RU2523391C1 - Method for prediction of risk of developing coronary artery restenosis after stenting procedure in patients with ischemic heart disease - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: evaluation of additional risk factors - measuring the epicardial fatty tissue thickness, blood leptin, lipoprotein A and glucose is followed by mathematical processing of the numerical values of the risk factors to calculate a probability of restenosis by formula: P=[exp(-8.0248+0.0893·X1+0.00354·X2+0.1397·X3+1.1194·X4+0.3286·X5+0.0665·X6)]/[1+exp(-8.0248+0.0893·X1+0.00354·X2+0.1397·X3+1.1194·X4+0.3286·X5+0.0665·X6)], wherein P is the probability of restenosis, %, X1 is a patient's leptin, ng/ml; X2 is lipoprotein A, mg/l; X3 is the epicardial fatty tissue thickness, millimetres; X4 is high-density lipoprotein cholesterol, mmole/l; X5 is blood glucose, mmole/l; X6 is interleukin-6, pcg/ml; -8.0248 is the absolute term of the equation. The invention enables more accurate prediction of the risk of developing restenosis following coronary stenting in patients with ischemic heart disease.

EFFECT: effectiveness of the presented method and its applicability in practical healthcare.

1 tbl, 2 ex, 1 dwg

Description

The present invention relates to medicine, cardiology and can be used to diagnose the risk of complications, which include the development of restenosis of coronary arteries (repeated stenosis or narrowing) in the stenting zone in patients with coronary heart disease in order to conduct preoperative correction of risk factors, reduce the risk of complications and improved stenting results.

A known method for predicting the development of restenosis in the stent in men of working age with primary uncomplicated myocardial infarction, see RF patent No. 2410019, M. cl. A61B 5/02, publ. 01/27/2011, which includes determining the likelihood of developing restenosis in the stent according to a mathematical formula taking into account the localization coefficients of the stent: the anterior descending artery, the right coronary artery, the envelope of the artery, the diagonal branch, the coefficient of resumption of angina, re-hospitalization, the development of a typical anginal attack during VEM, ST segment depression is oblique or horizontal.

The disadvantages of this method include assessing the risk of developing restenosis indirectly after stenting and only in a limited circle of patients - men of working age, in addition, the signs used for prediction are subjective and different specialists can be evaluated differently.

There is a better method for assessing the risk of adverse cardiovascular events in patients with coronary heart disease who underwent endovascular myocardial revascularization within one year after the intervention, see RF patent for invention No. 2444297, M. cl. A61B 10/00, G01N 33/68 (prototype), characterized by assessing the level of anti-inflammatory cytokines IL-6 and FNO, low-density plasma lipoprotein cholesterol, triglycerides, high-density lipoprotein cholesterol, taking into account the fact of smoking and statin use. If there are more than four factors considered: the initial level of pro-inflammatory IL-6 cyclotins is more than 5 ng / ml and FNO is more than 50 ng / ml, low-density lipoprotein cholesterol is more than 2.5 mmol / l, high-density lipoprotein cholesterol is less than 1.0 mmol / l , triglycerides more than 1.7 mmol / l, smoking, statin rejection, patients are classified as high risk of adverse cardiovascular events.

The disadvantages of the prototype include not only the assessment of specifically restenosis, but also the evaluation of many other complications after stenting: including the occurrence of repeated heart attack, cerebrovascular accident, cardiac arrhythmias and others. Moreover, the assessment of risk factors included in the forecasting model in the prototype is carried out only in two gradations more or less than normal values.

The technical result of the invention is the creation of a more accurate and specifically directed method for predicting the risk of developing restenosis after coronary stenting in patients with coronary heart disease.

The technical result is achieved by combining well-known features common with the prototype, such as examining a patient, diagnosing risk factors with detecting deviations from their normal state, including the factor of the presence of high density lipoprotein cholesterol and interleukin-6 in the blood and new signs that the diagnosis is carried out by determining the values of additional risk factors - measuring the thickness of epicardial adipose tissue, levels of leptin, lipoprotein "a" and glucose in the blood, after which mathematical processing of the numerical values of risk factors with obtaining the probability of occurrence of restenosis according to the proposed prediction formula of the following form:

$$P=\frac{\exp \left(-\mathrm{8,0248}+\mathrm{0,0893}\cdot X\mathrm{one}+\mathrm{0,00354}\cdot \mathrm{<figure-callout\; id="2"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}2+0.1397\cdot \mathrm{<figure-callout\; id="3"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}3+\mathrm{1,1194}\cdot X\mathrm{four}+0.3286\cdot X56+0.0665\cdot X6\right)}{\mathrm{one}+\exp \left(-\mathrm{8,0248}+\mathrm{0,0893}\cdot X\mathrm{one}+\mathrm{0,00354}\cdot \mathrm{<figure-callout\; id="2"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}2+0.1397\cdot \mathrm{<figure-callout\; id="3"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}3+\mathrm{1,1194}\cdot X\mathrm{four}+0.3286\cdot X5+0.0665\cdot X6\right)}$$

where P is the probability of occurrence of restenosis in%, X1 is the patient's leptin value in ng / ml: X2 is the lipoprotein “a” value in mg / l; X3 is the thickness of the epicardial adipose tissue in millimeters; X4 - the value of high density lipoprotein cholesterol in mmol / l; X5 - blood glucose value in mmol / l; X6 is the value of interleukin-6 in PCG / ml, -8.0248 is the free term of the equation.

The novelty of the proposed method is the diagnosis by determining the values of additional risk factors - measuring the thickness of the epicardial adipose tissue, the levels of leptin, lipoprotein "a" and glucose in the blood, and then carry out mathematical processing of the numerical values of the risk factors to obtain the likelihood of restenosis using the proposed prediction formula for the following type:

$$P=\frac{\exp \left(-\mathrm{8,0248}+\mathrm{0,0893}\cdot X\mathrm{one}+\mathrm{0,00354}\cdot \mathrm{<figure-callout\; id="2"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}2+0.1397\cdot \mathrm{<figure-callout\; id="3"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}3+\mathrm{1,1194}\cdot X\mathrm{four}+0.3286\cdot X56+0.0665\cdot X6\right)}{\mathrm{one}+\exp \left(-\mathrm{8,0248}+\mathrm{0,0893}\cdot X\mathrm{one}+\mathrm{0,00354}\cdot \mathrm{<figure-callout\; id="2"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}2+0.1397\cdot \mathrm{<figure-callout\; id="3"\; label="X"\; filenames="00000004.png"\; state="\{\{state\}\}">X</figure-callout>}3+\mathrm{1,1194}\cdot X\mathrm{four}+0.3286\cdot X5+0.0665\cdot X6\right)}$$

where P is the probability of occurrence of restenosis in%, X1 is the patient's leptin value in ng / ml; X2 - the value of lipoprotein "a" in mg / l; X3 is the thickness of the epicardial adipose tissue in millimeters; X4 - the value of high density lipoprotein cholesterol in mmol / l; X5 - blood glucose value in mmol / l; X6 is the value of interleukin-6 in PCG / ml, -8.0248 is a free member of the equation.

Thus, the diagnosis with the assessment of additional risk factors by determining the thickness of the epicardial adipose tissue, the values of leptin, lipoprotein “a” and glucose in the blood in combination with the known risk factors - the presence of high density lipoprotein cholesterol and interleukin-6 in the blood - allows a high degree of accuracy and a wide range of values to determine the degree of risk of developing restenosis after coronary stenting.

The calculation of the likelihood of restenosis using the mathematical dependence developed by the authors allows predicting coronary artery restenosis in patients with different values of risk factors and any degree of deviation of these risk factors from their normal state.

According to the patent information research, no combination of known features with new features was found in the sources of patent and scientific and technical information, and therefore, these features can be attributed to those having novelty. The specified combination does not follow explicitly from the existing level of technology and technology and has an inventive step. Embodiments of the proposed method, indicated in the description, indicate its industrial applicability.

Figure 1 presents a graph of diagnostic efficiency, showing the sensitivity and specificity of the proposed forecasting method.

Curve 1 ROC curve - (Receiver Operator Characteristic) in the vertical and horizontal coordinates in percent reflects the data obtained using the statistical program when the values of the studied risk factors are entered into the computer. Area 3 between curve 1 and conditional line 2 reflects the quality of the model used to predict restenosis. The larger the area 3, the higher the quality of the proposed model. In the proposed model, the area 3 obtained above conditional line 2 reaches 0.8150 from the entire area located above conditional line 2 and reflects the high degree of significance of the proposed mathematical forecasting model.

The proposed method is as follows.

When examining a patient using known methods, the specific values of six risk factors for the development of restenosis after coronary stenting are determined. Using ultrasound equipment, the thickness of epicardial adipose tissue is determined, then, according to generally known methods, the patient’s leptin value in ng / ml is determined on standard equipment in a medical laboratory; the value of lipoprotein "a" in mg / l; the value of high density lipoprotein cholesterol in mmol / l; blood glucose in mmol / l; and the value of interleukin-6 in PCG / ml.

The proposed forecasting method was obtained by analyzing the own data of 186 patients who were examined in the Altai Regional Cardiology Dispensary. Before stenting, all patients underwent an examination, including the diagnosis of 30 known cardiovascular risk factors, including the diagnosis of risk factors included in the mathematically expressed dependence (thickness of epicardial adipose tissue, leptin, lipoprotein “a”, blood glucose, high density lipoprotein cholesterol , interleukin-6.) The data obtained were subjected to statistical processing by the method of logistic regression, as a result of which the authors equation was obtained, which included the 6 most significant facts risk factors with a true prediction percentage of 82.5%.

Table 1 presents a step-by-step procedure for including risk factors in a mathematical model. The relative contribution of individual risk factors is expressed by the value of the Wald Chi-Square statistic, and table 1 also shows the regression coefficients for each risk factor included in the Estimate mathematical model. As a criterion of agreement between the actual distribution of observations by individual gradations and the forecast based on the logistic regression equation, the percent of correct prediction (Concordant) and the D-Zomer coupling coefficient (Somers'D) were used. When evaluating the regression equation, we used the method of stepwise inclusion of risk factors, which ranks the attribute in accordance with their contribution to the model. The significance level for including predictors in the regression equation was set so that the achieved significance level of the Chi-Square criterion (Pr> Chi-Square) for each risk factor at the end of the step-by-step procedure did not exceed 10%. A general assessment of model agreement and real data was evaluated using the Hosmer and Lemeshov Goodness-of-Fit Test.

Step order of inclusion of the selected predictors (risk factors) (Variable) with the percentage of correct predictions at each step is shown in Table 1. The percentage of correct predictions for the total set of predictors predicting the development of restenosis was 82.5%.

Table 1 Results of a step-by-step procedure of logistic regression Step Variable Chi square Estimate Concordant Pr> Chi-Square one Leptin 15,327 0,0893 66.2% <0.0001 2 Lipoprotein a 25,815 0,00354 76.3% <0.0001 3 Epicardial adipose tissue thickness 27,789 0.1397 78.1% <0.0001 four Blood sugar 30,202 0.3286 79.3% <0.0001 5 Interleukin-6 31,451 0.0665 80.6% <0.0001 6 HDL cholesterol 31,813 1,1194 82.5% <0.0001 Note: Association of Predicted Probabilities and Observed Responses

Cocordant = 82.5%, Somers'D 0.630, Hosmer and Lemeshovv Goodness-of-Fit Test = 7.3416 with 8 DF (p = 0.5003).

Examples of the proposed method

Example 1. Patient N, 56 years old, was hospitalized in the cardiology department No. 2 of the Altai Regional Cardiology Dispensary with a diagnosis of coronary heart disease (CHD): Angina pectoris III f. class Stenosis of the anterior descending artery (PNA) 80% in / 3. Ventricular and supraventricular extrasystoles. Chronic heart failure (CHF) I tbsp., II f. class In addition to the standard general clinical examination, patient N was determined by the thickness of epicardial adipose tissue, which was 10 mm, and an analysis of laboratory parameters was also performed: blood glucose was 6.7 mmol / l, lipoprotein level “a” was 600 mg / l, leptin level was 23.6 ng / ml, the level of interleukin-6, 13.5 PCG / ml, high density lipoprotein cholesterol 0.8 mmol / L. Using the proposed mathematical forecasting model built into the MS Excel, 2007 program, the probability of developing restenosis in this patient was calculated, the risk of development was 83%, which is a high gradation of risk. Stenting of the anterior descending artery was postponed for 1 month, during which the maximum possible correction of risk factors for restenosis was performed (statins, hypoglycemic therapy, plasmapheresis, anti-inflammatory therapy, diet therapy aimed at weight loss). After 1 month, 6 risk factors included in the mathematical model were re-determined: the thickness of the epicardial adipose tissue was 8 mm, the HDL cholesterol level was 1.0 mmol / L, the blood glucose was 5.6 mm / L, the lipoprotein level “a” was 400 mg / l, leptin level of 16.5 ng / ml, interleukin-6 level of 5.6 pg / ml. When re-evaluating the likelihood of developing restenosis, the risk of this complication was already 33%. The patient was successfully stented for the anterior descending artery, and the prescribed measures for the prevention of risk factors were continued. The patient was observed for 1 year in the office of rehabilitation treatment of the Altai Cardiac Center, no signs of restenosis of the stented coronary artery (PNA) were detected.

Example 2. Patient Y, 65 years old, was admitted to the cardiology department No. 2 of the Altai Cardiology Dispensary with a diagnosis of coronary heart disease: angina pectoris III f. class Stenosis of the right coronary artery 90% in sr / 3. A persistent form of atrial fibrillation. Chronic heart failure I st., II f. In addition to the standard general clinical examination, the patient was evaluated 6 risk factors that are included in the proposed mathematical model. The thickness of the epicardial adipose tissue was 8 mm, HLPV 0.9 mmol / l, blood glucose 7.1 mmol / l, lipoprotein “a” 500 mg / l, leptin 25.8 ng / ml, interleukin-6 14.5 pcg / ml In this patient, the estimated risk of developing restenosis was 81%. Despite the high probability of restenosis, the patient underwent stenting of the right coronary artery. After 7 months, the patient was re-admitted with the resumption of the clinic of angina pectoris, during the control diagnostic coronarography, coronary artery restenosis was diagnosed in the stenting area with a narrowing of the lumen of the coronary artery by 65%. The patient underwent balloon stent angioplasty.

To assess the practical effectiveness of using the proposed forecasting model, the risk of restenosis was calculated and the mathematically obtained risk was compared with the real event of restenosis. In the time elapsed since the end of the data set for building models, in the Altai Regional Cardiology Dispensary, stenting was performed in 120 patients who predicted the development of restenosis using this model, and then assessed the actual development of restenosis. Using the proposed method for predicting and the estimated risk of more than 50%, restenosis actually occurred in 80% of cases, which indicates the adequacy of the proposed method and the possibility of its use in practical health care, with a set of preventive interventions aimed at correcting risk factors with an estimated risk of restenosis of more than 50% . For the convenience and simplicity of the practical application of the proposed mathematical model, the authors created a probabilistic calculator of the development of restenosis based on the spreadsheet editor MS Excel as part of the standard software package MS Office 2007, using the above equation, into which (the calculator) the numerical characteristics of the analyzed forecast criteria are entered, and the desired value “Y” is displayed as a percentage automatically.

It was decided at the end of the examination of the application materials to predict coronary artery restenosis using the proposed method in order to reduce the risk of complications after stenting.

Claims (1)

1. A method for predicting the risk of developing coronary artery restenosis after stenting in patients with coronary heart disease, including examining a patient, diagnosing risk factors with deviations from their normal state, including the presence of high density lipoprotein cholesterol and interleukin-6 in the blood, characterized in that the diagnosis is carried out by determining the values of additional risk factors - the thickness of the epicardial adipose tissue, the levels of leptin, lipoprotein “a” and glucose in the blood, p What follows is carried out mathematical processing of the numerical values of risk factors to obtain the likelihood of restenosis by the appended predicting risk following:
$$P=\frac{\exp \left(-\mathrm{8,0248}+\mathrm{0,0893}\cdot X\mathrm{one}+\mathrm{0,00354}\cdot X2+0.1397\cdot X3+\mathrm{1,1194}\cdot X\mathrm{four}+0.3286\cdot X5+0.0665\cdot X6\right)}{\mathrm{one}+\exp \left(-\mathrm{8,0248}+\mathrm{0,0893}\cdot X\mathrm{one}+\mathrm{0,00354}\cdot X2+0.1397\cdot X3+\mathrm{1,1194}\cdot X\mathrm{four}+0.3286\cdot X5+0.0665\cdot X6\right)}$$

where P is the probability of occurrence of restenosis in%, X1 is the patient's leptin value in ng / ml; X2 - the value of lipoprotein "a" in mg / l; X3 is the thickness of the epicardial adipose tissue in millimeters; X4 - value of high density lipoprotein cholesterol in mmol / l; X5 - blood glucose value in mmol / l; X6 is the value of interleukin-6 in PCG / ml, -8.0248 is the free term of the equation.

Images