clinical data of patients with acute decompensated cirrhosis and acute-on-chronic liver failure, including anthropometric data, vital signs, laboratory data, and albumin binding capacity obtained by specific detection; the albumin binding capacity obtained by specific detection is to assess albumin binding capacity by experimentally measuring the remaining binding amount of the specific fluorescent marker dansylsarcosine at albumin binding site II.

[0007] The model building module is used to build a predictive model for death one year after admission in patients with acute decompensated cirrhosis and chronic-on-acute liver failure. The predictive model is built using a competing risk model and a convenient prediction tool is created by visualizing the predictive model using a dynamic nomogram.

[0008] The evaluation prediction module is used to assess the predictive performance of the model using receiver operating characteristic (ROC) curves.

[0009] Furthermore, the prognostic factors in the competing risk model for predicting long-term mortality in patients with acute decompensated cirrhosis and acute-on-chronic liver failure include albumin-binding capacity, mean arterial pressure, white blood cell count, and international normalized ratio.

[0010] Furthermore, the method described above for predicting the long-term prognosis of patients with acute decompensated cirrhosis and acute-on-chronic liver failure is used to predict 1-year survival rate.

[0011] Furthermore, the prediction model formula is as follows:

[0012] Prognostic score = -0.0136 * albumin binding capacity -0.0210 * mean arterial pressure +0.0767 * white blood cell count +0.7374 * international normalized ratio

[0013] The albumin binding capacity (%) is calculated as follows: fluorescence in the standard ultrafiltrate / fluorescence in the sample ultrafiltrate × 100. A prognostic score ≥ -1.48 indicates a high-risk group for death, while a prognostic score < -1.48 indicates a low-risk group for death.

[0014] Another aspect of the present invention provides a method for predicting the long-term prognosis of patients with liver disease, the method for predicting the long-term prognosis of patients with decompensated cirrhosis and acute-on-chronic liver failure, comprising the following steps:

[0015] Step S1: Collect clinical data of patients with acute decompensated cirrhosis and acute-on-chronic liver failure, including anthropometric data, vital signs, laboratory data, and albumin binding capacity obtained by specific tests.

[0016] Step S2: Construct a competing risk model to predict the mortality of patients with acute decompensated cirrhosis and acute-on-chronic liver failure one year after admission. The prediction model is constructed using a competing risk model, and a convenient prediction tool is created by visualizing the prediction model using a dynamic nomogram.

[0017] Step S3: Evaluate the predictive performance of the model using the receiver operating characteristic curve.

[0018] Furthermore, the albumin binding capacity obtained by the specific detection is evaluated by experimentally measuring the remaining binding amount of the specific fluorescent marker dansylsarcosine at albumin binding site II.

[0019] Furthermore, the prognostic factors in the competing risk model for predicting long-term mortality in patients with acute decompensated cirrhosis and acute-on-chronic liver failure include albumin-binding capacity, mean arterial pressure, white blood cell count, and international normalized ratio.

[0020] Furthermore, the method described above for predicting the long-term prognosis of patients with acute decompensated cirrhosis and acute-on-chronic liver failure is used to predict 1-year survival rate.

[0021] Furthermore, the prediction model formula is as follows:

[0022] Prognostic score = -0.0136 * albumin binding capacity -0.0210 * mean arterial pressure +0.0767 * white blood cell count +0.7374 * international normalized ratio

[0023] The albumin binding capacity (%) is calculated as follows: fluorescence in the standard ultrafiltrate / fluorescence in the sample ultrafiltrate × 100. A prognostic score ≥ -1.48 indicates a high-risk group for death, while a prognostic score < -1.48 indicates a low-risk group for death.

[0024] Compared with existing technologies, this system and method have the following advantages:

[0025] 1. This invention can accurately assess the long-term survival probability of hospitalized patients with acute decompensated cirrhosis and acute-on-chronic liver failure through rapid and convenient laboratory testing, promptly identify high-risk individuals for death, provide early treatment to prevent disease deterioration or perform liver transplantation as soon as possible to ensure maximum survival; for low-risk individuals, it can provide symptomatic treatment for corresponding complications and control the occurrence and development of subsequent complications, thereby prolonging the survival of cirrhosis patients.

[0026] 2. This invention has a high efficacy in predicting the 1-year mortality rate of patients with acute decompensated cirrhosis and acute-on-chronic liver failure, which is significantly higher than the conventional Model for End-stage Liver Disease (MELD) score, with an area under the AUC curve of 0.80 vs. 0.76 (p = 0.02).

[0027] 3. This invention establishes a predictive model for the 1-year prognosis of patients with acute decompensated cirrhosis and acute-on-chronic liver failure, and transforms it into a web-based tool. By inputting the values ​​of four parameters, the corresponding score and 1-year survival probability can be obtained, thereby completing the risk assessment of cirrhosis patients. The operation is simple and convenient, and the accuracy is high. Attached Figure Description

[0028] Figure 1 This is a flowchart of a method for predicting the long-term prognosis of patients with liver disease.

[0029] Figure 2 Dynamic line graph for testing queue prognostic model.

[0030] Figure 3 ROC curve for evaluating the performance of the prognostic prediction model for the test cohort.

[0031] Figure 4 ROC curves were used to evaluate the effectiveness of the cohort prognostic prediction model. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] like Figure 1 The diagram shown is a flowchart of the method of the present invention. The present invention provides a method for predicting the long-term prognosis of patients with acute decompensated cirrhosis and acute-on-chronic liver failure based on human serum albumin binding capacity. The specific prediction model construction method includes the following steps:

[0034] Step S1: Collect clinical data of patients with acute decompensated cirrhosis and acute-on-chronic liver failure, including anthropometric data, vital signs, laboratory data, and albumin binding capacity obtained by specific tests.

[0035] Step S2: Construct a competing risk model and dynamic nomogram to predict mortality one year after admission in patients with acute decompensated cirrhosis and acute-on-chronic liver failure.

[0036] Step S3: Evaluate the predictive performance of the model using the Receiver Operating Characteristic curve (ROC).

[0037] Clinical cohort construction in this invention: This study prospectively and non-selectively screened patients with acute decompensated cirrhosis and acute-on-chronic liver failure admitted to the Department of Infectious Diseases, Ruijin Hospital, affiliated with Shanghai Jiao Tong University School of Medicine, between June 2016 and November 2018, ultimately enrolling 385 patients for data analysis. All procedures involved were performed in accordance with the principles of the Declaration of Helsinki and the Guidelines of the International Conference on Harmonized System of Clinical Practice. This invention involves human participants, and the research protocol was reviewed and approved by the Ethics Committee of Ruijin Hospital, affiliated with Shanghai Jiao Tong University School of Medicine, with ethics approval number:

[2018] Lun Shen No.

[162] .

[0038] The inclusion criteria for this invention are: (1) age 18-80 years; (2) cirrhosis: cirrhosis diagnosed in the past or during the current hospitalization. The diagnostic criteria for cirrhosis are: pathological diagnosis of liver tissue; or comprehensive judgment by imaging (magnetic resonance, CT, B-mode ultrasound), gastrointestinal endoscopy (esophageal-gastric varices, portal hypertension signs), liver elastography and corresponding clinical manifestations and / or laboratory test indicators of cirrhosis; (3) the current hospitalization is non-selective, and the reason for admission is one or more of the following combinations of cirrhosis decompensation events: ascites (grade 2-3), upper gastrointestinal bleeding, jaundice (total bilirubin ≥5mg / dL), hepatic encephalopathy, bacterial / fungal infection.

[0039] The exclusion criteria for this invention are: (1) age < 18 years or age > 80 years; (2) post-liver transplant; (3) received commercial albumin solution infusion, plasma infusion, plasma exchange, artificial liver, etc. within 15 days prior to enrollment; (4) hepatocellular carcinoma at any stage and other advanced malignant tumors; (5) acute or chronic extrahepatic diseases that affect short-term prognosis and are combined with other systems, such as chronic kidney disease in the uremia stage (requiring dialysis treatment), chronic left heart failure, obstructive pulmonary disease, chronic respiratory failure, etc.; (6) long-term use of immunosuppressants for non-liver disease reasons, such as nephrotic syndrome, rheumatic diseases, anti-rejection reaction after organ transplantation, etc.; (7) long-term use of anticoagulants, such as warfarin for atrial fibrillation, etc.; (8) this admission is selective admission, such as only to complete the scheduled medical treatment, including but not limited to liver biopsy, splenectomy, TIPS (transjugular intrahepatic portosystemic shunt), HVPG measurement (hepatic venous pressure gradient), etc. Gradient), endoscopic banding, MDT consultation (Multi-Disciplinary Team), simple follow-up examination, etc.; (9) HIV antibody positive;

[0040] (10) Pregnant or breastfeeding women; (11) Patients refuse to participate in this study and refuse to sign the informed consent form; (12) None of the above conditions apply, but the patient is temporarily unable to sign the informed consent form due to coma or other reasons, and there is no legal representative to sign it for them. Based on the condition, it is judged that the patient may not be able to wake up and sign the informed consent form later.

[0041] Within 48 hours of enrollment, 5 mL of peripheral venous blood was collected from each patient, and the serum was immediately separated and frozen at -80°C. Anthropometric data, medical history, vital signs, and laboratory data of the enrolled patients were collected using the electronic medical record information system. Clinical data collection included: (1) gender, age, height, weight, and blood pressure;

[0042] (2) History of hypertension, diabetes, etiology of cirrhosis, history of decompensated cirrhosis, occurrence of decompensated events during hospitalization, occurrence of acute-on-chronic liver failure during hospitalization, admission time, discharge time, and survival outcome 1 year after admission; (3) Laboratory test indicators: white blood cell count (WBC), hemoglobin (Hb), platelet count (Plt), alanine aminotransferase

[0043] (ALT), aspartate aminotransferase (AST), total bilirubin (TB), serum albumin (Alb), serum creatinine (Cr), prothrombin time (PT), international normalized ratio (INR), and C-reactive protein (CRP). The MELD score of the patient was calculated using laboratory data. Albumin binding capacity was assessed by experimentally measuring the remaining binding amount of the specific fluorescent marker dansylsarcosine at albumin binding site II. The formula for calculating albumin binding capacity (%) was: Albumin binding capacity (%) = Fluorescence in standard ultrafiltrate / Fluorescence in sample ultrafiltrate × 100.

[0044] Statistical methods included: all statistical tests were performed using R software (version 4.3.2). The Shapiro test was used to test for normality. Continuous variables with normal distribution were expressed as mean ± standard deviation (SD), and skewed distributions were expressed as median (IQR). Categorical variables were described using counts (percentages). Comparisons between groups were performed using the Student t-test, Mann-Whitney U test, or χ² test. 2Tests were performed. For groups of two or more, ANOVA or the Kruskal-Wallis test was used. Competing risk regression analysis was used to identify risk factors associated with 1-year mortality in patients with acute decompensated cirrhosis, with transplantation as a competing risk. A backward stepwise approach was then used among variables with p < 0.05 in univariate analysis to select the optimal variable for further multivariate analysis and nomogram plotting. Parameters with strong correlations (Spearman correlation coefficients greater than 0.5) were not included in multivariate analysis due to multicollinearity. Receiver operating curves were used to assess the prognostic performance of the model, assuming that the survival outcome for liver transplant recipients was death and that no one survived at the end of the follow-up period. The cumulative incidence function was estimated using the Gray method, mortality risk was stratified, and the optimal threshold was determined based on the Youden index. In all analyses, the significance level was set at two-sided p < 0.05.

[0045] Of the 385 patients included in the final analysis, 332 had acute decompensated cirrhosis and 53 had acute-on-chronic liver failure. Based on one-year survival outcomes after admission, 251 survived, 106 died, and 28 underwent liver transplantation. Table 1 shows that comparing the baseline characteristics of survivors and non-survivors at one year of age revealed statistically significant differences in history of hypertension, mean arterial pressure at admission, white blood cell count, C-reactive protein, alanine aminotransferase, aspartate aminotransferase, total bilirubin, serum albumin, international normalized ratio, MELD score, and albumin-binding capacity.

[0046] Table 1. Baseline demographic, clinical, and laboratory data of patients with acute decompensated cirrhosis and acute-on-chronic liver failure upon admission:

[0047]

[0048]

[0049] As shown in Table 2, multivariate competing risk analysis identified the following independent predictors of 1-year mortality in hospitalized patients with acute decompensated cirrhosis and acute-on-chronic liver failure: albumin-binding capacity, mean arterial pressure, white blood cell count, and international normalized ratio (USNR). Their sub-distribution hazard ratios were: albumin-binding capacity (sHR 0.986; 95% CI, 0.979–0.994; p<0.001), mean arterial pressure (sHR 0.979; 95% CI, 0.962–0.997; p = 0.022), white blood cell count (sHR 1.080; 95% CI, 1.049–1.111; p<0.001), and USNR (sHR 2.091; 95% CI, 1.656–2.639; p<0.001).

[0050] Table 2. Multivariate independent risk factors for mortality 1 year after admission in patients with acute decompensated cirrhosis and acute-on-chronic liver failure:

[0051] variable Regression coefficient Sub-distribution hazard (95% CI) p Albumin binding capacity, per 1% -0.0136 0.986(0.979-0.994) <0.001 Mean arterial pressure, per 1 mmHg -0.0210 0.979(0.962-0.997) 0.022 <![CDATA[White blood cell count, per 10 9 cells / L]]> 0.0767 1.080(1.049-1.111) <0.001 International Normalized Ratio (INR), per 1 0.7374 2.091(1.656-2.639) <0.001

[0052] A prognostic prediction model was constructed using the regression coefficients of the selected independent risk factors. The model formula is: Prognostic Score = -0.0136 * Albumin-Binding Capacity - 0.0210 * Mean Arterial Pressure + 0.0767 * White Blood Cell Count + 0.7374 * International Normalized Ratio. The optimal Youden index yielded a cut-off value of -1.48 for this model. A score ≥ -1.48 indicates a high risk of death, while a score < -1.48 indicates a low risk of death. Simultaneously, a dynamic nomogram was created using the selected independent risk factors as a convenient web-based prediction tool. By selecting the values ​​of these four parameters, the model score and the corresponding 1-year survival probability are automatically calculated. Figure 2 The image shows a dynamic nocturnal plot of the prognostic model for the test cohort. The area under the curve (AUC) of the nocturnal plot is 0.80, significantly better than the MELD score. Figure 3 As shown.

[0053] To validate the predictive performance of the model, a validation cohort of 142 patients was included, comprising 132 patients with acute decompensated cirrhosis and 10 patients with acute-on-chronic liver failure. At one-year follow-up after admission, 110 patients survived, 27 died, and 5 underwent liver transplantation. The model was analyzed against the validation cohort, yielding an AUC of 0.76. Figure 4 As shown, the model has a good predictive ability for expected survival outcomes.

[0054] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A system for predicting long-term prognosis of a patient with liver disease, characterized by, The prognosis prediction system for decompensated cirrhosis and acute-on-chronic liver failure patients comprises a data collection module, a model construction module, and an evaluation prediction module, wherein: The data collection module is used to collect clinical data of patients with acute decompensated cirrhosis and acute-on-chronic liver failure, including anthropometric data, vital signs, laboratory data, and albumin binding capacity obtained by specific detection; the albumin binding capacity obtained by specific detection is evaluated by measuring the residual binding amount of specific fluorescent marker dansyl sarcosine to albumin binding site II through experiments; The model construction module is used to construct a prediction model for the death of patients with acute decompensated cirrhosis and acute-on-chronic liver failure one year after admission, the prediction model is constructed by using a competing risk model, and a dynamic nomogram is used to visualize the prediction model to make a convenient prediction tool; the prediction model formula is: Albumin binding capacity model = -0.0136*albumin binding capacity-0.0210*mean arterial pressure+0.0767*white blood cell count+0.7374*international normalized ratio; wherein the albumin binding capacity (%) = fluorescence in standard ultrafiltrate / florescence in sample ultrafiltrate*100; The evaluation prediction module is used to evaluate the prediction performance of the model by using a receiver operating characteristic curve.

2. The system for predicting long-term prognosis of a liver disease patient according to claim 1, wherein The prognostic factors of the prediction model for the death of patients with acute decompensated cirrhosis and acute-on-chronic liver failure one year after admission include albumin binding capacity, mean arterial pressure, white blood cell count, and international normalized ratio.

3. The system for predicting long-term prognosis of a liver disease patient according to claim 1, wherein The prediction model for the death of patients with acute decompensated cirrhosis and acute-on-chronic liver failure one year after admission is used to predict the one-year survival rate.

4. A method for predicting long-term prognosis of a patient with liver disease, characterized by, The prognosis prediction method for decompensated cirrhosis and acute-on-chronic liver failure patients comprises the following steps: Step S1, collect clinical data of patients with acute decompensated cirrhosis and acute-on-chronic liver failure, including anthropometric data, vital signs, laboratory data, and albumin binding capacity obtained by specific detection; the albumin binding capacity obtained by specific detection is evaluated by measuring the residual binding amount of specific fluorescent marker dansyl sarcosine to albumin binding site II through experiments; Step S2, construct a competing risk model for predicting the death of patients with acute decompensated cirrhosis and acute-on-chronic liver failure one year after admission, the prediction model is constructed by using a competing risk model, and a dynamic nomogram is used to visualize the prediction model to make a convenient prediction tool; the prediction model formula is: Albumin binding capacity model = -0.0136*albumin binding capacity-0.0210*mean arterial pressure+0.0767*white blood cell count+0.7374*international normalized ratio; wherein the albumin binding capacity (%) = fluorescence in standard ultrafiltrate / florescence in sample ultrafiltrate*100; Step S3, evaluate the prediction performance of the model by using a receiver operating characteristic curve.

5. The method of long-term prognosis prediction of liver disease patient according to claim 4, characterized in that, The prognostic factors of the competing risk model for the death of patients with acute decompensated cirrhosis and acute-on-chronic liver failure one year after admission include albumin binding capacity, mean arterial pressure, white blood cell count, and international normalized ratio.

6. The method of long-term prognosis prediction of liver disease patient according to claim 4, characterized in that, The prognosis prediction method is used to predict the 1-year survival rate. The prognosis prediction method is used to predict the 1-year survival rate.

Images