IgA nephropathy treatment effect prediction method based on NLR

Abstract

The invention relates to the technical field of nephropathy prediction, and discloses an NLR-based IgA nephropathy treatment effect prediction method, which comprises the following steps: collecting index factors of a patient; the index factors comprise an NLR value, a serum creatinine level, a uric acid level and a glomerular filtration rate; the index factors are preprocessed; screening a reserved index factor according to a first threshold value, constructing a final index change rate interval according to a second threshold value and the index change rate interval, screening out a final data set according to the final index change rate interval, obtaining the index factor of the patient in real time, inputting the trained prediction model, and outputting the response of the patient to glucocorticoid treatment; on the basis of the existing NLR, the serum creatinine level, the uric acid level and the glomerular filtration rate which have significant influence on the glucocorticoid treatment response and index data in a specific range are screened as a training set, and the prediction result of the subsequent patient on the glucocorticoid treatment response through the model trained by the training set is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of nephropathy prediction, and particularly to a method for predicting the treatment effect of IgA nephropathy based on NLR. Background Art

[0002] IgA nephropathy is defined as obvious IgA deposition in the glomerular mesangial region. IgA nephropathy is one of the most common primary glomerular diseases globally, with diverse clinical manifestations, including asymptomatic microscopic hematuria and even rapidly progressive nephritis. Approximately 15 - 20% of patients develop end-stage renal disease (ESRD) within 10 years after onset, and this proportion will reach 25 - 30% if not controlled. Therefore, these patients usually require alternative treatment. Currently, the most commonly used treatment methods for IgA nephropathy are angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs). For patients with persistent proteinuria greater than 1 gram per day, glucocorticoids may be beneficial. However, the effectiveness of glucocorticoids varies greatly, and there are no parameters to predict their therapeutic response. The neutrophil-to-lymphocyte ratio (NLR) reveals an imbalance between two different immune systems. Neutrophil count (NC) reflects persistent inflammation, while lymphocyte count (LC) represents the regulatory immune pathway and malnutrition. Many evidences show that elevated NLR is associated with poor outcomes, such as malignant tumors, acute systemic inflammation, and heart diseases. However, no study has investigated the usefulness of NLR in patients with IgA nephropathy. Due to its anti-inflammatory properties, glucocorticoids are commonly used to treat IgA nephropathy.

[0003] In recent years, a number of studies have shown that the neutrophil-to-lymphocyte ratio (NLR), serum creatinine (Scr), uric acid (UA), and estimated glomerular filtration rate (eGFR) are potential predictors of glucocorticoid response. In particular, NLR has been proven to be an effective predictor of the response of IgAN patients to glucocorticoid treatment. However, most of the existing studies focus on the evaluation of single indicators and lack a method for comprehensive evaluation of multiple indicators.

[0004] A single indicator is difficult to comprehensively reflect the complex pathological state of patients. For example, NLR can reflect the inflammatory state, but cannot fully reflect renal function and metabolic state. Similarly, Scr and eGFR can reflect renal function, but cannot comprehensively reflect inflammatory and metabolic states. Therefore, although NLR is an effective predictor, it may not be able to comprehensively capture the pathological state of patients when used alone, resulting in inaccurate prediction results. Although serum creatinine (Scr), uric acid (UA), and estimated glomerular filtration rate (eGFR) are potential predictors of glucocorticoid response, the impact of these indicators on the prediction results is not as significant as that of NLR when used alone. Therefore, relying solely on these indicators for prediction may not have sufficient accuracy. However, by comprehensively evaluating multiple biomarkers such as NLR, Scr, UA, and eGFR, more comprehensive patient condition information can be provided, which helps to improve the accuracy and reliability of prediction. Comprehensive evaluation of multiple indicators can provide more comprehensive information, but how to effectively screen and integrate multiple indicators, reduce noise data, and improve the robustness of the model remains a challenge. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a method for predicting the treatment effect of IgA nephropathy based on NLR, including: S1. Collect the indicator factors of patients; the indicator factors include: NLR value, serum creatinine level, uric acid level, and glomerular filtration rate; S2. Preprocess the indicator factors; S3. Obtain a training set from the preprocessed indicator factors; S31. Analyze each preprocessed indicator factor through the ROC curve to obtain the area under the working characteristic curve of each preprocessed indicator factor; according to the area under the working characteristic curve of each preprocessed indicator factor, determine the prediction value of each preprocessed indicator factor; use the prediction value of the NLR value as the reference value; use the prediction values of the serum creatinine level, uric acid level, and glomerular filtration rate as the screening values; S32. Analyze the preprocessed NLR value through the ROC curve to obtain the optimal cut-off value of NLR; S33. Construct a first threshold according to the reference value, the optimal cut-off value of NLR, and the screening values; S34. Screen and reserve indicator factors according to the first threshold; S35. Calculate the index change rate of the prediction value of the reserved indicator factor and the reference value; calculate the standard deviation and mean according to the index change rate, and obtain the index change rate interval according to the standard deviation and mean; S36. Construct a second threshold according to the prediction value of the reserved indicator factor, the reference value, and the optimal cut-off value of NLR; S37. Construct the final index change rate interval according to the second threshold and the index change rate interval, and screen out the final data set according to the final index change rate interval; S4. Construct a training set based on the data set, train the prediction model, and obtain the trained prediction model; S5. Obtain the index factors of the patient in real time, input the trained prediction model, and output the patient's response to glucocorticoid treatment.

[0006] Further, the calculation formula of the first threshold is: ; In the formula, L 1 represents the first threshold, Y NLR represents the predictive value of the NLR value, that is, the benchmark value, M represents the best cut-off value of NLR, Y si represents the predictive value of the i-th screening value, and n represents the maximum number of screening values.

[0007] Further, screen and reserve the index factors according to the first threshold. Specifically: compare each screening value with the first threshold respectively. If the current screening value is less than or equal to the first threshold, define the index factor corresponding to the current screening value as the reserved index factor; if the current screening value is greater than the first threshold, delete the current screening value and the index factor corresponding to the current screening value.

[0008] Further, the calculation formula of the second threshold is: ; In the formula, L 2 represents the second threshold, H j represents the predictive value of the j-th reserved index factor, M represents the best cut-off value of NLR, Y NLR represents the predictive value of the NLR value, and p represents the maximum number of reserved index factors.

[0009] Further, screening out the final data set according to the final index change rate interval specifically includes: determining the reserved index factors in the final index change rate interval and the predictive values corresponding to the reserved index factors according to the index change rates of the reserved index factors.

[0010] Further, the training set includes: the NLR value of each patient, the reserved index factor, the predictive value corresponding to the reserved index factor, the index change rate of the reserved index factor, the patient's response to glucocorticoid treatment; the patient's response to glucocorticoid treatment is used as the prediction label.

[0011] Further, construct the final index change rate interval [Q 上 , Q 下 , specifically as follows: Q 上 =C 上 ×L 2 ; Q 下 =C 下 ×L 2 ; Wherein, Q 上 represents the upper limit value of the final change rate range of the index, Q 下 represents the lower limit value of the final change rate range of the index, C 上 represents the upper limit value of the change rate range of the index, C 下 represents the lower limit value of the change rate range of the index, L 2 represents the second threshold value.

[0012] Furthermore, the NLR value is the ratio of neutrophils to lymphocytes.

[0013] The embodiments of the present invention have the following technical effects: Based on the NLR, which is the most effective predictor of the response to glucocorticoid treatment, the present invention screens the serum creatinine level, uric acid level, glomerular filtration rate, and index data within a specific range that also have a significant impact on the response to glucocorticoid treatment as the training set. The selected training set covers multi-dimensional indicators and can comprehensively reflect the patient's inflammation, metabolic status, renal function, and can comprehensively capture the patient's pathological state. The model trained by this training set is more accurate in predicting the subsequent response of patients to glucocorticoid treatment.

[0014] The present invention comprehensively considers: the benchmark value related to the most effective predictor of the response to glucocorticoid treatment, the optimal cut-off value of NLR, and the predictive value of other indicators to construct a first threshold. The indicators screened by the first threshold represent relatively high independence. If the predictive value of a certain indicator is greater than the first threshold, it means that the contribution of this indicator to the prediction model is small, and there is great uncertainty in distinguishing the treatment response, and it will not significantly improve the predictive ability of the model.

[0015] The present invention screens out independent indicators through the first threshold. Due to the influence of aspects such as the patient's physiology, environment, behavior, and medical intervention, the independent indicators fluctuate, resulting in a change in the predictive value and being greater than the first threshold. Therefore, not all the predictive values of independent indicators have a significant impact on the prediction result. The present invention constructs a second threshold based on the predictive value, benchmark value, and optimal cut-off value of NLR of independent indicators, screens out the predictive values and corresponding independent indicators that are effective or have a significant impact on the prediction result from the independent indicators, and applies them to the prediction model, significantly improving the prediction accuracy. Description of the Drawings

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0017] Figure 1 It is a flowchart of a method for predicting the therapeutic effect of IgA nephropathy based on NLR provided by an embodiment of the present invention; Figure 2 It is an ROC curve graph of various index factors provided by an embodiment of the present invention; Figure 3 It is a structural diagram of a system for predicting the therapeutic effect of IgA nephropathy based on NLR provided by an embodiment of the present invention. Specific Embodiments

[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0019] Figure 1 It is a schematic structural diagram of a method for predicting the therapeutic effect of IgA nephropathy based on NLR provided by an embodiment of the present invention. Refer to Figure 1 , specifically including: S1. Collect the index factors of the patient; the index factors include: NLR value, serum creatinine level, uric acid level, and glomerular filtration rate.

[0020] The NLR value is the ratio of neutrophils to lymphocytes.

[0021] Retrospective studies were conducted on the clinical data and laboratory test results of primary IgA patients admitted to the Department of Nephrology, the Second Hospital of Tianjin Medical University from 2000 to 2018.

[0022] The patient inclusion criteria are as follows: (1) Age ≥ 18 years old; (2) The pathological results conform to the IgA diagnostic criteria; (3) Persistent proteinuria > 1 g / d after 6 months of ACEI or ARB treatment. The exclusion criteria are as follows: (1) Secondary IgA; (2) End-stage renal disease; (3) Malignant tumor with or without metastasis; (4) Active infection; (5) Uncontrolled diabetes; (6) Coronary heart disease, heart failure; (7) Other diseases.

[0023] The glucocorticoid treatment regimen is as follows: Methylprednisolone is administered by intravenous injection at a dose of 1 g / day for three consecutive days, and the drug is given in the 1st, 3rd, and 5th months respectively; Prednisone is taken orally at a dose of 0.5 mg / kg / day every other day for a total of 6 months. The patient is followed up in the outpatient clinic every month, and remission is defined as a urinary protein quantification < 1 g / 24 h during the follow-up period.

[0024] Clinical data include gender, age, systolic blood pressure (SBP), diastolic blood pressure (DBP), body mass index (BMI), and the response to glucocorticoid treatment; Laboratory tests include albumin (ALB), uric acid (UA), serum creatinine (Scr), urinary protein quantification (Proteinuria), nocturia frequency (NC), and leakage urine volume (LC). The laboratory test results include that the patient achieves remission with glucocorticoid treatment and that the patient does not achieve remission with glucocorticoid treatment.

[0025] S2. Preprocess the said index factors.

[0026] As shown in Table 1 below, a preprocessing method for index factors is provided.

[0027] Table 1 Preprocessing method for index factors

[0028] S3. Obtain a training set from the preprocessed index factors.

[0029] Figure 2 It is an ROC curve graph of each index factor provided by an embodiment of the present invention. Refer to Figure 2 , S3 includes the following sub-steps: S31. Analyze each preprocessed index factor through the ROC curve to obtain the area under the working characteristic curve of each preprocessed index factor; According to the area under the working characteristic curve of each preprocessed index factor, determine the predictive value of each preprocessed index factor; Take the predictive value of the NLR value as the reference value; Take the predictive values of the serum creatinine level, uric acid level, and glomerular filtration rate as the screening values.

[0030] The prior art determines independent factors through the Cox proportional hazards model, which are the NLR value, serum creatinine level, uric acid level, and glomerular filtration rate in this embodiment. The Cox proportional hazards model outputs the hazard ratio (HR) of each variable and its 95% confidence interval, plots the ROC curve for each indicator, and calculates the area under the ROC curve (AUROC) to quantify the predictive ability of the model. The range of the AUROC value is from 0 to 1, and the larger the value, the stronger the predictive ability of the model. The above is the prior art and will not be elaborated in detail in this embodiment. This embodiment preferably uses the index area as the predictive value. Since NLR is the most effective predictor, the predictive value of NLR is used as the benchmark value in this embodiment to better evaluate the independent predictive value of other indicators. If the predictive value of a certain indicator is significantly lower than that of NLR, it indicates that the contribution of this indicator in prediction is small and it may not be an independent predictor. The high predictive value of NLR ensures the stability of the model. Even after introducing other indicators, NLR can still serve as a stable reference point to ensure that the overall performance of the model is not affected by the fluctuations of a single indicator. The predictive values of serum creatinine level, uric acid level, and glomerular filtration rate are used as screening values.

[0031] S32. Analyze the preprocessed NLR value according to the ROC curve to obtain the optimal NLR cut-off value.

[0032] In some embodiments, the optimal NLR cut-off value can be obtained by analyzing the ROC curve and the Youden index method, that is, according to the sensitivity and specificity in the ROC curve, using the Youden index = (sensitivity + specificity) - 1, and obtaining the optimal cut-off value when the Youden index is maximized.

[0033] Continue to combine Figure 2 , it can be concluded from the above ROC curve analysis that the remission rate of patients in the NLR < 2.43 group is significantly higher than that of patients in the NLR ≥ 2.43 group. Therefore, 2.43 is used as the optimal cut-off value in this embodiment.

[0034] S33. Construct a first threshold according to the benchmark value, the optimal NLR cut-off value, and the screening values; ; In the formula, L 1 represents the first threshold, Y NLR represents the predictive value of the NLR value, that is, the benchmark value, M represents the optimal NLR cut-off value, Y si represents the predictive value of the i-th screening value, and n represents the maximum number of screening values.

[0035] S34. Screen and reserve indicator factors according to the first threshold.

[0036] Screen the reserved index factors according to the first threshold, specifically: compare each screening value with the first threshold respectively. If the current screening value is less than or equal to the first threshold, define the index factor corresponding to the current screening value as the reserved index factor; if the current screening value is greater than the first threshold, delete the current screening value and the index factor corresponding to the current screening value.

[0037] In this embodiment, comprehensive consideration is given to: the benchmark value related to the most effective prediction factor for glucocorticoid treatment response, the best cut-off value of NLR, and the prediction value of other indicators, to construct a first threshold. The indicators screened out by the first threshold represent relatively high independence. If the prediction value of a certain indicator is greater than the first threshold, it means that the contribution of this indicator to the prediction model is small, and there is great uncertainty in distinguishing treatment responses, and it will not significantly improve the prediction ability of the model. In this embodiment, this indicator is directly excluded, and only the indicators that contribute more to the prediction model are retained.

[0038] S35. Calculate the index change rate of the prediction value of the reserved index factor and the benchmark value; calculate the standard deviation and mean according to the index change rate, and obtain the index change rate interval according to the standard deviation and mean.

[0039] By calculating the index change rate of the prediction value of the reserved index factor and the NLR benchmark value, the relative importance of each indicator relative to NLR can be evaluated. It helps to further screen out independent prediction factors that have a significant impact on the prediction result. The calculation of the change rate enables the prediction values of different indicators to be compared on the same scale, ensuring the standardization of the evaluation process.

[0040] S36. Construct a second threshold according to the prediction value of the reserved index factor, the benchmark value, and the best cut-off value of NLR.

[0041] ; In the formula, L 2 represents the second threshold, H j represents the prediction value of the jth reserved index factor, M represents the best cut-off value of NLR, Y NLR represents the prediction value of the NLR value, and p represents the maximum number of reserved index factors.

[0042] In this embodiment, independent indicators are screened out through the first threshold. Due to the influence of aspects such as physiology, environment, behavior, and medical intervention of patients, the independent indicators fluctuate, resulting in a change in the prediction value and being greater than the first threshold. Therefore, not all the prediction values of the independent indicators have a significant impact on the prediction result. The present invention constructs a second threshold based on the prediction value of the independent indicator, the benchmark value, and the best cut-off value of NLR, screens out the prediction values that are effective or have a significant impact on the prediction result and the corresponding independent indicators from the independent indicators, and applies them to the prediction model, significantly improving the prediction accuracy.

[0043] S37. Construct the final index change rate interval [Q 上 , Q 下 according to the second threshold and the index change rate interval, and screen out the final data set according to the final index change rate interval.

[0044] Q 上 = C 上 × L 2 ; Q 下 = C 下 × L 2 ; In the formula, Q 上 represents the upper limit value of the final index change rate interval, Q 下 represents the lower limit value of the final index change rate interval, C 上 represents the upper limit value of the index change rate interval, C 下 represents the lower limit value of the index change rate interval, and L 2 represents the second threshold.

[0045] The introduction of the second threshold can further optimize the index change rate interval, specifically: narrow the index change rate interval to obtain a smaller range interval to ensure that the data points within this interval have a significant impact on the treatment response.

[0046] Screen out the final data set according to the final index change rate interval, specifically including: determine the reserved index factors within the final index change rate interval and the corresponding prediction values of the reserved index factors according to the index change rates of the reserved index factors.

[0047] S4. Construct a training set according to the data set, train the prediction model, and obtain the trained prediction model; The training set includes: the NLR value, reserved index factors, corresponding prediction values of the reserved index factors, index change rates of the reserved index factors, and the response of the patient to glucocorticoid treatment for each patient after being screened by the first threshold and the second threshold; the response of the patient to glucocorticoid treatment is used as the prediction label.

[0048] As shown in Table 2 below, a comparison of the training results of the single NLR model and the multi-index model of the present invention is provided.

[0049] Table 2 Comparison of training results of single NLR model and multi-index model of the present invention

[0050] As can be seen from Table 2, the prediction ability of the multi-index model of the present invention on the training set and the validation set is significantly better than that of the single-index model (an increase of about 20%), indicating that integrating multi-dimensional indexes such as NLR, serum creatinine, uric acid, and glomerular filtration rate can capture the pathological state of patients more comprehensively, thereby improving the prediction accuracy.

[0051] The improvement in sensitivity (from 65% to 82%) means that more patients who are actually effective in treatment can be correctly identified, avoiding treatment delays; the improvement in specificity (from 70% to 88%) can reduce the wrong recommendation for ineffective patients, reduce the risk of side effects of glucocorticoids, and optimize the treatment plan.

[0052] The multi-index model proposed by the present invention is significantly superior to the traditional single-index method in terms of the accuracy, stability, and clinical practicability of predicting the response of IgA nephropathy to glucocorticoid treatment, verifying the technical advantages of integrating multi-dimensional biomarkers.

[0053] S5. Obtain the index factors of the patient in real time, input them into the trained prediction model, and output the response of the patient to glucocorticoid treatment.

[0054] As Figure 3 shown, this embodiment also provides a prediction system for the treatment effect of IgA nephropathy based on NLR, including the following modules: Index factor collection module: used to collect the index factors of the patient; Preprocessing module: connected to the index factor collection module, used to preprocess the index factors; Training set acquisition module: connected to the preprocessing module, used to obtain the training set from the preprocessed index factors; Prediction model: connected to the training set acquisition module, used to construct a training set according to the data set, train the prediction model, and obtain the trained prediction model; Prediction module: connected to the prediction model, used to obtain the index factors of the patient in real time, input them into the trained prediction model, and output the response of the patient to glucocorticoid treatment.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the therapeutic effect of IgA nephropathy based on NLR, characterized in that: include: S1. Collect the patient's indicator factors; The index factors include: NLR value, serum creatinine level, uric acid level and glomerular filtration rate; S2. Preprocessing the index factors; S3, obtaining a training set from the preprocessed indicator factors; S31. Analyze each pretreatment indicator factor by ROC curve to obtain the area under the operating characteristic curve of each pretreatment indicator factor; determine the predictive value of each pretreatment indicator factor according to the area under the operating characteristic curve of each pretreatment indicator factor; use the predictive value of NLR value as the benchmark value; use the predictive value of serum creatinine level, the predictive value of uric acid level and the predictive value of glomerular filtration rate as the screening value; S32, analyzing the pre-treated NLR value according to the ROC curve to obtain the optimal NLR cutoff value; S33, constructing a first threshold value according to the benchmark value, the NLR optimal cutoff value, and the screening value; S34. Filter the reserved indicator factor according to the first threshold; S35, calculating the index change rate of the predicted value of the reserved index factor and the benchmark value; calculating the standard deviation and mean according to the index change rate, and obtaining the index change rate interval according to the standard deviation and mean; S36, constructing a second threshold value according to the predicted value, the benchmark value, and the optimal cutoff value of NLR of the reserved indicator factor; S37, constructing a final indicator change rate interval according to the second threshold and the indicator change rate interval, and screening out a final data set according to the final indicator change rate interval; S4. Construct a training set according to the data set, train the prediction model, and obtain a trained prediction model; S5. Obtain the patient's indicator factors in real time, input them into the trained prediction model, and output the patient's response to glucocorticoid treatment.

2. The method for predicting the therapeutic effect of IgA nephropathy based on NLR according to claim 1, characterized in that: The calculation formula of the first threshold is: ； Where L1 represents the first threshold, Y NLR represents the predictive value of NLR value, that is, the benchmark value, M represents the optimal cutoff value of NLR, and Y si represents the predicted value of the ith screening value, and n represents the maximum number of screening values.

3. The method for predicting the therapeutic effect of IgA nephropathy based on NLR according to claim 1, characterized in that: The reserved index factor is filtered according to the first threshold, specifically: each filtering value is compared with the first threshold respectively, if the current filtering value is less than or equal to the first threshold, the index factor corresponding to the current filtering value is defined as the reserved index factor; if the current filtering value is greater than the first threshold, the current filtering value and the index factor corresponding to the current filtering value are deleted.

4. The method for predicting the therapeutic effect of IgA nephropathy based on NLR according to claim 1, characterized in that: The calculation formula of the second threshold is: ； Where L2 represents the second threshold, H j represents the predictive value of the jth reserved indicator factor, M represents the optimal cutoff value of NLR, and Y NLR represents the predictive value of the NLR value, and p represents the maximum number of reserved indicator factors.

5. The method for predicting the therapeutic effect of IgA nephropathy based on NLR according to claim 1, characterized in that: The final data set is screened out according to the final indicator change rate interval, specifically including: determining the reserved indicator factors in the final indicator change rate interval and the prediction values ​​corresponding to the reserved indicator factors according to the indicator change rates of the reserved indicator factors.

6. The method for predicting the therapeutic effect of IgA nephropathy based on NLR according to claim 5, characterized in that: Construct the final change rate interval of the indicator [Q 上 , Q 下 ], specifically: Q 上 =C 上 ×L2; Q 下 =C 下 ×L2; In the formula, Q 上 Represents the upper limit of the final change rate range of the indicator, Q 下 Represents the lower limit of the final change rate range of the indicator, C 上 Represents the upper limit of the index change rate interval, C 下 represents the lower limit of the index change rate interval, and L2 represents the second threshold.

7. The method for predicting the therapeutic effect of IgA nephropathy based on NLR according to claim 1, characterized in that: The training set includes: each patient's NLR value, reserved indicator factor, the prediction value corresponding to the reserved indicator factor, the indicator change rate of the reserved indicator factor, and the patient's response to glucocorticoid treatment; the patient's response to glucocorticoid treatment is used as a prediction label.

8. The method for predicting the therapeutic effect of IgA nephropathy based on NLR according to claim 1, characterized in that: The NLR value is the ratio of neutrophils to lymphocytes.

Images