Multifactorial prediction of adverse outcome of acute coronary syndrome in patients with post-COVID syndrome

Full Text

Introduction. According to global registries, acute coronary syndrome, both with and without ST-segment elevation, is the leading cause of morbidity and mortality in the Russian Federation and worldwide [1, 2, 3]. There are prognostic scales for assessing the risk of developing an unfavorable outcome of ACS in the world, but they only assess the outcome of acute coronary syndrome, without a combination with post-COVID syndrome. For example, the GRACE model is one of the most well-known. Along with the hospital risk of ACS complications, it allows you to assess the likelihood of a long-term unfavorable prognosis. Moreover, the discriminant power of the GRACE model significantly exceeds models based on the results of randomized clinical trials. However, in light of new data, the absence of biochemical parameters of peripheral blood, electrocardiography (ECG) and echocardiography (EchoCG) data, and genetic parameters for analysis in the GRACE model indicate its incompleteness and inconsistency with modern trends in science and practical healthcare [5]. It is known that the pandemic of a new coronavirus infection has adversely affected the course and outcome of cardiovascular diseases [4]. The World Health Organization (WHO) has identified post-COVID syndrome as a separate nosology, according to which it represents signs and symptoms that developed during or after a new coronavirus infection and continue after 12 weeks, which cannot be explained by any other cause [4]. Thus, patients with ACS in combination with PCS need a comprehensive prognostic model for assessing the adverse outcome of ACS for further prevention and individual rehabilitation. Objective: to build a multivariate model for predicting adverse outcomes of acute coronary syndrome with and without ST-segment elevation in patients with ACS. Material and methods. Study design: prospective cohort study. The study group included 118 people, including 61 men and 57 women. The average age of women was 57.5 ± 6.2 years, men - 53.7 ± 8.3 years. The comparison group consisted of 121 patients with acute coronary syndrome without a history of new coronavirus infection. All patients were comparable in gender and age, and were urgently delivered to the regional vascular center by an ambulance team. Upon admission, all patients were diagnosed with acute coronary syndrome. This diagnosis was made based on the clinical guidelines "Acute coronary syndrome with ST segment elevation of the electrocardiogram [2] and Acute coronary syndrome without ST segment elevation of the electrocardiogram" [3], approved by the Scientific and Practical Council of the Ministry of Health of the Russian Federation. The inclusion criterion for the study was the presence of a history of new coronavirus infection (NCI) that met the criteria for the diagnosis of "Post-COVID syndrome" specified in the methodological guidelines "Characteristics of the course of long-COVID infection. Therapeutic and rehabilitation measures" [6, 7]. In accordance with the amendments made to the International Classification of Diseases (ICD-10), post-COVID syndrome occurs in individuals after coronavirus infection with confirmed SARS-CoV-2 infection 3 months after the onset of COVID-19. In patients included in the study, the diagnosis of previous COVID-19 was established in accordance with the recommended laboratory diagnostic methods specified in the temporary clinical guidelines "Prevention, diagnosis and treatment of a new coronavirus infection (COVID-19)", version 18 (10.26.2023), approved by the Scientific and Practical Council of the Ministry of Health of the Russian Federation [6, 7]. The comparison group consisted of 121 patients (including 62 men and 59 women) diagnosed with ACS without post-COVID syndrome (there was no indication in the anamnesis of the presence of a diagnosis of NCI confirmed by a PCR smear or the detection of immunoglobulins of classes A, M, G (IgA, IgM and IgG) to SARS-CoV-2 by the immunochemical method [6, 7]). To predict the risk of developing an unfavorable ACS, logistic regression analysis was used. To build a logistic regression model, the following equation was used (1) P = 1 / (1 + e-y), (1) where P is the probability of the index event; e is the base of natural logarithms (Euler's number), equal to 2.718; y is the standard regression equation. The standard regression equation was represented by the following formula (2) y = a + b1X1 + b2X2+...+bnXn, (2) where a is a constant; b are the regression coefficients; X are the initial variables. The value of X was represented by quantitative or qualitative variables. Qualitative variables were taken as a binary variable, where 1 is the presence of a factor and 0 is the absence of a factor. Using logistic regression by the stepwise inclusion of statistically significant factors (variables), a predictive model was built. The statistical significance of the model is determined was divided by the χ² criterion. At p ˂ 0.05, the null hypothesis of model insignificance was rejected. The cutoff threshold after forming the index event development model was 0.5. The sensitivity and specificity of the model were assessed using ROC analysis. The result was interpreted by constructing ROC curves with an assessment of the area under the ROC curve (AUC). The study was approved by the local ethics committee of the Novosibirsk State Medical University of the Ministry of Health of the Russian Federation (protocol No. 155 dated November 29, 2023, Novosibirsk), and also approved at a meeting of the problem commission (protocol No. 1 "Current issues of prevention, diagnosis and treatment of internal diseases" dated October 25, 2023). Each patient signed informed consent to participate in the study in accordance with the ethical requirements of the World Health Organization. Results. Modeling of an unfavorable outcome (unfavorable ACS) was carried out on the basis of calculating the probability of its development, using logistic regression analysis. To build a logistic regression model, equation (1) was used. The standard regression equation is presented by formula (2). The initial variables were clinical and anamnestic (gender, age, weight, degree of obesity (if any), duration of pain syndrome, nature of pain, localization of pain, presence of arterial hypertension (AH), presence of coronary heart disease (CHD), presence of chronic heart failure (CHF), presence of a previous new coronavirus infection (NCI) and its absence, severity of the previous NCI, the previous NCI wave (alpha, delta, omicron), type of therapy, smoking, severity of the previous cardiovascular event), instrumental, laboratory parameters. Using logistic regression by the stepwise inclusion of statistically significant factors (variables), a prognostic model was constructed. The statistical significance of the model was determined by the X2 criterion. At p ˂ 0.05, the null hypothesis of the insignificance of the model was rejected. The value of 0.5 was adopted as the cutoff threshold after forming the model for the development of the index event. It should be noted that, taking into account the significance criterion (Wald), the most statistically significant predictors were: the value of hypokinesis, the combination of ACS and PCS (Table 1). Then followed the tyrosine kinase level, the presence of the ACE 2 TT/AA gene and the presence of CHF. Table 1 - The main results of the analysis of binary logistic regression of the prognosis of the development of unfavorable ACS Predictor B (regression coefficient) МSE (root mean square error) Wald (Wald statistics, X2) p (significance level) Exp (B) Belonging to the group "ACS and PCS" (X1) -1.689 0.48 12.362 0.0004 0.185 soluble fms-like tyrosine kinase-1 (X2), pg\ml 0.039 0.01 8.237 0.004 1.04 Presence of CHF (X3) 0.870 0.39 4.894 0.027 2.388 Hypokinesia (X4), damage to segments of the heart muscle 0.082 0.02 17.983 0.00002 1.085 Presence of ACE 2 TT/AA gene polymorphism (X5) -1.286 0.51 6.419 0.011 0.276 Thus, the predicted probability of developing a combined endpoint, including the total number of cardiovascular complications and fatal outcomes, was presented as formula (3): P = 1 / (1 + 2.718 - (13.153 - 1.689 × X1 + 0.039 × X2 + 0.870 × X3 + 0.082 × X4 - 1.286 × X5), (3) where X1 is the belonging to the group "ACS with PCS», X1 = 0 – patient with ACS without PCS, X1 = 1 – patient with ACS and PCS, X2 – soluble fms-like thyroxine kinase-1 (X2), pg\ml, X3 – belonging to the group «Presence of CHF», X3 = 0 – patient without signs of CHF, X3 = 1 – patient with signs of CHF, X4 – hypokinesia, X5 – presence of ACE 2 TT/AA gene polymorphism, X5=0 – patient does not have this gene, X5=1 – patient has this gene. When obtaining the final result, for clarity, the resulting number is multiplied by 100%. The Hosmer – Lemeshow goodness-of-fit criterion for this prognostic model was X2 = 38.416, p = 0.0000, which characterizes very high significance. Then the ROC curve was constructed. According to the ROC curve construction data, the area under the ROC curve is 0.8 (Figure 1), so the quality of the model can be assessed as good - an acceptable model. The sensitivity of the model (the proportion of correctly classified patients with the development of adverse ACS) was 93.5, and the specificity (the proportion of correctly classified patients without adverse ACS) was 21.8. The total proportion of correctly predicted complications is 76.6%. Thus, the resulting model perfectly predicts the presence of adverse ACS, but poorly predicts its absence. Figure 1 - ROC curve graph for predicting the development of adverse ACS in patients with PCS ROC curve graph for predicting the development of adverse ACS in patients with PCS Let's give a clinical example of the applicability of the constructed model. The patient is a 68-year-old woman, with a history of hypertension stage III, arterial hypertension 3, risk of cardiovascular complications 4. Complaints of increased blood pressure to 180/95 mm Hg, does not control blood pressure, takes medications irregularly, does not undergo medical examinations, is seen by doctors irregularly. She was admitted to the regional vascular center by ambulance with complaints of pressing burning pain behind the sternum for 30 minutes, radiating to the left arm, under the left shoulder blade, to the left half of the lower jaw, not relieved by taking nitroglycerin. At the prehospital stage, morphine was administered to relieve pain, acetylsalicylic acid 300 mg, clopidogrel 600 mg, heparin 5000 U. According to the ECG upon admission - sinus rhythm, heart rate 100 bpm. The electrical axis of the heart is horizontal. Voltage is unchanged. ST segment elevation in standard lead III up to 2 mm, avF, chest leads V 4 – V6. Troponin over 10 (normal up to 0.3 ng/ml). Upon admission, the condition is assessed as severe. Hemodynamics are stable, pressure is maintained independently. BP 105/60 mm Hg. HR 98 beats per minute. Pulse is rhythmic, weak filling. Respiratory rate 20 per minute, independent breathing is harsh, conducted along all lung fields, isolated weak wheezing is heard. SpO2 95%. The abdomen is soft and painless on palpation. There are no focal neurological symptoms. Thus, a diagnosis of acute coronary syndrome with ST segment elevation was established. Bypassing the emergency room, the patient was examined by a cardiologist, an anesthesiologist-resuscitator, an X-ray endovascular surgeon, and taken to the X-ray operating room. A study was conducted to calculate the probability of an unfavorable outcome using a multivariate regression model, which yielded the following results: X1 - the patient has not had a new coronavirus infection (no PCS) (0) X2 - the level of soluble fms-like tyrosine kinase-1 is 70 pg / ml, X3 - the patient has no signs of CHF (0) X4 - echocardiography data did not reveal hypokinesis zones (0) X5 - genetic analysis to determine the nucleotide sequence variant did not show the presence of ACE 2 TT / AA gene polymorphism (0) Thus, our multivariate analysis formula looks like this: P = 1 / (1 + 2.718 - (13.153 - 1.689 × 0 + 0.039 × 70 + 0.870 × 0 + 0.082 × 0 -1.286×0), y= 13.153+ (-1.689*0) + 0.039*70 + 0.870*0 + 0.082*0 + (-1.286*0), y= 13.153 +0 + 2.73 +0+0+0 y= 15.883. P = 1 / 1 + 2.718 - 15.883, => 1/1 + 1.27 = 1/2.27 = 0.44, => P= 0.44*100% = 44% - the probability of developing an unfavorable outcome in this patient. Unlike the standard situation, in this clinical case, taking into account the unfavorable prognosis of ACS, the treatment was changed, namely, increased antiplatelet therapy: ticagrelor 90 mg 2 times a day, heparin was added 1000 U intravenously bolus in the first 12 hours, then transfer to infusion, acetylsalicylic acid 100 mg. A double dose of antiplatelet therapy was given in full, the "door-to-balloon" time was 49 minutes (the time from admission to a medical facility (by ambulance or from the emergency room) to recanalization of the coronary artery with a balloon or stent), based on the results of coronary angiography, angioplasty and stenting of the infarction-related vessel - the right coronary artery - were performed. The operation was successful, the patient was transferred to the intensive care unit in a stable condition for the purpose of cardiac monitoring and observation of vital functions. Given the high risk of death after the operation, the patient's condition was constantly monitored. 12 hours after the operation, the patient developed shortness of breath, interruptions in the heart's work, pressing pain behind the sternum, an attack of profuse sweating. The monitor shows atrial fibrillation, 146-166 beats per minute. The patient was immediately given 150 mg of cordarone intravenously by jet stream, then 300 mg of cordarone + 5% 250.0 mg of glucose solution intravenously by drip, the rhythm was restored after 15 minutes, the patient was transported to the X-ray operating room, where coronary angiography data did not reveal stent obstruction. In this clinical case, constant monitoring of the condition, due to the high risk of death, allowed for timely detection and prompt treatment of a complex heart rhythm disorder. On the 3rd day, taking into account the stabilization of the condition and relief of life-threatening heart rhythm disturbance, the patient was transferred to the cardiology department of the RSC to continue treatment: clopidogrel 75 mg once a day (taking into account the heart rhythm disturbance), rivaroxaban 20 mg in the morning (normal SCF), acetylsalicylic acid 100 mg per day, atorvastatin 80 mg. On the seventh day, in a stable condition and without signs of cardiovascular circulatory failure, she was discharged from the hospital under the supervision of a general practitioner and a cardiologist at the place of residence and for the purpose of placing her on preferential provision of vital drugs. Discussion. To date, the literature has described single prognostic models aimed at assessing the outcome of ACS [3, 8]. One of the first prognostic models was the prognostic system proposed in 1962 [9]. It is based on calculation of the prognostic index based on the characteristics of the acute period of ACS. The obtained data predicted the possibility of unfavorable development of the disease within 28 days from the onset of the disease. Modern prognostic models include a model created using the regression analysis method based on the GRACE IM registry data [9, 10]. The model includes 8 indicators that were obtained after analyzing the registry data: patient age, heart failure class according to the Killip classification, systolic blood pressure level, heart rate, creatinine level, diagnostic level of myocardial necrosis biomarkers, ST segment changes, the presence of at least 1 episode of cardiac arrest [5]. The GRACE model is one of the most well-known. Along with the hospital risk of ACS complications, it allows you to assess the likelihood of a remote unfavorable prognosis. Moreover, the discriminant power of the GRACE model significantly exceeds models based on the results of randomized clinical trials. However, in light of new data, the absence of biochemical parameters of peripheral blood, ECG and ECHO-kg data, and genetic parameters for analysis in the GRACE model indicate its incompleteness and inconsistency with modern trends in science and practical healthcare [5]. Another domestic model aimed at assessing adverse outcomes in patients with ST-segment elevation myocardial infarction takes into account the importance of renal function [11]. According to the authors, renal dysfunction is one of the main factors determining the unfavorable prognosis of myocardial infarction. However, none of the existing prognostic models aimed at assessing the risk of adverse outcomes of acute coronary syndrome take into account the presence of post-COVID syndrome in patients, unlike our multivariate regression model. Conclusion. A multifactorial model for predicting the risk of adverse outcome of acute coronary syndrome with and without ST segment elevation in patients with ACS has been obtained. The use of the proposed model in healthcare optimizes the tactics of management and treatment of patients in this nosological group, and will also help prevent the risk of not only a fatal outcome, but also the development of complications, and improve the prevention and cardiac rehabilitation of patients.

About the authors

Valentina Kozik

Federal State Budgetary Educational Institution of Higher Education Novosibirsk State Medical University of the Ministry of Health of the Russian Federation (Federal State Budgetary Educational Institution of Higher Education NGMU of the Ministry of Health of the Russian Federation)

Author for correspondence.
Email: valiyta90@mail.ru
ORCID iD: 0000-0001-7128-7887
SPIN-code: 2648-5319

PhD student of NGMU
Cardiologist
Secretary of the executive committee "Young Cardiologists" of the Russian Society of Cardiology in the Siberian Federal District

Russian Federation

Lubov Shpagina

Федеральное государственное бюджетное образовательное учреждение высшего образования «Новосибирский государственный медицинский университет» Минздрава России

Email: mkb-2@yandex.ru
ORCID iD: 0000-0003-3446-8018
SPIN-code: 5773-6649

Head of the Department of Hospital Therapy and Medical Rehabilitation, Professor, Doctor of Medical Sciences

Russian Federation, 630091, Russian Federation, Novosibirsk region, Novosibirsk, Krasny prospect, 52

Iliya Shpagin

Федеральное государственное бюджетное образовательное учреждение высшего образования «Новосибирский государственный медицинский университет» Минздрава России

Email: dr.ilya.shpagin@gmail.com
ORCID iD: 0000-0002-3109-9811
SPIN-code: 2892-6184

Professor of the Department of Hospital Therapy and Medical Rehabilitation, Doctor of Medical Sciences

Russian Federation, 630091, Russian Federation, Novosibirsk region, Novosibirsk, Krasny prospect, 52

References

Supplementary files