Voriconazole-related liver injury risk prediction model and construction method thereof

Abstract

The invention belongs to the technical field of adverse drug reaction prediction, particularly relates to a voriconazole-related liver injury risk prediction model and a construction method thereof, and discloses a Nomogram-based voriconazole-related liver injury risk prediction model. Comprising the following steps: collecting clinical data of a patient to form a data set, and judging whether the voriconazole-related liver injury occurs or not according to a diagnostic standard; randomly dividing the data set into a training set and a verification set according to a ratio of 7: 3; obtaining independent risk factors of the voriconazole related liver injury from the training set through a statistical method; establishing a Nomogram prediction model according to the independent risk factors; analyzing the evaluation model through an ROC curve, a calibration curve and a DCA curve; performing internal verification on the model through a Bootstrap method; and performing external verification on the model by using the verification set. Through the established prediction model, the occurrence risk of the voriconazole-related liver injury can be accurately predicted.

Description

Technical Field

[0001] The present invention belongs to the technical field of adverse drug reaction prediction, and specifically relates to a risk prediction model for voriconazole-related liver injury and a construction method thereof. Background Art

[0002] Voriconazole (VCZ) is a triazole broad-spectrum antifungal drug that achieves its antifungal effect by inhibiting the biosynthesis of ergosterol in fungal cells. It has been recommended as a first-line drug for the treatment and prevention of invasive fungal disease (IFD). Liver injury is one of the most common adverse reactions of voriconazole. Previous studies have shown that among 363 critically ill patients treated with VCZ, 101 patients developed VCZ-related ≥ grade 2 hepatotoxicity. Another study reported that the incidence of abnormal liver biochemical indicators related to VCZ was 12.2%.

[0003] There have been some studies on the influencing factors of VCZ-related liver injury. Most studies have shown that VCZ-related liver injury is significantly associated with VCZ trough concentration (C VCZmin ). In addition, some studies have shown that VCZ-related liver injury may be related to CYP2C19 gene polymorphism, inflammatory status and concomitant medication, but not to factors such as gender, age, and weight. However, univariate analysis can only show the impact of a single factor on liver injury, and clinical patients are often in a complex environment, including different pathophysiological states, different treatment environments, and even different living habits. Therefore, it is necessary to integrate the influence of multiple factors and establish a risk prediction model to help doctors judge the risk probability of VCZ-related liver injury in patients early, so as to formulate the correct treatment plan. However, the current risk prediction model has rarely been reported in terms of adverse reactions.

[0004] Voriconazole-N-oxide (VNO) is the main metabolite of voriconazole. In recent years, VNO has received extensive attention, and some studies have proposed the contribution of VNO plasma concentration measurement to VCZ therapeutic drug monitoring (TDM). However, the role of VNO plasma trough concentration and metabolic ratio (MR) (ratio of VNO to VCZ trough concentration) in voriconazole liver injury remains unclear. In addition, the liver is an important organ for lactate metabolism, and approximately 65% ​​of lactate is metabolized by the liver. Hyperlactatemia may increase the burden on the liver and further damage liver function. In turn, liver dysfunction can also lead to reduced metabolic excretion of lactate, forming a vicious cycle. In short, the effect of lactate on liver function is significant, and liver damage can also affect lactate metabolism. However, there are currently no reports on the effect of lactate levels on voriconazole-related liver injury.

[0005] Based on the above problems and defects in the prior art, the present invention aims to establish a more convenient and accurate prediction model for voriconazole-related liver injury based on Nomogram, so as to carry out early identification and intervention of patients at high risk of liver injury and provide a reference for rational clinical drug use. Summary of the invention

[0006] In view of this, in order to solve the problems existing in the prior art, the first object of the present invention is to provide a method for constructing a risk prediction model for voriconazole-related liver injury.

[0007] In order to achieve the above purpose, the present invention adopts the following technical solution:.

[0008] A method for constructing a risk prediction model for voriconazole-related liver injury, collecting and screening clinical data of patients treated with voriconazole to form a data set, and the data set is divided into a training set and a validation set at a ratio of 7:3; the training set is used to construct a nomogram prediction model for voriconazole-related liver injury, and the ROC curve, calibration curve, and DCA curve are used to analyze and evaluate the model; the Bootstrap method is used to internally validate the model; and the validation set is used to externally evaluate the model.

[0009] It is worth noting that in order to comprehensively consider the influence of multiple factors, establish a risk prediction model, and help doctors judge the risk probability of VCZ-related liver damage in patients at an early stage, so as to formulate a correct treatment plan, the present invention discloses a risk prediction model for voriconazole-related liver damage based on Nomogram, including: collecting clinical data of patients and judging whether voriconazole-related liver damage occurs according to diagnostic criteria; randomly dividing the data set into a training set and a validation set at a ratio of 7:3; obtaining independent risk factors for voriconazole-related liver damage in the training set by statistical methods; establishing a Nomogram prediction model for voriconazole-related liver damage based on independent risk factors; evaluating the model by ROC curve, calibration curve, and DCA curve; internally verifying the model by the Bootstrap method; and externally verifying the model by the validation set. The established prediction model can accurately predict the risk of voriconazole-related liver damage. The present invention provides a simple and accurate tool for clinicians to judge the risk of liver damage in patients treated with voriconazole, and can provide a reference for rational clinical drug use.

[0010] Furthermore, the specific steps of the method for constructing the risk prediction model of voriconazole-related liver injury include:

[0011] S1: Collect clinical data of patients treated with voriconazole, including demographics and laboratory test and treatment details;

[0012] S2: Establish inclusion and exclusion criteria, screen the collected patient data one by one, generate a data set for patients who meet the criteria, and randomly divide the data set into a training set and a validation set at a ratio of 7:3; group the data set according to the diagnostic criteria for voriconazole-related liver injury, with patients who meet the diagnostic criteria being the liver injury group and those who do not meet the diagnostic criteria being the non-liver injury group;

[0013] S3: Univariate and binary logistic regression analysis was used to screen out independent risk factors for voriconazole-related liver injury in the liver injury group and the non-liver injury group in the training set;

[0014] S4: Construct a nomogram prediction model for predicting the risk of voriconazole-related liver injury based on the information obtained in step S3.

[0015] Furthermore, in step S1, the clinical data collected include: age, gender, height, weight, BMI, smoking history, drinking history, dosage, medication days, drug source, severe pneumonia, infection, hypertension, diabetes, heart disease, sepsis, hypoalbuminemia, tumor, Aspergillus, Candida, mixed infection, infection with unknown pathogens, voriconazole trough concentration (C VCZmin ), voriconazole N-oxide concentration (C VNO ), metabolic ratio (C VNO / CVCZmin ), ALB (plasma albumin), UREA (urea), CREA (creatinine), UA (uric acid), EGFR (glomerular filtration rate), WBC (white blood cell count), NEUT% (neutrophil percentage), LYM% (lymphocyte percentage), NEUT# (absolute neutrophil count), LYM# (absolute lymphocyte count), RBC (red blood cell count), HGB (hemoglobin), PLT (platelet count), LA (lactic acid), CRP (C-reactive protein), PCT (procalcitonin), IL-6 (interleukin-6), antibacterial drugs, proton pump inhibitors, and glucocorticoids; in addition, the results of liver function biochemical tests before and after medication were collected to determine whether the patient had voriconazole-related liver injury.

[0016] It is worth noting that in the present invention, the missing values ​​of the three variables CRP, PCT, and IL-6 were greater than 25%, so these three variables were eliminated and not included in the subsequent analysis.

[0017] Furthermore, in step S2, the inclusion and exclusion criteria established include:

[0018] (1) Inclusion criteria: treatment with voriconazole; age ≥ 18 years; voriconazole treatment drug monitoring; complete clinical data and examination results;

[0019] (2) Exclusion criteria: patients taking concomitant medications that affect voriconazole concentrations; patients diagnosed with liver diseases such as cirrhosis, drug-induced liver injury, viral hepatitis, liver cancer, or liver failure at admission; patients taking voriconazole and receiving hemodialysis or other treatments that affect blood volume; patients with non-steady-state trough concentrations; pregnant or lactating women; and patients with incomplete clinical data and examination results.

[0020] Furthermore, in step S2, the diagnostic criteria for voriconazole-related liver injury are:

[0021] If the patient has no abnormal liver function before taking the drug, (1) ALT ≥ 5×ULN; (2) ALP ≥ 2×ULN (especially with elevated GGT and excluding the elevated ALP level caused by bone disease); (3) ALT ≥ 3×ULN and TBil ≥ 2×ULN, and meets one of the above three conditions, acute drug-induced liver injury can be diagnosed; if the patient has abnormal liver function before taking the drug, and after taking the drug, the liver function is more than 1 times higher than the available average level before taking the drug and cannot be explained by underlying liver disease, drug-induced liver injury is suspected. After excluding other causes of abnormal liver function biochemical indicators, drug-induced liver injury can be diagnosed; the Roussel Uclaf Causality Assessment Method (RUCAM) is used to evaluate the causal relationship between voriconazole and liver injury. If the RUCAM scale score is ≥ 6 points, it is considered that the patient's liver injury is caused by voriconazole.

[0022] Furthermore, the step S3 includes:

[0023] S31: Univariate analysis was performed using SPSS 27.0. Qualitative data were expressed as cases (%). Inter-group comparison was performed using χ 2 Test; Normally distributed quantitative data are expressed as mean ± standard deviation The t test was used; the non-normally distributed quantitative data were expressed as median (quartile) [M (P25, P75)], and the Mann-Whitney U test was used; 9 risk factors were finally screened out, including: C VCZmin , C VNO , metabolic ratio (MR), sepsis, eGFR, WBC, NEUT#, LA, and baseline liver function (P < 0.05);

[0024] S32: SPSS27.0 was used to perform Pearson correlation analysis on the risk factors screened in step S31; the results of the correlation analysis showed that NEUT%, WBC, C VCZmin , C VNO , MR have a strong correlation, excluding NEUT%, C VCZmin , C VNO , finally included WBC, MR, LA, eGFR, sepsis and baseline liver function for binary logistic regression analysis;

[0025] S33: SPSS27.0 was used to perform binary logistic regression analysis on the risk factors screened in step S32; the results showed that independent risk factors included MR, LA and baseline liver function (P < 0.05).

[0026] It is worth noting that the present invention proposes for the first time to use MR, LA and baseline liver function (P < 0.05) to jointly predict voriconazole-related liver injury. The application of MR can better reflect the patient's liver metabolic capacity. The metabolism of LA in the body mainly depends on the liver. Excessive production of LA may exceed the liver's metabolic capacity, ultimately leading to abnormal liver function. At the same time, patients with abnormal baseline liver function may also be more susceptible to drug-induced liver injury. When taking medication, special attention should be paid to changes in liver function indicators. The present invention simultaneously combines three independent risk factors to predict voriconazole-related liver injury, aiming to accurately identify the probability of liver injury in patients treated with voriconazole in the early stage of treatment, provide a scientific reference for the adjustment of subsequent treatment plans, and further improve the safety of voriconazole treatment for patients.

[0027] Furthermore, in step S4, the independent risk factors were included based on the results of binary logistic regression analysis and a Nomogram model was constructed using RStudio 4.4.1.

[0028] Furthermore, the specific steps of the method for constructing the risk prediction model of voriconazole-related liver injury also include model evaluation, namely:

[0029] S5: performing internal analysis on the Nomogram prediction model established in step S4, including: ROC curve analysis, calibration curve analysis, and DCA curve analysis;

[0030] S6: internally verify the Nomogram prediction model established in step S4 by using the Bootstrap method (n=1000 times);

[0031] S7: Use the validation set to perform external evaluation on the Nomogram prediction model established in step S4.

[0032] It is worth noting that the methods used in constructing the model in the present invention include: using the hospital's electronic medical record system (HIS system) to query the patient's clinical data; using Excel to statistically analyze the patient's clinical data; using SPSS27.0 to perform balance test, univariate analysis and correlation analysis; and using RStudio 4.4.1 to construct, evaluate and verify the model.

[0033] Wherein, the step S5 comprises: the Nomogram model established in step S4 is evaluated by drawing an ROC curve to evaluate the discrimination of the model, drawing a calibration curve to evaluate the calibration of the model, and drawing a decision analysis curve to evaluate the clinical net benefit of the model.

[0034] The step S6 includes: the Nomogram model established in step S4 is internally validated by the Bootstrap method (n=1000 times); the mean ± standard deviation of the AUC of the logistic regression model established by calculating 1000 simulated data sets ( ) was used to compare the AUC of the original model to evaluate the internal validation results of the model.

[0035] The step S7 includes: placing the Nomogram model established in step S4 into the validation set data to evaluate the model (ie, ROC curve, calibration curve and DCA curve) to evaluate the performance of the model in the external data set.

[0036] Therefore, the present invention establishes a Nomogram model of voriconazole-related liver injury, and evaluates the model through ROC curve, calibration curve and DCA curve, performs internal validation of the model through the Bootstrap method (n=1000 times); performs external validation of the model through a validation set; and finally applies the established Nomogram model to specific clinical patients.

[0037] The second object of the present invention is to provide a risk prediction model for voriconazole-related liver injury obtained by the construction method as described above.

[0038] A risk prediction model for voriconazole-related liver injury, wherein the independent risk factors of the risk prediction model for voriconazole-related liver injury include the metabolic ratio of voriconazole, the patient's lactate level and baseline liver function status, and the calculation formula for the probability of voriconazole-related liver injury predicted by the model is:

[0039] Logit(P)=-1.7191-4.4969×MR+0.3638×LA+1.8532×BLF

[0040] Wherein: P represents the probability of voriconazole-related liver injury; MR represents the metabolic ratio of voriconazole (voriconazole N-oxide concentration / voriconazole trough concentration); LA represents lactic acid; BLF represents baseline liver function, with normal baseline liver function recorded as 0 and abnormal baseline liver function recorded as 1.

[0041] When applied, the Nomogram model disclosed in the present invention is used to input clinical data of clinical patients, obtain the probability of voriconazole-related liver injury in the patient, and draw a case nomogram. Independent risk factors are used as variables, and the scores corresponding to each variable under different values ​​are calculated, and the total score after adding up the scores of all variables is given to predict the probability of voriconazole-related liver injury in each patient.

[0042] Compared with the univariate analysis in the prior art that can only indicate the impact of a single factor on liver damage, the risk prediction model for voriconazole-related liver damage based on Nomogram disclosed in the present invention combines multiple risk factors, and uses metabolic ratio, lactate and baseline liver function to jointly predict voriconazole-related liver damage, aiming to accurately identify the probability of occurrence of liver damage in patients treated with voriconazole in the early stage of treatment, and provides a simple and accurate tool for clinicians to judge the risk of liver damage in patients treated with voriconazole, which can provide a reference for rational clinical drug use and further enhance the accuracy and applicability of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 The present invention is a flowchart of a method for constructing a risk prediction model for voriconazole-related liver injury.

[0045] Figure 2 It is a Nomogram prediction model diagram in Example 1 of the present invention, wherein MR represents the voriconazole metabolic ratio (voriconazole N-oxide concentration / voriconazole trough concentration); LA represents lactic acid; BLF represents baseline liver function; Points represents the score corresponding to each item; Total Points represents the total score of the patient; Linear Predictor represents the linear prediction value; risk represents the risk of voriconazole-related liver injury. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] The word "embodiment" used here as an "exemplary" does not necessarily mean that any embodiment described is superior to or better than other embodiments. Unless otherwise specified, the performance index tests in the embodiments of this application are performed using conventional test methods in the art. It should be understood that the terms described in this application are only used to describe specific implementation methods and are not used to limit the content disclosed in this application.

[0048] Unless otherwise specified, the technical and scientific terms used in this document have the same meanings as commonly understood by ordinary technicians in the technical field to which this application belongs; other experimental methods and technical means not specifically specified in this application refer to experimental methods and technical means commonly used by ordinary technicians in this field.

[0049] In order to better illustrate the content of the present application, numerous specific details are provided in the specific examples below. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In the embodiments, some methods, means, instruments, equipment, etc. well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present application.

[0050] Under the premise of no conflict, the technical features disclosed in the embodiments of the present application can be combined arbitrarily, and the resulting technical solutions belong to the contents disclosed in the embodiments of the present application.

[0051] The present invention belongs to the technical field of drug adverse reaction prediction, and specifically relates to a risk prediction model for voriconazole-related liver injury and a method for constructing the same. The present invention discloses a risk prediction model for voriconazole-related liver injury based on Nomogram, including: collecting clinical data of patients and judging whether voriconazole-related liver injury occurs according to diagnostic criteria; randomly dividing the data set into a training set and a validation set at a ratio of 7:3; obtaining independent risk factors for voriconazole-related liver injury in the training set by statistical methods; establishing a Nomogram prediction model for voriconazole-related liver injury based on independent risk factors; evaluating the model by ROC curve, calibration curve, and DCA curve; internally validating the model by the Bootstrap method; and externally validating the model by the validation set. The established prediction model can accurately predict the risk of voriconazole-related liver injury. The present invention provides a simple and accurate tool for clinicians to judge the risk of liver injury in patients treated with voriconazole, and can provide a reference for rational clinical drug use.

[0052] In order to better understand the present invention, the present invention is further specifically described below through the following examples, but it should not be understood as a limitation of the present invention. Some non-essential improvements and adjustments made by technicians in this field based on the above invention content are also considered to fall within the protection scope of the present invention.

[0053] Example 1

[0054] A risk prediction model for voriconazole-related liver injury and its construction method ( Figure 1 )

[0055] S1: Systematic data of patients who received voriconazole and underwent TDM in Ningxia Medical University General Hospital were collected, including demographics, laboratory tests, and treatment details; clinical data collected included age, gender, height, weight, BMI, smoking history, drinking history, dosage, medication days, drug source, severe pneumonia, infection, hypertension, diabetes, heart disease, sepsis, hypoalbuminemia, tumor, Aspergillus, Candida, mixed infection, infection with unknown pathogens, voriconazole trough concentration (C VCZmin ), voriconazole N-oxide concentration (C VNO ), metabolic ratio (C VNO / C VCZmin), ALB, UREA, CREA, UA, EGFR, WBC, NEUT%, LYM%, NEUT#, LYM#, RBC, HGB, PLT, LA, CRP, PCT, IL-6, antimicrobial drugs, proton pump inhibitors, and glucocorticoids; among them, the missing values ​​of CRP, PCT, and IL-6 were greater than 25%, so these three variables were eliminated and not included in the subsequent analysis; in addition, the results of biochemical tests of liver function before and after medication were collected to determine whether the patients had voriconazole-related liver injury; a total of 176 patients' clinical data were collected.

[0056] S2: Inclusion and exclusion criteria were established according to the present invention, and the collected patient data were screened one by one. A data set was generated for patients who met the criteria. The data set was randomly divided into a training set and a validation set at a ratio of 7:3, and the data set was grouped according to the diagnostic criteria for voriconazole-related liver injury. Patients who met the diagnostic criteria were grouped into the liver injury group, and those who did not meet the diagnostic criteria were grouped into the non-liver injury group.

[0057] The inclusion and exclusion criteria for patients were as follows:

[0058] (1) Inclusion criteria:

[0059] a) Treatment with voriconazole;

[0060] b) Age>18 years old;

[0061] c) Conduct voriconazole therapeutic drug monitoring;

[0062] d) The clinical data and examination results are complete.

[0063] (2) Exclusion criteria:

[0064] a) Patients are taking drugs that affect voriconazole concentration;

[0065] b) Patients diagnosed with liver diseases at the time of admission, such as cirrhosis, drug-induced liver injury, viral hepatitis, liver cancer or liver failure;

[0066] c) Patients taking voriconazole undergo hemodialysis or other treatments that affect blood volume;

[0067] d) non-steady-state valley concentration;

[0068] e) Pregnant or breastfeeding women;

[0069] f) Clinical data and examination results are incomplete.

[0070] According to the inclusion and exclusion criteria, a total of 17 patients were excluded (6 patients underwent CRRT or hemodialysis treatment; 3 patients were diagnosed with liver disease at admission; 4 patients had concurrent drugs that interacted with voriconazole; 3 patients were aged <18 years; 1 patient had non-steady-state trough concentrations). Finally, a data set was generated for 159 patients, 111 patients (69.8%) were used as training sets to build the model, and 48 patients (30.2%) were used as validation sets for external validation. The training set and validation set patients were compared for balance and differences. The results showed that there was no significant difference between the training set and the validation set. The validation set can be used to validate the constructed prediction model. The results are shown in Table 1.

[0071] Table 1 Baseline characteristics and balance test of patients in the training set and validation set

[0072]

[0073]

[0074] Note: BMI: body mass index; C VCZmin : voriconazole trough concentration; C VNO : voriconazole N-oxide concentration; MR: metabolic ratio (voriconazole N-oxide concentration / voriconazole trough concentration); ALB: albumin; UREA: urea; CREA: creatinine; UA: uric acid; eGFR: estimated glomerular filtration rate; WBC: white blood cell count; NEUT%: neutrophil percentage; LYM%: lymphocyte percentage; NEUT#: absolute neutrophil count; LYM#: absolute lymphocyte count; RBC: red blood cell count; HGB: hemoglobin; PLT: platelet count; LA: lactate.

[0075] The diagnostic criteria for voriconazole-related liver injury are as follows:

[0076] If the patient has no abnormal liver function before taking the drug, (1) ALT ≥ 5 × ULN; (2) ALP ≥ 2 × ULN (especially with elevated GGT and excluding ALP level caused by bone disease); (3) ALT ≥ 3 × ULN and TBil ≥ 2 × ULN, if one of the above three conditions is met, acute drug-induced liver injury can be diagnosed; if the patient has abnormal liver function before taking the drug, and after taking the drug, the liver function is more than 1 times higher than the average level available before taking the drug and cannot be explained by underlying liver disease, drug-induced liver injury is suspected. If other causes of abnormal liver function biochemical indicators are excluded, drug-induced liver injury can be diagnosed; use Roussel The Uclaf Causality Assessment Method (RUCAM) evaluated the causal relationship between voriconazole and liver injury. A RUCAM scale score ≥ 6 points was considered to be caused by voriconazole in the patient. According to the diagnostic criteria, 19 patients (17.1%) developed voriconazole-related liver injury in the training set, and 92 patients (82.9%) did not develop voriconazole-related liver injury. In the validation set, 7 patients (14.6%) developed voriconazole-related liver injury, and 41 patients (85.4%) did not develop voriconazole-related liver injury.

[0077] S3: Univariate and binary logistic regression analysis was used to screen out independent risk factors affecting voriconazole-related liver injury in the liver injury group and the non-liver injury group.

[0078] Specifically, the S3 steps are as follows:

[0079] S31: Univariate analysis was performed using SPSS 27.0. Qualitative data were expressed as cases (%). Inter-group comparison was performed using χ 2 Normally distributed quantitative data are expressed as mean ± standard deviation. The t test was used; the non-normally distributed quantitative data were expressed as median (quartile) [M(P25, P75)] and the Mann-Whitney U test was used; among them, P < 0.05 was considered to indicate that the independent risk factor had a significant difference between the two groups.

[0080] The results of univariate analysis are shown in Table 2. VCZmin , C VNO , metabolic ratio (MR), sepsis, eGFR, WBC, NEUT#, LA, and baseline liver function (P<0.05), indicating that there were significant differences between the voriconazole-related liver injury group and the non-liver injury group.

[0081] Table 2 Univariate analysis of voriconazole-related liver injury

[0082]

[0083]

[0084] Note: BMI: body mass index; C VCZmin : voriconazole trough concentration; C VNO : voriconazole N-oxide concentration; MR: metabolic ratio (voriconazole N-oxide concentration / voriconazole trough concentration); ALB: albumin; UREA: urea; CREA: creatinine; UA: uric acid; eGFR: estimated glomerular filtration rate; WBC: white blood cell count; NEUT%: neutrophil percentage; LYM%: lymphocyte percentage; NEUT#: absolute neutrophil count; LYM#: absolute lymphocyte count; RBC: red blood cell count; HGB: hemoglobin; PLT: platelet count; LA: lactate.

[0085] S32: SPSS27.0 was used for Pearson correlation analysis; in order to avoid the influence of collinearity, the risk factors with differences in univariate analysis were subjected to Pearson correlation analysis. The results showed that NEUT%, WBC, C VCZmin , C VNO , MR have a strong correlation, excluding NEUT%, C VCZmin , C VNO Finally, WBC, MR, LA, eGFR, sepsis and baseline liver function were included in binary logistic regression analysis.

[0086] S33: The risk factors screened in step S32 were subjected to binary logistic regression analysis using SPSS27.0; the results showed that independent risk factors included MR, LA and baseline liver function, as shown in Table 3.

[0087] Table 3 Binary Logistic regression analysis of voriconazole-related liver injury

[0088]

[0089] Note: β: regression coefficient; SE(β): standard error of regression coefficient; OR: odds ratio.

[0090] S4: Based on the information obtained in step S3, a nomogram model was constructed to predict the risk of voriconazole-related liver injury. The results are shown in Figure 2 .

[0091] Specifically, RStudio 4.4.1 was used to construct a Nomogram model. In the constructed Nomogram model, MR, LA and baseline liver function were used as independent risk factors as variables. The scores corresponding to each variable at different values ​​were calculated, and the total score after adding up the scores of all variables was calculated to predict the probability of each patient having substandard trough concentration of voriconazole.

[0092] S5: Investigate and evaluate the Nomogram model established in step S4 to obtain the best Nomogram model for predicting voriconazole-related liver injury.

[0093] Specifically, step S5 is as follows:

[0094] S51: The receiver operating characteristic curve (ROC curve) is used to evaluate the discrimination of the nomogram model: The ROC curve is used to evaluate the effect of certain indicators in distinguishing two different categories of samples. The curve is drawn with the true positive rate (sensitivity) as the vertical axis and the false positive rate (1-specificity) as the horizontal axis. By drawing the ROC curve and calculating the area under the curve (AUC), the classification or diagnostic effects of different indicators can be compared. The closer the AUC is to 1, the better the classification or diagnostic effect of the indicator.

[0095] The constructed Nomogram model had a consistency index C-index of 0.875 and an AUC of 0.875 for predicting voriconazole-related liver injury, indicating that the constructed Nomogram model had good accuracy and discrimination.

[0096] S52: Use the calibration curve to evaluate the calibration of the Nomogram model: The calibration curve is used to evaluate the difference between the model prediction results and the actual results. It is a curve drawn with the predicted probability as the horizontal axis and the actual probability as the vertical axis; the Ideal line represents the ideal reference line (the model prediction probability and the actual probability are completely consistent), Apparent is the current model curve, and Bias-corrected is the curve calculated by multiple sampling (Bootstrap = 1000 times). Therefore, the higher the consistency between the Apparent line and the Bias-corrected line, the closer it is to the Ideal line, which means that the calibration of the Nomogram model is better.

[0097] The predicted probability of the constructed Nomogram model was consistent with the actual probability. Visual inspection showed that the constructed Nomogram model had good calibration, indicating that the model can accurately predict the occurrence of voriconazole-related liver injury.

[0098] S53: Use decision curve (DCA curve) to evaluate the clinical benefit of the Nomogram model: The DCA curve is a method to evaluate the application value of the prediction model in actual clinical decision-making. It is a curve drawn with threshold probability as the horizontal axis and net benefit as the vertical axis. In the figure, the None line means no intervention for anyone, the All line means intervention for anyone, and the train line represents the overall net benefit of the Nomogram model within the threshold range. Therefore, the larger the area from the train line to the None line and the All line, the wider the range, which means that the clinical benefit of the Nomogram model is better.

[0099] The constructed Nomogram model has good clinical benefits in the range of 0.1-1.0, indicating that the model has good clinical practicality.

[0100] Thus, a risk prediction model for voriconazole-related liver injury was obtained, and the formula for establishing the model was:

[0101] Logit(P)=-1.7191-4.4969×MR+0.3638×LA+1.8532×BLF,

[0102] Wherein: P represents the probability of voriconazole-related liver injury; MR represents the metabolic ratio of voriconazole (voriconazole N-oxide concentration / voriconazole trough concentration); LA represents lactic acid; BLF represents baseline liver function, with normal baseline liver function recorded as 0 and abnormal baseline liver function recorded as 1.

[0103] A nomogram model was constructed based on this formula, which has good discrimination, calibration and clinical practicality.

[0104] Furthermore, the present invention also verifies the clinical utility value of the established Nomogram model:

[0105] (1) The Bootstrap method is used to internally validate the Nomogram model. The Bootstrap method generates a large number of virtual “resampled” data sets by repeatedly extracting from the original data set (with replacement), and uses these data sets to simulate possible situations.

[0106] The constructed Nomogram model was bootstraped 1000 times to obtain the AUC of 1000 simulated data sets. It is 0.875±0.047, and the AUC of the ROC curve corresponding to the original model is 0.875, indicating that the model has good internal validation results.

[0107] (2) Use the validation set to externally validate the Nomogram model; that is, put the established Nomogram model into the validation set data to evaluate the model (i.e., ROC curve, calibration curve, and DCA curve) to evaluate the performance of the model in the external data set.

[0108] Therefore, external validation of the constructed Nomogram model showed good discrimination, calibration and clinical practicality.

[0109] (3) Predict voriconazole-related liver injury in a single patient based on the Nomogram model established in step S4. VCZmin :9.07mg / L、C VNO: 0.86mg / L, calculated MR: 0.09, lactic acid (LA): 7.2mmol / L, abnormal baseline liver function (LFT) at admission; ALT before medication: 96.5U / L; ALT on the 7th day after medication: 667.8U / L, according to the diagnostic criteria for voriconazole-related liver injury, it met the diagnosis of voriconazole-related liver injury.

[0110] The constructed Nomogram model showed a comprehensive score of 220 points, and the risk of predicted voriconazole-related liver injury was 92.5%. The results showed that the constructed Nomogram model had a high prediction accuracy for voriconazole-related liver injury in patients.

[0111] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for constructing a risk prediction model for voriconazole-related liver injury, characterized in that: Clinical data of patients treated with voriconazole were collected and screened to form a data set, which was divided into a training set and a validation set at a ratio of 7:

3. The training set was used to construct a nomogram prediction model for voriconazole-related liver injury, and the ROC curve, calibration curve, and DCA curve were used to analyze and evaluate the model. The Bootstrap method was used to internally validate the model. The validation set was used for external evaluation of the model.

2. The construction method according to claim 1, characterized in that: The specific steps include: S1: Collect clinical data of patients treated with voriconazole, including demographics and laboratory test and treatment details; S2: Establish inclusion and exclusion criteria, screen the collected patient data one by one, generate a data set for patients who meet the criteria, and randomly divide the data set into a training set and a validation set at a ratio of 7:3; group the data set according to the diagnostic criteria for voriconazole-related liver injury, with patients who meet the diagnostic criteria being the liver injury group and those who do not meet the diagnostic criteria being the non-liver injury group; S3: Univariate and binary logistic regression analysis was used to screen out independent risk factors for voriconazole-related liver injury in the liver injury group and the non-liver injury group in the training set; S4: Construct a nomogram prediction model for predicting the risk of voriconazole-related liver injury based on the information obtained in step S3.

3. The construction method according to claim 2, characterized in that: In step S1, the collected clinical data include: age, gender, height, weight, BMI, smoking history, drinking history, dosage, medication days, drug source, severe pneumonia, infection, hypertension, diabetes, heart disease, sepsis, hypoalbuminemia, tumor, Aspergillus, Candida, mixed infection, infection with unknown pathogens, voriconazole trough concentration, voriconazole N-oxide concentration, metabolic ratio, ALB, UREA, CREA, UA, EGFR, WBC, NEUT%, LYM%, NEUT#, LYM#, RBC, HGB, PLT, LA, CRP, PCT, IL-6, antibacterial drugs, proton pump inhibitors, and glucocorticoids; in addition, the biochemical test results of liver function before and after medication are collected to determine whether the patient has voriconazole-related liver injury.

4. The construction method according to claim 2, characterized in that: In step S2, the inclusion and exclusion criteria established include: (1) Inclusion criteria: treatment with voriconazole; age ≥ 18 years; voriconazole treatment drug monitoring; complete clinical data and examination results; (2) Exclusion criteria: patients taking concomitant medications that affect voriconazole concentrations; patients diagnosed with liver diseases such as cirrhosis, drug-induced liver injury, viral hepatitis, liver cancer, or liver failure at admission; patients taking voriconazole and receiving hemodialysis or other treatments that affect blood volume; patients with non-steady-state trough concentrations; pregnant or lactating women; and patients with incomplete clinical data and examination results.

5. The construction method according to claim 2, characterized in that: In step S2, the diagnostic criteria for voriconazole-related liver injury are: If the patient has no abnormal liver function before taking the drug, (1) ALT ≥ 5 × ULN; (2) ALP ≥ 2 × ULN; (3) ALT ≥ 3 × ULN and TBil ≥ 2 × ULN, if one of the above three conditions is met, acute drug-induced liver injury can be diagnosed; If the patient has abnormal liver function before medication, and after medication, the liver function increases more than 1 times the average level before medication and cannot be explained by underlying liver disease, drug-induced liver injury is suspected. Drug-induced liver injury can be diagnosed after excluding other causes of abnormal liver function biochemical indicators; The Roussel Uclaf causality assessment method was used to evaluate the causal relationship between voriconazole and liver injury. A RUCAM scale score of ≥6 points was considered to be caused by voriconazole.

6. The construction method according to claim 2, characterized in that: The step S3 includes: S31: Univariate analysis was performed using SPSS 27.0, qualitative data were presented using case studies, and inter-group comparisons were performed using χ 2 Test; Normally distributed quantitative data were expressed as mean ± standard deviation, and t test was used; non-normally distributed quantitative data were expressed as median [M (P25, P75)], and Mann-Whitney U test was used; 9 risk factors were finally screened out, including: C VCZmin , C VNO , metabolic ratio, sepsis, eGFR, WBC, NEUT#, LA, and baseline liver function (P < 0.05); S32: SPSS27.0 was used to perform Pearson correlation analysis on the risk factors screened in step S31; the results of the correlation analysis showed that NEUT%, WBC, C VCZmin , C VNO , MR have a strong correlation, excluding NEUT%, C VCZmin , C VNO , finally included WBC, MR, LA, eGFR, sepsis and baseline liver function for binary logistic regression analysis; S33: SPSS27.0 was used to perform binary logistic regression analysis on the risk factors screened in step S32; the results showed that independent risk factors included MR, LA and baseline liver function (P < 0.05).

7. The construction method according to claim 2, characterized in that: In step S4, the independent risk factors were included based on the results of binary logistic regression analysis and a Nomogram model was constructed using RStudio 4.4.

1.

8. The construction method according to claim 2, characterized in that: The specific steps also include: S5: internal evaluation of the Nomogram prediction model established in step S4, including: ROC curve analysis, calibration curve analysis, DCA curve analysis; S6: internally verify the Nomogram prediction model established in step S4 by using the Bootstrap method (n=1000 times); S7: Use the validation set to perform external evaluation on the Nomogram prediction model established in step S4.

9. The risk prediction model for voriconazole-related liver injury obtained by the construction method according to any one of claims 1 to 8, characterized in that: The independent risk factors of the risk prediction model for voriconazole-related liver injury include the metabolic ratio of voriconazole, the patient's lactate level and baseline liver function status, and the calculation formula for the probability of voriconazole-related liver injury predicted by the model is: Logit(P)=-1.7191-4.4969×MR+0.3638×LA+1.8532×BLF, Wherein: P represents the probability of voriconazole-related liver injury; MR represents the metabolic ratio of voriconazole (voriconazole N-oxide concentration / voriconazole trough concentration); LA represents lactic acid; BLF represents baseline liver function, with normal baseline liver function recorded as 0 and abnormal baseline liver function recorded as 1.

Images