RU2770819C1 - Method for determining the probability of occurrence of a hospital adverse event after coronary bypass surgery - Google Patents

Abstract

FIELD: medicine.SUBSTANCE: invention relates to the field of medicine, namely cardiology. The patient’s Disse insulin resistance index is determined, as well as age in years, the size of the left atrium, the time of aortic compression in minutes, after which the probability of an adverse event P is calculated according to the formula, substituting the values obtained into it.EFFECT: method makes it possible to predict the hospital outcomes of coronary bypass surgery.1 cl, 2 tbl, 3 ex

Description

The present invention relates to the field of medicine, namely cardiology, and can be used in the assessment of operational risk in preparing patients for open surgery on the vessels of the heart by determining additional blood parameters and calculating the insulin resistance index.

Insulin resistance (IR) is a violation of the metabolic response to endogenous or exogenous insulin. This condition leads to an increased concentration of insulin in the blood plasma compared to the physiological values for the available glucose concentration. Over time, insulin resistance leads to depletion of beta cells and the development of type 2 diabetes. Insulin resistance negatively affects the course of cardiovascular diseases. It is known that prediabetes and diabetes mellitus are more common among patients with coronary heart disease.

The most accurate method, recognized as the "gold standard" for assessing insulin resistance, is the euglycemic hyperinsulinemia clamp proposed by Andres R. et al. in 1966 and developed by DeFronzo K. et al. In 1979 [DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance // American Journal of Physiology, 1979, 237 (3): 214-223]. For the assessment of insulin resistance, the clamp is considered the most reliable and reproducible both in DM and in healthy people. The clamp test technique involves injecting insulin at a constant rate intravenously to achieve a sufficient level of hyperinsulinemia (50-400, on average 100 µU/ml) to suppress hepatic glucose production and insulin secretion and maintain glycemic levels at a constant normal level by changes in the rate of glucose administration. Typically, the rate of insulin infusion is 40 mU/m2 ^of body surface per minute, or approximately 1 mU/kg/min. Measurement of glycemia is performed every 5-10 minutes on glucose analyzers or constant monitoring of the level of glycemia is used using an artificial pancreas apparatus (“Biostator”). After 120-240 minutes, dynamic equilibrium is reached: the rate of glucose administration is equal to the rate of its uptake by tissues. Thus, the total amount of glucose administered over the last 60-120 minutes of the study at steady state characterizes the insulin sensitivity index. A constant level of glycemia and the rate of glucose infusion into the state of dynamic equilibrium of the introduction and consumption of glucose are maintained for 60 minutes. The total duration of the study is 4-6 hours.

Another option for measuring insulin resistance is the minimal model developed by Bergman et al. in 1979, where frequent determinations of glucose and insulin are carried out during an intravenous glucose tolerance test for 180 minutes, or a shortened version (FSIGTT, OSIG) [Bergman R.N., Ider Y.Z., Bowden R., Cobelli C. Quantitative estimation of insulin sensitivity // American Journal of Physiology, 1979, 236 (6): 667-677; Mari A., Pacini G., Murphy E., Nolan J.J. et al. A model-basedmethods for assessing insulin sensitivity from the oral glucose tolerance test. // Diabetes Care, 2001, 24: 539-548. The results are entered into a computer model (MINMOD) based on certain assumed principles of glucose and insulin kinetics. In healthy people, the results significantly correlate with the data of the clamp method, however, in diabetes mellitus, there are serious limitations to its use. Due to the weakening of stimulated insulin secretion in response to glucose administration, initial hyperglycemia, and a sharp decrease in insulin sensitivity, often the indices of the minimum model are close to zero. In addition, there is a greater variability of results than with the clamp, and its cost remains high, the minimum test is technically complex and lengthy.

In addition, various indices have been proposed for assessing IR, calculated from the ratio of fasting plasma insulin and glucose concentrations (HOMA-IR, Sago, Quantitative Insulin Sensitivity Chek Index (QUICKI) [A.Yu. Mayorov, K.A. Urbanova, G.R. Galstyan Methods for Quantifying Insulin Resistance Obesity and Metabolism 2009;6(2):19-23 https://doi.org/10.14341/2071-8713-5313] They are technically easy to use, the simplest formulas are applied in clinical practice for an approximate determination of insulin sensitivity, are used for scientific research.The Disse insulin resistance index used in this invention also refers to the calculated indices, it was proposed by Disse E. in 2008, and includes, in addition to glucose and insulin indicators, lipid parameters: free fatty acids, serum high-density lipoprotein cholesterol (HDL-C), total cholesterol [Disse E, Bastard JP, Bonnet F, Maitrepierre C, Peyrat J, Louche-Pelissier C, Laville M A lipid-parameter-based index for estimating insulin sensitivity and identifying insulin resistance in a healthy population. Diabetes Metab. Nov 2008; 34(5): 457-63. doi: 10.1016/j.diabet.2008.02.009].

The second part of the invention is to use the resulting value of the Disse insulin resistance index to predict the results of coronary bypass surgery. Coronary artery bypass surgery is a major open heart surgery in which narrowed or closed coronary vessels are bypassed by introducing healthy blood vessels. Bypass surgery is one of the most common surgical procedures in the treatment of cardiovascular diseases. Coronary artery bypass grafting is necessary if one or more of the main coronary arteries are narrowed or blocked and there is significant circulatory failure of the heart that cannot be corrected by minimally invasive percutaneous intervention or medication. Predicting the outcomes of coronary bypass surgery is an important task in cardiology and cardiovascular surgery. Scales for assessing the risk of surgery based on the assessment of the severity of the patient's condition before surgery have been developed.

Currently accepted worldwide is the EuroScore II (European System for Cardiac Operative Risk Evaluation) scale designed to assess the risk of an adverse outcome of coronary bypass surgery. It takes into account the following parameters: age, female gender, elevated serum creatinine, extracardiac artery pathology, chronic lung disease, serious neurological disorders, previous cardiac surgery, myocardial infarction, left ventricular dysfunction, chronic congestive heart failure, pulmonary hypertension, active endocarditis, unstable angina pectoris, emergency surgery, critical condition of the patient before surgery, postinfarction rupture of the interventricular septum, surgery on the thoracic aorta. Each of these factors was evaluated. From the sum of these scores, the EuroScore risk indicator is formed. An algorithm for determining the probability of death was also used, based not on a simple summation of risk factors, but on a more complex and flexible logistic analysis. The EuroScore II scale is suitable for calculating the risk of mortality during surgery and is not suitable for assessing the risk of non-lethal complications, which are important for the further prognosis and quality of life of the patient and greatly increase the cost of staying in a cardiac surgery hospital.

Closest to the claimed method of prediction is the method for predicting the outcomes of coronary artery bypass grafting with the determination of insulin resistance, described in the article by researchers Sato 2010 [N. Sato, G. Carvalho, T. Sato, R. Lattermann, T. Matsukawa, T. Schricker. The Association of Preoperative Glycemic Control, Intraoperative Insulin Sensitivity, and Outcomes after Cardiac Surgery. Clin Endocrinol Metab 2010; 95: 4338-4344, https://doi.org/10.1210/jc.2010-0135]. In this study, patients with and without diabetes mellitus in two cardiac surgery clinics (Canada and Japan) preparing for heart surgery were subjected to insulin resistance testing by euglycemic clamp test, the methodology of which is described above. The findings of this study are that clamp-assessed insulin resistance is associated with an increased risk of complications, regardless of the patient's diabetic state [N. Sato, 2010]. The clamp test used to predict the outcomes of CABG in this study has a number of disadvantages described above: the complexity and duration of the procedure, the requirement for a large amount of technical, material, time and labor resources - simultaneous intravenous administration of insulin and glucose is carried out for 4-6 hours with constant rate and glycemic assessment.

Such shortcomings are deprived of the calculated method for determining insulin resistance - the Disse index, which requires only a single sampling of 10 ml of blood from the cubital vein in addition to the routine examination before surgery. The technical result of the invention is a method for predicting hospital outcomes of coronary artery bypass grafting using the calculated index IR Disse.

708 consecutive patients who underwent coronary bypass surgery in 2011-2012 at the surgical clinic of the Research Institute for Complex Problems of Cardiovascular Diseases were included. A further sample included 383 patients in whom additional blood markers were determined, namely free fatty acids and fasting insulin in plasma, and insulin resistance indices were calculated Disse, Quantitative Insulin Sensitivity Chek Index (QUICKI), revised - QUICK! 284 (74%) - men, median age - 59.0 [54.5; 64.0]. Patients with type 2 diabetes were 125 (32.6%), 67 were prediabetic (17.4%). Isolated CABG without combined operations was performed in 322 (84.0%) patients, CABG under cardiopulmonary bypass in 346 (90.3%), "on a beating heart" - in 37 (9.7%).

The calculation of the Disse index was made on the basis of several blood parameters, performed with a single blood sampling from the cubital vein on an empty stomach on any day before the operation. The Disse index is calculated by the equation Disse=12×(2.5×{[serum high-density lipoprotein cholesterol (HDL-C) / total cholesterol (mmol/l)] - [serum free fatty acids (FFA) (mmol/l)]} ) - fasting serum insulin (IU / ml). The value of the Disse index is always less than zero, an increase in the index value corresponds to a decrease in insulin resistance (IR). The median of the Disse index was -12.87 [-20.13; -8.31]. There are no reference values for this index, as well as for other calculated indices of insulin resistance. The means used (equipment, reagents): standard equipment of the clinical diagnostic laboratory, for the determination of free fatty acids - Thermo Fisher Sientific reagents (Germany), for insulin - the Accu-Bind ELISA Microwells test system from Monobind Inc BCM Diagnostics (Germany).

Statistical processing was carried out using the standard software package STATISTICA 8.0. The normality of distribution of quantitative data was checked using the Shapiro-Wilk test. Due to the fact that the distribution of all quantitative characteristics differed from normal, they were described using the median, indicating the upper and lower quartiles (25th and 75th percentiles).

Direct selection binary logistic regression (likelihood ratio) was used to determine predictors of long hospital stay or poor outcome. The level of critical significance (p) in the regression analysis was taken equal to 0.05. The composite endpoint was hospital stay after surgery for more than 10 days or the occurrence of any significant postoperative complication, including death.

The novelty of the invention.

The Disse Insulin Resistance Index has never been used to assess the risk of heart surgery. Binary logistic regression analysis was performed to identify factors associated with hospital outcome - the composite endpoint - (Table 2). The composite endpoint was a hospital stay of more than 10 days or any of the significant perioperative complications listed in Table 1. The following factors were included in the initial analysis model: gender, age, BMI, diabetes mellitus, any CMD, overweight or obesity, Disse index , echocardiography parameters (aorta size, LV myocardial mass, LV myocardial mass index, Echocardiography parameters - E / A ratio, Vf, e'), biochemical parameters (glucose, triglycerides), heart rate, heart failure, NYHA grade, a also the duration of the operation and the time of cross-clamping of the aorta.

An increase in left atrial size, age, aortic cross-clamping time, and a decrease in Disse index were associated with a significant increase in the likelihood of developing a combined endpoint. All other factors included in the model showed no association with hospital outcome. The addition of the Disse index increased the significance of the model (Nagelkerke's R-square at step 4 was 0.743 (Table 2).

Note to table 2:

a. Variables introduced in step 1: LA.

b. Variables introduced in step 2: Age.

c. Variables introduced in step 3: AoClampTime.

d. Options entered in step 4: Disse.

LA - left atrial size, AoClampTime - aortic clamping time, Disse - Disse index

In addition, other well-known calculated indexes for determining insulin resistance - Cago, QUICKI (Quantitative Insulin Sensitivity Chek Index), Revised-QUICKI - were determined in this study, but none of them showed an association with the studied outcome.

The calculation formula is as follows:

P=1÷(1+(exp- ^{(-0.144*(ind. Disse)+0.276*age +4.429*left atrium+(-0.073)*aortic time-30.057))} ,

It is based on the logit transformation P=1÷(1+exp ^-y )

where P is the probability of an adverse event occurring, e is the base of natural logarithms 2.71; y - standard regression equation (y=b0+b1x1+b2x2+…+bnxn). In this case, y=b0+b1*Disse index+b2*age (years)+b3*left atrial size (cm)+b4*aortic cross-clamping time (min), b 0 is a constant (Table 1). Substituting all the values from the binary regression equation (table 1) we get the formula described above.

Described prediction method:

Example 1. A 60-year-old man, aortic cross-clamping time 57 minutes, Disse index - 4.95, left atrial size 3.6 cm. This patient had no complications or hospital stay for more than 10 days. We calculate according to the formula according to the patient's indicators:

P=1÷(1+(exp- ^{(-0.144*(ind. Disse)+0.276*age +4.429*left atrium+(-0.073)*aortic time-30.057))} ,

y=-30.057+(-0.144*(-4.950)+0.276*60+4.429*3.6+(-0.073)*57=-1.714

P=1÷(1+(exp- ^{(-0.144*(-4.950)+0.276*60+4.429*3.6+(-0.073)*57-30.057))} =1÷(1+(exp ^{(-1.714 )} )=0.152

The probability of occurrence of the event of interest is 15.2%.

Example 2. A 57-year-old man, aortic cross-clamping time 55 minutes, Disse index - 16.034, left atrial size 4.3 cm. This patient had a significant hospital complication (progression of renal failure against the background of chronic kidney disease) and he was in the hospital after CABG 21 day.

P=1÷(1+(exp- ^{(-0.144*(ind. Disse)+0.276*age +4.429*left atrium+(-0.073)*aortic time-30.057))} ,

y=-30.057+(-0.144*(-16.034)+0.276*57+4.429*4.3+(-0.073)*55=3.032

P=1÷(1+(exp- ^{(-30.057+(-0.144*(-7.429)+0.276*52+4.429*4.3+(-0.073)*44} =1÷(1+(exp ^{(-3.032 )} )=0.954.

The probability of the occurrence of the event of interest is 95.4%.

Example 3. A 52-year-old man, aortic cross-clamping time 44 minutes, Disse index - 16.034, left atrial size 4.3 cm. This patient had a significant hospital complication from a sternal wound and was in the hospital for 30 days.

y=-30.057+(-0.144*(-7.429)+0.276*52+4.429*4.3+(-0.073)*44=1.197

P=1÷(1+(exp- ^{(-30.057+(-0.144*(-7.429)+0.276*52+4.429*4.3+(-0.073)*44} =1÷(1+(exp ^{(-1.197 )} )=0.954.

The probability of the occurrence of the event of interest is 76.8%.

Claims (4)

A method for determining the probability of an adverse event occurring after coronary bypass grafting, characterized in that the patient's insulin resistance index Disse, age in years, size of the left atrium, aortic clamping time in minutes are determined, after which the probability of an adverse event is calculated according to the formula, substituting the obtained values into it : P=1÷(1+(exp- ^{(-0.144*(ind. Disse)+0.276*age +4.429*left atrium+(-0.073)*aortic time-30.057))} , where P is the probability of an adverse event occurring, e - base of natural logarithms 2.71; (-30.057) is a constant.