阅读说明：本技术 一种新型的用于急性前葡萄膜炎个体化复发风险预测的模型及其应用 (Novel model for predicting individual recurrence risk of acute anterior uveitis and application thereof ) 是由 周猛 张子城 贺赫 王毓琴 纪皙文 于 2021-07-02 设计创作，主要内容包括：本发明公开了一种新型的用于急性前葡萄膜炎个体化复发风险预测的模型及其应用,所述预测模型中包括如下危险因素：强直性脊柱炎、HLA-B27、单核细胞计数、高密度脂蛋白、低密度脂蛋白,所述预测模型可实现对急性前葡萄膜炎患者复发风险水平高低的有效判断,为急性前葡萄膜炎复发风险的评估和个体化治疗方案的选择提供了客观的支持,具有良好的应用前景。(The invention discloses a novel model for predicting the individualized recurrence risk of acute anterior uveitis and application thereof, wherein the prediction model comprises the following risk factors: the prediction model can realize effective judgment of the recurrence risk level of the acute anterior uveitis patients, provides objective support for the assessment of the recurrence risk of the acute anterior uveitis and the selection of an individual treatment scheme, and has good application prospect.)

1. A group of risk factors for recurrence of acute anterior uveitis, wherein the risk factors include ankylosing spondylitis, HLA-B27, monocyte count, high density lipoprotein, and low density lipoprotein.

2. A predictive model of risk of recurrence of acute anterior uveitis comprising the following risk factors: ankylosing spondylitis, HLA-B27, monocyte count, high density lipoprotein, low density lipoprotein.

3. The predictive model of claim 2, wherein the predictive model uses the regression equation to calculate a relapse risk score of 5RF-panel_score：

5RF-panel_scoreAnkylosing spondylitis 0.09230+ HLA-B27 + 0.19863+ monocyte count (-0.59456) + high density lipoprotein 0.36348+ low density lipoprotein (-0.12934) + 0.3287.

4. The predictive model of claim 3, wherein the values for ankylosing spondylitis, HLA-B27 in the regression equation are as follows:

(ii) a value of 1 for ankylosing spondylitis with no ankylosing spondylitis and a value of 0 for no ankylosing spondylitis;

HLA-B27 has a positive assignment of 1 and HLA-B27 has a negative assignment of 0.

5. The predictive model of claim 3, wherein the relapse risk score is 5RF-panel_scoreIncluding low and high risk of recurrence;

preferably, the relapse risk score is 5RF-panel_score>0.328, high risk of recurrence;

preferably, the relapse risk score is 5RF-panel_score<0.328, low risk of recurrence.

6. A method for assessing the risk of individualized relapse in acute anterior uveitis, comprising the steps of:

(1) obtaining clinical data of a subject, the clinical data including ankylosing spondylitis, HLA-B27, monocyte counts, high density lipoproteins, low density lipoproteins;

(2) inputting the clinical data of step (1) into the predictive model of any one of claims 2-5 to obtain a relapse risk score of 5RF-panel_score；

(3) A relapse risk score of 5RF-panel obtained according to step (2)_scoreAssessing the risk of relapse in the subject:

the relapse risk score of 5RF-panel_score>0.328, high risk of recurrence;

the relapse risk score of 5RF-panel_score<0.328, low risk of recurrence.

7. A prediction device of risk of recurrence of acute anterior uveitis, the prediction device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor;

wherein the processor, when executing the computer program, runs the regression equation recited in claim 3.

8. A computer-readable storage medium, characterized in that the computer-readable storage medium comprises a stored computer program;

wherein the apparatus in which the computer-readable storage medium is located is controlled to perform the regression equation recited in claim 3 when the computer program is executed.

9. Use of the risk factor of claim 1 in the prediction of the risk of recurrence of acute anterior uveitis.

10. The use of any one of the following aspects, wherein said use comprises:

(1) use of the risk factor of claim 1 in a product for predicting the risk of recurrence of acute anterior uveitis;

(2) use of the risk factor of claim 1 in the construction of a model for predicting the risk of recurrence of acute anterior uveitis;

preferably, the predictive model is the predictive model of any one of claims 2-5;

(3) use of the prediction model of any one of claims 2 to 5 in the construction of a system for predicting the risk of relapse of acute anterior uveitis.

Technical Field

The invention belongs to the field of biomedicine, relates to a model for predicting disease recurrence risk, and particularly relates to a novel model for predicting acute anterior uveitis individualized recurrence risk and application thereof.

Background

Uveitis (Uveitis), also called Uveitis, is a kind of intraocular inflammatory disease mainly involving the eye pigment membrane such as iris, ciliary body, choroid, etc., and also is a kind of intractable eye diseases with multiple gene involvement, multiple etiology, complex pathogenesis and various clinical manifestations, wherein Uveitis can accumulate peripheral tissues to cause retinitis, optic disc inflammation, vitritis, etc., and can seriously impair vision and even be blinded (Thorne JE, Suhler E, Sup M, et al. Presence of nociception in the unknown stages: a clinical-based analysis [ J ]. JAMA hthalmol.2016; 134: 1237-1245.). Primary Uveitis is classified into anterior Uveitis (mainly cumulative iris, ciliary body), intermediate Uveitis (mainly cumulative peripheral choroid, retina), posterior Uveitis (mainly cumulative choroid, retina and even optic disc), and pan Uveitis based on the anatomical classification according to the classification criteria of the International Uveitis disease Study Group (IUSG) and the Uveitis naming standard working Group (SUN).

Acute Anterior Uveitis (AAU) accounts for about 80-85% of all Uveitis lesions (Gritz DC, Wong IG. Inc. and present of Uveitis in Northern California; the Northern California Epidemiology of Uveitis Study [ J ]. Ophthalmology.2004; 111: 491-. Currently, the etiology of AAU is unknown, and may be related to various factors such as genes, race, region, environment, and lifestyle. The main risk factors include autoimmune diseases such as ankylosing spondylitis, Behcet's disease, juvenile idiopathic arthritis, rheumatoid arthritis, psoriasis, auto-inflammatory bowel diseases such as ulcerative colitis and Crohn's disease, infectious diseases such as herpes simplex virus or herpes zoster virus, Human Immunodeficiency Virus (HIV), cytomegalovirus, treponema pallidum, Brucella, tubercle bacillus, leprosy bacillus and the like.

Although the prognosis of AAU is mostly good after active treatment, the AAU is easy to relapse, the vision of AAU patients is seriously damaged by the relapse, but the relapse risks of different AAU patients are greatly different, some AAU patients are easy to relapse, and some AAU patients can not relapse for a long time, clinical AAU patients usually worry about whether the AAU relapses and causes serious psychological burden, but clinicians cannot accurately inform the patients whether the AAU relapses, and the current research is mostly concentrated in the research on the pathogenesis and clinical treatment scheme of the AAU. In some previous studies, indexes related to the risk of AAU recurrence were explored, and it was found that the risk of AAU recurrence may be related to age, sex, number of attacks, human leukocyte antigen B27(HLA-B27) and various bacterial antibodies in blood, but the research results are difficult to replicate and the sample size is small, so there is still a need for a method for simply, rapidly and effectively predicting the risk of AAU recurrence, which provides objective support for the assessment of the risk of AAU recurrence and the selection of individualized treatment scheme.

Disclosure of Invention

In view of the above, in order to fill the gap in the art regarding a method for easily, quickly and efficiently determining the risk of relapse of AAU, the present invention aims to provide a novel model for individualized risk prediction of relapse of acute anterior uveitis and applications thereof.

The above object of the present invention is achieved by the following technical solutions:

in a first aspect of the invention, a panel of risk factors for recurrence of acute anterior uveitis is provided.

Further, the risk factors include ankylosing spondylitis, HLA-B27, monocyte count, high density lipoprotein, and low density lipoprotein.

The Ankylosing spondylitis, AS well AS the Ankylosing spondylitis and the AS, in the invention refers to a disease with inflammation of sacroiliac joints and spinal attachment points AS main symptoms, multiple joint lesions exist, most of the disease firstly invades the sacroiliac joints and then ascends to the cervical vertebrae, a few patients are invaded by the cervical vertebrae or a plurality of spinal column segments at the same time and can invade the peripheral joints, the joints at early lesions have inflammatory pain, accompanied by muscle spasm around the joints and stiff feeling, the disease is obvious in the morning and can also be manifested AS night pain, the pain is relieved by moving or taking an analgesic, along with the development of the disease, the joint pain is relieved, the activity of each spinal column segment and the joints is limited and malformed, and the whole spinal column and lower limbs at late stage become stiff bows.

The term "HLA-B27" as used herein refers to Human leukocyte antigen (Human leukocyte antigen), which belongs to one of HLA-B loci. HLA antigens are expressed products of Major Histocompatibility Complex (MHC) of human, and are mainly responsible for mutual recognition between cells and induction of immune response in the immune system, and the function of regulating immune response is classified into three categories according to the structure and distribution of HLA antigens: the class I molecules are HLA-A, HLA-B, HLA-C series antigens and are widely distributed on the surfaces of nucleated cells of various tissues, including platelets and reticulocytes, and mature red blood cells generally do not contain HLA antigens; the II molecules are HLA-D/DR, HLA-DP and HLA-DQ series antigens, and are mainly expressed on B cells and antigen presenting cells, and the two antigens are related to transplantation, wherein the II antigens are more important; class III molecules are complement components.

The term "monocyte count" as used herein, in the same way as "Monocytes" and "MO", also known as differential leukocyte count, means that different types of leukocytes are counted separately and percentages are calculated, specifically, the absolute value of a monocyte is generally expressed as MO # or MONO # on an assay result sheet, and the normal range of the absolute value of a monocyte is (0.12-0.8). times.10⁹/L。

The "High-density lipoprotein", as well as "High-density lipoprotein" and "HDL", in the present invention is a complex lipoprotein composed of lipids and proteins and their carried regulatory factors, also called a1 lipoprotein, and is generally expressed as HDL-C on the test result list, and the normal range of HDL is 0.7-2.0 mmol/L.

The "Low-density lipoprotein", as well as "Low-density lipoprotein" and "LDL", in the present invention is a lipoprotein particle that carries cholesterol into cells of peripheral tissues, and can be oxidized into oxidized Low-density lipoprotein, generally expressed as LDL-C on the test result, and the normal range of LDL is 0-3.4 mmol/L.

In a second aspect of the invention, a predictive model of the risk of recurrence of acute anterior uveitis is provided.

Further, the predictive model includes the following risk factors: ankylosing spondylitis, HLA-B27, monocyte count, high density lipoprotein, low density lipoprotein.

Further, the predictive model calculates a relapse risk score of 5RF-panel using the following regression equation_score：

5RF-panel_scoreAnkylosing spondylitis 0.09230+ HLA-B27 + 0.19863+ monocyte count (-0.59456) + high density lipoprotein 0.36348+ low density lipoprotein (-0.12934) + 0.3287.

Further, the assignment of ankylosing spondylitis, HLA-B27 in the regression equation is as follows:

(ii) a value of 1 for ankylosing spondylitis with no ankylosing spondylitis and a value of 0 for no ankylosing spondylitis;

HLA-B27 has a positive assignment of 1 and HLA-B27 has a negative assignment of 0.

Further, the monocyte count, the high density lipoprotein and the low density lipoprotein in the regression equation are specific numerical values of the monocyte count, the high density lipoprotein and the low density lipoprotein respectively.

Further, the relapse risk score of 5RF-panel_scoreIncluding low and high risk of recurrence;

preferably, the relapse risk score is 5RF-panel_score>0.328, high risk of recurrence;

preferably, the relapse risk score is 5RF-panel_score<0.328, low risk of recurrence.

A third aspect of the invention provides a method for assessing the risk of individualized relapse in acute anterior uveitis.

Further, the method comprises the steps of:

(1) obtaining clinical data of a subject, the clinical data including ankylosing spondylitis, HLA-B27, monocyte counts, high density lipoproteins, low density lipoproteins;

(2) inputting the clinical data of step (1) into the predictive model of the second aspect of the invention to obtain a relapse risk score of 5RF-panel_score；

(3) A relapse risk score of 5RF-panel obtained according to step (2)_scoreAssessing the risk of relapse in the subject:

the relapse risk score of 5RF-panel_score>0.328, high risk of recurrence;

the relapse risk score of 5RF-panel_score<0.328, low risk of recurrence.

A fourth aspect of the invention provides a device for predicting the risk of recurrence of acute anterior uveitis.

Further, the prediction apparatus comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor;

wherein the processor, when executing the computer program, runs the regression equation as described in the second aspect of the invention.

A fifth aspect of the invention provides a computer-readable storage medium.

Further, the computer readable storage medium includes a stored computer program;

wherein the apparatus in which the computer-readable storage medium is located is controlled to execute the regression equation in the second aspect of the present invention when the computer program is executed.

A sixth aspect of the invention provides the use of a risk factor according to the first aspect of the invention in the prediction of the risk of recurrence of acute anterior uveitis.

A seventh aspect of the invention provides the use of any one of the following aspects:

(1) the use of a risk factor according to the first aspect of the invention in a product for predicting the risk of recurrence of acute anterior uveitis;

(2) the application of the risk factors in the first aspect of the invention in the construction of a model for predicting the risk of recurrence of acute anterior uveitis;

preferably, the predictive model is the predictive model of the second aspect of the invention;

(3) the application of the prediction model in the first aspect of the invention in the construction of a prediction system for the risk of recurrence of acute anterior uveitis.

The invention also provides a method for constructing the prediction model of the second aspect of the invention.

Further, the method comprises the steps of:

(1) obtaining clinical data of a sample of a subject;

(2) identifying risk factors for recurrence of acute anterior uveitis according to the clinical data in step (1);

(3) and (3) carrying out model training by using a mechanical algorithm and constructing by using a Logistic regression method to obtain the prediction model of the second aspect of the invention.

Further, the machine learning algorithm in the step (3) includes at least one of a support vector machine algorithm, a naive bayes algorithm, a decision tree algorithm, a random forest algorithm, a neural network algorithm and a regression algorithm.

Further, when the significance difference analysis is carried out, the significance analysis is carried out by adopting significance test, and in a specific embodiment, clinical indexes with data in different distribution modes are tested by adopting different test methods: if the clinical index data is normally distributed (such as MO, HDL and LDL), Student's t test (Student's t-test) is adopted; if the clinical index data is not normally distributed (such as age), using Wilcoxon signed-rank test (Wilcoxon signed-rank test); if the clinical index is classified data (such as ankylosing spondylitis and HLA-B27), adopting chi-square test; the specific process of the inspection is a conventional technical means.

Further, the method further comprises the step of performing performance evaluation on the constructed prediction model, wherein the specific performance evaluation mode is as follows: and drawing an ROC curve on the basis of the training set and the test set based on the obtained prediction model of the acute anterior uveitis recurrence risk, and determining that the performance of the model meets the requirements if the values of specificity TPR, sensitivity TNR, accuracy ACC and AUC in the ROC curve are respectively greater than respective preset values. Wherein the preset value is set according to conventional experience in the art and is not particularly limited herein. For example, in one specific embodiment, a model is deemed satisfactory where the AUC value is greater than 0.7.

The invention has the advantages and beneficial effects that:

compared with the prior art, the invention provides a novel model for predicting the individual recurrence risk of the acute anterior uveitis, the model can effectively judge the recurrence risk level of the acute anterior uveitis patients, high-risk groups and low-risk groups can be clearly distinguished according to the recurrence risk score obtained by calculation of the model, the judgment result can guide clinicians to adjust intervention measures and treatment schemes adopted by the acute anterior uveitis patients, the clinician and the patients can quickly master the recurrence risk early warning information, and objective support is provided for evaluation of the recurrence risk of the acute anterior uveitis and selection of individual treatment schemes.

Drawings

Embodiments of the invention are described in detail below with reference to the attached drawing figures, wherein:

FIG. 1 is a heat map of the different distribution of clinical factors in AAU data for 233 AAU patients;

FIG. 2 is a forest plot of HR and 95% confidence intervals for each clinical feature calculated by the one-factor Cox model;

FIG. 3 is a diagram of the construction results of the 5RF-panel model in the training queue, wherein, A is a diagram: 5RF-panel_scoreAnd a heat map visualizing clinical factors, panel B: 5RF-panel_scoreA graph of the results of the correlation with three clinical factors (MO, HDL and LDL);

FIG. 4 is a diagram of the construction results of the 5RF-panel model in the training queue, wherein, A is a diagram: 5RF-panel model scoreGraph B: using 5RF-panel_scoreA result graph of Kaplan-Meier survival analysis without recurrence time is carried out on the high-risk group and the low-risk group in the training queue;

FIG. 5 is a diagram of the construction result of the 5RF-panel model in the training queue, wherein, A is a diagram: 5RF-panel difference box plot between high risk group and low risk group in training cohort, B plot: training 5RF-panel difference histograms between high risk groups and low risk groups in the cohort;

FIG. 6 is a graph of the results of 5RF-panel model validation in the test queue, wherein Panel A: 5RF-panel_scoreAnd a heat map visualizing clinical factors, panel B: 5RF-panel_scoreA graph of the results of the correlation with three clinical factors (MO, HDL and LDL);

FIG. 7 is a graph of the results of 5RF-panel model validation in the test queue, wherein Panel A: ROC curve results plot for 5RF-panel model score, panel B: using 5RF-panel_scoreA result graph of Kaplan-Meier survival analysis without recurrence time is carried out on the high risk group and the low risk group in the test queue;

FIG. 8 is a graph of the results of 5RF-panel model validation in the test queue, wherein Panel A: 5RF-panel differential box plot between high risk group and low risk group in test cohort, Panel B: 5RF-panel difference histograms between high risk groups and low risk groups in the test cohort;

FIG. 9 is a graph comparing the results of the 5RF-panel model with conventional clinical features in predicting the risk of AAU recurrence, wherein, A is a graph: ROC curve results plot for 5RF-panel model and traditional clinical features, panel B: comparing the 5RF-panel model to the conventional clinical profile for recurrence-free time-to-live difference plots;

FIG. 10 is a graph comparing the results of the 5RF-panel model with conventional clinical features in predicting the risk of AAU recurrence, wherein, A is a graph: graph of Kaplan-Meier survival analysis results without time to relapse scored for the 5RF-panel model of the ankylosing spondylitis (Yes, No) subtype, panel B: differential box plot of 5RF-panel model score for the medium and low risk groups of ankylosing spondylitis (Yes, No) subtype;

FIG. 11 is a graph comparing the results of the 5RF-panel model with conventional clinical features in predicting the risk of AAU recurrence, wherein, A is a graph: graph of the results of Kaplan-Meier survival analysis without time to relapse on the 5RF-panel model score for HLA-B27(Yes, No) subtype, panel B: differential box plot of 5RF-panel model scores for the high risk group and the low risk group in HLA-B27(Yes, No) subtype;

FIG. 12 is a graph showing the results of ROC curves of a five risk factor (5RF-panel) model and another relapse risk assessment model established in the present invention, wherein A is a graph: the five risk factors (5RF-panel) model established by the invention, B picture: another model for risk assessment of relapse (ankylosing spondylitis, HLA-B27, HDL, LDL, TG).

Detailed Description

The present invention is further illustrated below with reference to specific examples, which are intended to be illustrative only and are not to be construed as limiting the invention. As will be understood by those of ordinary skill in the art: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents. The following examples are examples of experimental methods not indicating specific conditions, and the detection is usually carried out according to conventional conditions or according to the conditions recommended by the manufacturers.

Example 1 screening of risk factors associated with acute anterior uveitis

1. Research population

233 cases of patients diagnosed as AAU according to the AAU standard of the international uveitis research group (IUSG) during the period from 1 month of 2015 to 2 months of 2020 by the affiliated eye vision hospital of the western medical university were retrospectively collected, clinical information of the patients was collected, and the age, sex, biochemical indicators, 1 st-year recurrence rate, and 2 nd-year recurrence rate of the patients were subjected to descriptive analysis, and the last follow-up time was 2021 month of 2021 year;

inclusion criteria were: (1) inclusion in active stage patients with significant ocular inflammatory manifestations; (2) patients with acute anterior uveitis with a course of less than three months;

exclusion criteria: (1) patients with intermediate and/or posterior uveitis and/or other ocular inflammatory diseases in addition to acute anterior uveitis; (2) patients with ankylosing spondylitis complicated with other systemic autoimmune diseases; (3) patients suffering from or undergoing treatment for a malignant tumor; (4) a pregnant or lactating patient; (5) patients with infectious diseases; (6) patients with secondary relapses were confirmed, but the specific time of secondary relapse was not clear during follow-up.

2. Data acquisition

During retrospective observation, patients were longitudinally followed for recurrence of acute anterior uveitis, whether the patients had a recurrence was determined by observing the recurrence of the inflammation at the scheduled out-patient visit or by assessing the acute symptoms provided by the patients to the clinician, by telephone follow-up of patients who had no recurrence record in the medical record system of the affiliated eye-care hospital of the university of Wenzhou medical, who had essentially no recurrence, and by recording the time of recurrence of patients who confirmed recurrence at other hospitals or clinics, and if not, rejecting the patient's records, assessing the potential risk factors of recurrence of acute anterior uveitis, including the age, gender, the level of HLA-B27, whether they had laboratory indices such as ankylosing spondylitis, and all specialist physicians assessing the ocular inflammation of patients have completed their physician' practice.

3. Screening for clinical risk factors associated with acute anterior uveitis

To determine the clinical risk factors associated with the recurrence of acute anterior uveitis, this example compared the clinical characteristics of a total of 233 patients in the relapsing and non-relapsing groups, determined the clinical characteristics with significant differences (P <0.05) using Wilcoxon rank-sum test and Student's t test, and then further evaluated the clinical risk factors associated with the recurrence of acute anterior uveitis (P <0.05) from the different candidate clinical characteristics obtained using the one-way Cox proportional risk regression model.

4. Results of the experiment

(1) Baseline characteristics of the study population

Clinical characteristics and laboratory indices of 233 AAU patients were summarized in terms of number of relapses (see table 1 and figure 1). The results showed that of 233 AAU patients, 73 (31.33%) had relapsed during the follow-up. All AThe median age (+ IQR) of AU patients is 39.70(32.12-49.16) years. 150 of all AAU patients (64.38%) were males, with the results meeting the age and gender characteristics of AAU; a total of 94 (40.34%) AAU patients had AS, with 55 (34.38%) of the AAU patient group that had relapsed once being accompanied by AS and 39 (53.42%) of the AAU patient group that had relapsed two or more times being accompanied by AS (P ═ 0.006); 129 of all patients (70.88%) were positive for HLA-B27, 80 of the group of patients who relapsed once (65.57%) were positive for HLA-B27, and 49 of the group of AAU patients who relapsed twice or more (81.67%) were positive for HLA-B27 (P ═ 0.025); the above results indicate that patients with concomitant AS are more likely to relapse than patients positive for HLA-B27; monocyte count MO (+ -SD) (10)⁹/L) MO was 0.53 ± 0.19 in the AAU patient group with one relapse and 0.45 ± 0.20 in the AAU patient group with two or more relapses (P ═ 0.007); the median of triglycerides TG (+ IQR) (mmol/L) was 1.47(0.94-2.1) in the AAU patient group with one recurrence, and the median of TGs was 1.01(0.74-1.53) (P ═ 0.023) in the AAU patient group with two or more relapses; the mean of high density lipoprotein HDL (+ -SD) (mmol/L) in the AAU patient group with one recurrence was 1.29 ± 0.28, and the mean of HDL in the AAU patient group with two or more recurrences was 1.47 ± 0.34(P ═ 0.004); the mean of low density lipoprotein LDL (+ -SD) (mmol/L) in the AAU patient group that relapsed once was 3.02 ± 0.80, and the mean of LDL in the AAU patient group that relapsed twice or more was 2.63 ± 0.77(P ═ 0.004).

(2) Screening for clinical risk factors associated with acute anterior uveitis

Results of a one-way Cox proportional risk regression model on 38 clinical signatures to assess their association with patient relapse-free survival are shown in fig. 2, showing that a set of clinical risk factors (HDL, MO, LDL, AS, HLA-B27) is significantly associated with patient relapse-free survival, with MO (HR 0.074, 95% CI 0.01-0.541, P0.01) and LDL (HR 0.583, 95% CI 0.36-0.945, P0.029) being associated with good survival of patients, and the other three clinical signatures HDL (HR 4.985, 95% CI 1.845-13.471, P0.002), AS (HR 1.04668, 95% CI 1.049-2.654, P0.03125) and HLA-B27(HR 1.3894, 95% CI 1. 3.747) being associated with patient relapse-free survival.

Table 1 baseline characteristics of AAU patients enrolled in the study

Wherein, NEU: (ii) neutrophil count; LYM: (ii) lymphocyte count; NLR: the ratio of neutrophils to lymphocytes; MO: counting the mononuclear cells; EO: eosinophil count; BASO: basophilic granulocyte count; CRP: c-reactive protein; SAA: serum amyloid a; AST: aspartate aminotransferase; ALT: alanine aminotransferase; AST/ALT: aspartate aminotransferase/alanine aminotransferase; GGT: glutamyl transpeptidase; ALP: alkaline phosphatase; TP: total protein; ALB: albumin; GLB: a globulin protein; A/G: albumin/globulin; TBIL: total bilirubin; DBIL: direct bilirubin; IBIL: indirect bilirubin; GLU: glucose; UA: uric acid; crea: creatinine; TG: a triglyceride; TCH (traffic channel): total cholesterol; HDL: high density lipoprotein; LDL: low density lipoprotein; RF: rheumatoid factor; VD (25-OH): 25-hydroxy vitamin D; ESR: erythrocyte sedimentation rate.

Example 2 establishment and validation of an individualized relapse risk assessment model for acute anterior uveitis

1. Establishment of five-risk factor (5RF-panel) model

The clinical characteristic raw data of samples without the indices described in the 5RF-panel model (AS, HLA-B27, MO, HDL and LDL) were excluded, and a total of 111 AAU patients were kept for further analysis, 111 patients were randomly and equally assigned to the training cohort (n 56) and the test cohort (n 55) of the relapsing and non-relapsing groups, and the 5RF-panel model was constructed using the logistic regression analysis method of the five clinical risk factor groups associated with relapse in the training cohort, AS follows:

5RF-panel_scoreankylosing spondylitis 0.09230+ HLA-B27 + 0.19863+ monocyte count (-0.59456) + high density lipoprotein 0.36348+ low density lipoprotein (-0.12934) +0.3287

Among them, 5RF-panel_scoreRepresenting a recurrence risk score, by a multiplier, and the risk factors in the regression equation are assigned the following values: (ii) a value of 1 for ankylosing spondylitis with no ankylosing spondylitis and a value of 0 for no ankylosing spondylitis; a positive assignment of 1 to HLA-B27 and a negative assignment of 0 to HLA-B27; monocyte count, high density lipoprotein, low density lipoprotein are the corresponding specific values.

In the training cohort, the relapse risk scores of the points with 0% false positive rate and 100% true positive rate were taken as a cutoff value (Cut-off) and applied directly to the training cohort, the test cohort and the overall cohort (111 patients) to classify the patients into low-risk and high-risk groups.

2. Prediction of relapse risk and assessment of performance using five risk factor (5RF-panel) model

To determine the effectiveness of the 5RF-panel model in clinically predicting recurrence of acute anterior uveitis, this example used the 5RF-panel model to score each AAU patient in the training cohort to obtain a recurrence risk score of 5RF-panel_score。

3. Verifying the reproducibility and durability of the five risk factor (5RF-panel) model

To determine the reproducibility and durability of the 5RF-panel model, this example used the 5RF-panel model to predict the risk of relapse of AAU patients in the test cohort, and first scored each AAU patient in the test cohort using the 5RF-panel model, and then analyzed its diagnostic efficacy (ROC curve) to further validate the predictive power of the 5RF-panel model.

4. Statistical analysis

Data are expressed as mean ± Standard Deviation (SD) in normal distribution, median and quartile Intervals (IQR) in abnormal distribution, and percentage in classification, differences between baseline characteristics in the two groups are comparatively analyzed by Wilcoxon signed rank test (abnormal distribution variable), Student's t test (normal distribution variable), and chi-square test (classification or binary variable), all baseline characteristics are stratified by recurrence number, potential risk factors associated with AAU recurrence are evaluated using a one-way Cox proportional risk regression model to obtain risk ratio (HR) and 95% Confidence Interval (CI), a risk factor group for individualized recurrence risk stratification is established using a logistic regression analysis model with R-package "stats", recurrence-free survival is evaluated using Kaplan-Meier method, survival distributions between the groups are compared using Log-rank test, the predictive performance of recurrence for each variable and the established risk factor group was determined using the area under the receiver operating characteristic curve (AUROC) analysis and compared using the Delong test, with P <0.05 being statistically significant for the differences, and all statistical analyses were performed using R language software (v4.0.3) and Bioconductor software (v 3.13).

5. Results of the experiment

The results show that 5RF-panel_scoreAnd MO (Pearson r ═ 0.47, P)<0.001) and LDL (Pearson r ═ 0.59, p<0.001) is significantly negatively correlated with HDL (Pearson r ═ 0.55, P<0.001) 5RF-panel between relapsing and non-relapsing groups with significant positive correlation_scoreSignificant differences in distribution (Wilcoxon test, P)<0.001), non-relapsing group 5RF-panel_scoreThe median of (a) was significantly lower than the relapsing group (0.234vs 0.421) (see fig. 3A and B);

the results of the ROC curve show that 5RF-panel_scoreBetter predictive power was provided in the training cohort with an AUC of 0.837 (95% CI ═ 0.728-0.947), which was divided into a high risk group (n ═ 19) and a low risk group (n ═ 37) according to the cut-off values defined in the method, the percentage of relapse free survival reaching 3 years being 89.2% in the low risk group and 39.0% in the high risk group (see fig. 4A); 55RF-panel using Kaplan-Meier (K-M) curve and Log-rank test_scoreThe results of the survival analysis are shown in FIG. 4B, and the relapse-free ratio of patients in the low risk group is significantly higher than that in the high risk group (log-rank test, P)<0.001, HR 45.874, 95% CI 5.232-402.2); the levels of the markers in the 5RF-panel model also differed significantly between the two risk groups (see fig. 5A), MO (Wilcoxon test, P ═ 0.043) and LDL (Wilcoxon test)P ═ 0.003) tends to be higher in the low risk group than in the high risk group, while HDL (Wilcoxon test, P ═ 0.002) is higher in the high risk group; the proportion of AS (chi-square test, P ═ 0.034) and HLA-B27 (chi-square test, P ═ 0.022) was significantly different in the two risk groups (see fig. 5B), and the above results indicate that 5RF-panel had a good effect AS a model for predicting the risk of AAU relapse;

FIG. 6A shows a 5RF-panel_scoreDistribution of classical clinical characteristic index and 5RF-panel index in the test cohort, 5RF-panel with MO (Pearson r ═ 0.55, P<0.001)、LDL(Pearson r＝-0.43，P<0.001) and HDL (Pearson r ═ 0.65, P<0.001), wherein 5RF-panel is in significant negative correlation with MO and LDL and in significant positive correlation with HDL; 5RF-panel of relapsing group_scoreIs significantly higher than that of the non-recurrent group (median: 0.535vs 0.316, Wilcoxon test, P)<0.001) (see fig. 6B); the results in FIG. 7A show 5RF-panel_scoreAn AUC value of 0.725 (95% CI ═ 0.561-0.889); according to 5RF-panel_scoreCan be divided into a low risk group (n ═ 24) and a high risk group (n ═ 31), the time difference between the two groups without recurrence is significant (Log-rank test, P ═ 0.024, HR ═ 51.982, 95% CI ═ 4.438-608.9) (see fig. 7B), the percentage of recurrence-free survival up to 3 years in the low risk group is 85.1%, and the percentage of recurrence-free survival in the high risk group is 55.7%; MO (Wilcoxon test, P)<0.001) and LDL (Wilcoxon test, P)<0.001) has a lower index in the high risk group, HDL (Wilcoxon test, P)<0.001) had a lower index in the low risk group and only the proportion of HLA-B27 was significantly different in the low risk group and the high risk group (chi-square test, P ═ 0.003) (see fig. 8A and B).

Comparative example 1 the personalized risk of relapse assessment model for acute anterior uveitis established in accordance with the present invention compares the performance with HLA-B27 and ankylosing spondylitis

1. Experimental methods

In order to compare the ability of the five risk factor (5RF-panel) model established by the invention to predict the recurrence risk of acute anterior uveitis with the ability of the clinical features (such as HLA-B27 and ankylosing spondylitis) currently used to predict the recurrence risk of acute anterior uveitis, the ROC curve analysis was performed on the whole cohort (111 patients) by using the model and the clinical features.

2. Results of the experiment

The results show that 5RF-panel_scoreThe AUC of (a) is 0.766 (95% CI ═ 0.668-0.863), significantly higher than HLA-B27(AUC ═ 0.596, 95% CI ═ 0.51-0.681, DeLong test: P ═ 0.003) and ankylosing spondylitis (AUC ═ 0.581, 95% CI ═ 0.476-0.687, DeLong test: P ═ 0.003) (see fig. 9A);

the results of the K-M curves show that the prognosis for patients in the low risk group is best compared to that predicted for HLA-B27 and ankylosing spondylitis, and that the recurrence-free survival rates (Log-rank test, P <0.001) for the low risk group and the high risk group differ most from those of the ankylosing spondylitis subgroup (Log-rank test, P ═ 0.2) and HLA-B27 subgroup (Log-rank test, P ═ 0.07) (see fig. 9B);

for 5RF-panel_scoreThe results of further correlation analysis with the classical clinical subgroup defined by HLA-B27 and ankylosing spondylitis showed that 5RF-panel_scorePatients in the ankylosing spondylitis (Yes) group (Log-rank test, P0.002, HR 32.963, 95% CI 2.658-408.9) or in the ankylosing spondylitis (No) group (Log-rank test, P0.008, HR 72.089, 95% CI 6.801-746.1) can still be classified into high-risk and low-risk groups; the percentages of recurrence-free survival up to 3 years in the low-risk group and high-risk group in the ankylosing spondylitis (Yes) group were 94.4% and 49.7%, respectively, and the percentages of recurrence-free survival up to 3 years in the low-risk group and high-risk group in the ankylosing spondylitis (No) group were 83.1% and 39.8%, respectively, and 5RF-panel in the high-risk group and low-risk group in the ankylosing spondylitis (Yes) group (Wilcoxon test, P ═ 0.006) and ankylosing spondylitis (No) group (Wilcoxon test, P ═ 0.005) were used for the 5RF-panel in the high-risk group and low-risk group_score5RF-panel of distinctly differently distributed, high-risk groups_scoreThe median was significantly higher than the low risk group (ankylosing spondylitis (Yes) group: 0.506vs 0.316, ankylosing spondylitis (No) group: 0.385vs 0.24) (see fig. 10A and B);

5RF-panel_scorethe same ability to effectively discriminate was shown on the HLA-B27 subgroup (see FIGS. 11A and B), 5RF-panell_scorePatients in the HLA-B27(Yes) group (Log-rank test, P0.002, HR 29.825, 95% CI 4.104-216.8) and patients in the HLA-B27(No) group (Log-rank test, P0.007, HR 136.175, 95% CI 4.1-4523) can be classified into a high risk group and a low risk group, the percentage of relapse-free survival up to 3 years in the low risk group and the high risk group in the HLA-B27(Yes) group patients is 81.6% and 51.1%, respectively, and the percentage of relapse-free survival up to 3 years in the low risk group and the high risk group in the HLA-B27(No) group is 92.9% and 33.3%, respectively; meanwhile, in HLA-B27(Yes) group (Wilcoxon test, P ═ 0.022) and HLA-B27(No) group (Wilcoxon test, P ═ 0.004), 5RF-panel was used for high risk group and low risk group_scoreThere is a significant difference in the distribution of (c); high risk group of 5RF-panel_scoreThe median was significantly higher than the low risk group (HLA-B27(Yes) group: 0.425vs 0.316, HLA-B27(No) group: 0.492vs 0.134).

Comparative example 2 comparison of Performance of the Individual model for assessing the Risk of relapse of acute anterior uveitis established according to the invention with that of another model for assessing the Risk of relapse

1. Experimental methods

In this example, another recurrence risk assessment model was constructed, the 5 risk factors included in the risk assessment model were ankylosing spondylitis, HLA-B27, HDL, LDL, and TG, and the ability of the five risk factor (5RF-panel) model established in the present invention to predict the recurrence risk of acute anterior uveitis was compared with the ability of the above recurrence risk assessment model to predict the recurrence risk of acute anterior uveitis, and this example performed ROC curve analysis on all AAU patients using the above models.

2. Results of the experiment

The results show that the five risk factors (5RF-panel) model established by the invention has a significantly higher AUC value (AUC is 0.766, and 95% CI is 0.668-0.863) for predicting the recurrence risk of acute anterior uveitis than another recurrence risk assessment model (AUC is 0.698, and 95% CI is 0.531-0.865) (see FIGS. 12A and B), and the five risk factors (5RF-panel) model established by the invention has a better capability for predicting the recurrence risk of acute anterior uveitis.

The above description of the embodiments is only intended to illustrate the method of the invention and its core idea. It should be noted that, for those skilled in the art, without departing from the principle of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications will also fall into the protection scope of the claims of the present invention.