Outcome of Postcardiac Surgery Acute Myocardial Infarction and Role of Emergency Percutaneous Coronary Interventions

Background. Cardiac surgery carries a well-known risk of perioperative myocardial infarction (MI), which is associated with high morbidity and both in-hospital and late mortality. )e rapid haemodynamics deterioration and presence of myocardial ischemia early after cardiac surgical operations is a complex life-threatening condition where rapid diagnosis and management is of fundamental importance. Objective. To analyse the factors associated with mortality of patients with postcardiotomy MI and to study the role of emergency coronary angiography in management and outcome. Methods. We retrospectively enrolled adult patients diagnosed to have postcardiotomy MI and underwent emergency coronary angiography at our tertiary care hospital between January 2016 and August 2019. Results. Sixty-one patients from consecutive 1869 adult patients who underwent cardiac surgeries were enrolled in our study. )e studied patients had a mean age of 49± 16.2 years with a mean BMI of 29.5± 6.6 and 65.6% of them were males. As compared to the survivors group, the nonsurvivors of perioperative MI had significant preoperative CKD, postoperative AKI, longer CPB time, frequent histories of previous PCI, previous cardiotomies, pre and postoperative ECMO use, higher median troponin I levels, higher peak and 24 hours median lactate levels. Regression analysis revealed that reoperation for revascularization (OR: 23; 95% CI: 8.27–217.06; P � 0.034) and hyperlactataemia (OR: 3.21; 95% CI: 1.14–9.04; P � 0.027) were independent factors associated with hospital mortality after perioperative MI. Hospital mortality occurred in 25.7% vs 86.7% (P< 0.001), AKI occurred in 37.1% vs 93.3% (P< 0.001), haemodialysis was used in 28.6% vs 80% (P � 0.002), and mediastinal exploration for bleeding was performed in 31.4% vs 80% (P � 0.006) in the PCI and reoperation groups, respectively, while there were no significant differences regarding gastrointestinal bleeding, cerebral strokes, or intracerebral bleeding. )e median peak troponin level was 795 (IQR 630–1200) vs 4190 (IQR 3700–6300) (P< 0.001) in the PCI and reoperation groups, respectively. Absence of significant angiographic findings occurred in 18% of patients. Conclusions. Perioperative MI is associated with significant morbidities and hospital mortality. Reoperation for revascularization and progressive hyperlactataemia are independent predictors of hospital mortality. Emergency coronary angiography is helpful in diagnosis and management of perioperative MI.


Background
Cardiac surgery carries a well-known risk of perioperative myocardial infarction (MI), which is associated with high morbidity and both in-hospital and late mortality [1,2]. Perioperative MI is also associated with an increased use of healthcare resources, and it correlates with an increased risk of progressive heart failure in the years after cardiac surgery [3][4][5]. e aetiology of perioperative MI can be classified as graft-related and nongraft-related causes [6]. e commonest graft-related causes are anastomotic stenosis, graft spasm, acute graft thrombosis, and graft kinking or overstretching [7]. e nongraft-related causes include inadequate intraoperative cardioprotection, direct myocardial trauma by sutures, microvascular embolisms related to reperfusion, myocardial damage by the release of oxygen free radicals, and distal coronary microembolization [8][9][10][11].
Different studies about postcardiotomy MI have been performed with a wide range of reported incidence, 1%-26%, depending on the definition used and the type of surgical operation [12][13][14][15]. Criteria for diagnosing perioperative MI after cardiac surgeries other than CABG are still controversial and have not been collected in the Fourth Universal Definition of MI [16,17]. e purpose of this study was to analyse the factors associated with mortality of patients with postcardiotomy MI and to study the role of emergency coronary angiography in management and outcome.

Study Design.
is retrospective study was conducted at our tertiary care hospital where approximately 500 adult cardiac surgeries had been performed annually, and a cardiac surgical intensive care unit with 10 beds exclusively dedicated to the postoperative care of adult patients who had undergone cardiac surgeries.
Data were collected from a consecutive cohort of 1869 adult patients who underwent cardiac surgeries at our tertiary care hospital between January 2016 and August 2019. Cardiac operations were performed using standard techniques via cardiopulmonary bypass (CPB) with mostly central cannulation and mild systemic hypothermia (>32°C). Once the patients were transferred to the intensive care unit, 12-lead electrocardiogram (ECG) and high sensitivity troponin I serum levels were assessed every 8 hours until 48 hours. Transesophageal echocardiography (TEE) assessments were performed at the end of surgery (mandatory) and then at ICU (when needed) to investigate biventricular contractility, regional wall motion abnormalities, valves function, or any signs of pericardial tamponade.
All the patients who underwent emergent coronary angiography during the postoperative ICU stay were enrolled in our study. According to our hospital protocol, emergency coronary angiography was performed after detection of 1 or more of the following postoperative events: fatal ventricular arrhythmias (VT/VF), new onset of persistent ST segment changes on ECG, cardiac arrest or unexplained hemodynamic deterioration, new regional wall motion abnormality on transthoracic or transesophageal echocardiogram.
Patients with persistant haemodynamic deterioration despite maximal inotropic support were subjected to emergency implantation of venoarterial extracorporeal membrane oxygenation (VA-ECMO), and then, coronary angiograms were performed. We conducted this study using the hospital's Integrated Compliance Information System (ICIS) after getting approval from our Institute Ethics Committee.

2.2.
e Studied Variables. Demographic and baseline characteristics of patients, clinical, laboratory, ECG, echocardiographic, and angiographic data were collected. e peak troponin I level after cardiac surgery was recorded with a normal reference of 15-52 ng/L. e peak arterial lactate levels and the levels after 24 hours of ICU admission were collected. e sequential organ failure assessment (SOFA) score was calculated for all patients upon ICU admission and after 48 hours for organ failure assessment. e worst values for each variable in the SOFA score in the 24 hour-period were used during calculation. e patients were divided according to hospital mortality into the survivors and nonsurvivors groups. According to revascularization modality, we divided the patients into the PCI and reoperation groups.

Statistical
Analysis. Data were summarized using mean with standard deviation or median with interquartile range (IQR) for quantitative variables and frequency with percentage for qualitative variables. Normality of data was checked using the Kolmogorov-Smirnov normality test. Comparisons between numerical data were done using Student's t-test or Mann-Whitney accordingly. Ordinal variables were compared using the chi square test. P values less than 0.05 were considered statistically significant. Graphs were used to illustrate some information. Regression analysis was performed to get the mortality predictors. Statistical tests were performed using the Statistical Package of Social Science Software program, version 23 (SPSS).

Demographic and Clinical Characteristics.
A total of 61 patients with diagnosed postcardiotomy myocardial infarction and subjected to emergency coronary angiography were enrolled in this study. e mean age was 49 ± 16.2 years with a mean BMI of 29.5 ± 6.6, and 65.6% of them were males with no statistically significant differences between survivors and nonsurvivors. As compared to the survivors group, the nonsurvivors group had significantly more frequent CKD, preoperative ECMO support, emergency cardiotomies, histories of PCI, and cardiac surgeries. e nonsurvivors group had significantly higher aPTT and lesser GFR, while there were no significant differences regarding MI history, LVEF, type of surgeries, and other laboratory data as compared to the survivors group (Table 1).

Operative and ICU Characteristics.
e cardiopulmonary bypass time was 200.4 ± 68.4 vs 168.2 ± 84.3 (P = 0.04) in the nonsurvivors and survivors groups, respectively. Postoperative VA-ECMO use was significantly more frequent in the nonsurvivors group but without a significant difference in ECMO support days. e nonsurvivors group had significantly higher mean SOFA scores upon ICU admission and after 48 hours assessment. e median peak lactate level and lactate after 24 hours of ICU admission were significantly higher at the nonsurvivors group. e median peak troponin level was 3850 (IQR 2700-6300) vs 680 (IQR 580-893) (P < 0.001) in the nonsurvivors and survivors groups, respectively. PCI was performed in 39.1% vs 68.4% (P � 0.02) in the nonsurvivors and survivors groups, respectively, while reoperation (CABG) was performed in 56.5% vs 5.3% (P < 0.001) in the nonsurvivors and survivors groups, respectively, without significant differences in the median time to angiography after cardiac surgery. Presence of insignificant coronary lesions that did not require intervention occurred (4.3% vs 26.3%, P � 0.04) in the nonsurvivors and survivors groups, respectively. ere were significantly frequent mediastinal explorations for bleeding, AKI, and use of haemodialysis in the nonsurvivors group.
ere were no significant differences between both groups regarding ischemic cerebral strokes, gastrointestinal bleeding, or ICU days (Table 2).

Criteria of Acute Myocardial Ischemia.
e development of new bundle branch block occurred in 21.7% vs 2.6% (P � 0.02), VT/VF in 87% vs 55.3% (P � 0.01), and reduction of LVEF in 23.8 ± 9.2 vs 38.7 ± 7.8 (P < 0.001) in the nonsurvivors and survivors groups, respectively. ere were no statistically significant differences between both groups regarding new echocardiographic RWMAs, angiographic findings, ST segment elevation or depression. ere were no statistically significant differences between the survivors and nonsurvivors regarding the angiographic data (Table 3). e development of RWMAs had 96.2% sensitivity and 37.5% specificity with 91.1% positive predictive value, 60% negative predictive value, and 88.5% accuracy for predicting significant coronary lesions. e presence of ST segment elevation had 67.9% sensitivity and 98% specificity with 98% positive predictive value, 32% negative predictive value, and 72.1% accuracy for predicting significant coronary lesions. e presence of ST segment depression had 49.1% sensitivity and 0% specificity with 76.5% positive predictive value, 0% negative predictive value, and 42.6% accuracy for predicting significant coronary lesions.

Hospital Outcome according to Revascularization
Methods. After emergency coronary angiography, 18% of cases showed insignificant coronary lesions. ere were no complications related to coronary catheterizations. According to revascularization methods, there were 2 groups (PCI and reoperation for CABG). ere were no statistically significant differences between both groups regarding demographic data or CPB and aortic cross-clamping times. e use of IABP and ECMO was significantly frequent in the reoperation group. e median time of angiography after surgery was 7 hours (IQR 4-10) vs 3 hours (IQR 2-6) (P � 0.01) in the PCI and reoperation groups, respectively. e LM lesions were diagnosed in 20% vs 46.7% (P < 0.001) in the PCI and reoperation groups, respectively, while CX lesions were diagnosed in (42.9% vs 6.7%, P < 0.001) in the PCI and reoperation groups, respectively ( Table 4, Figures 1-3).
Hospital mortality occurred in 25.7% vs 86.7% (P < 0.001), AKI occurred in 37.1% vs 93.3% (P < 0.001), haemodialysis used in 28.6% vs 80% (P � 0.002), and mediastinal exploration for bleeding occurred in 31.4% vs 80% (P � 0.006) in the PCI and reoperation groups, respectively, while there were no significant differences regarding gastrointestinal bleeding, cerebral strokes, or intracerebral bleeding (Table 5, Figure 4). e reoperation group had significantly higher mean SOFA scores upon ICU admission and after 48 hours assessment as compared to the PCI group. e median peak lactate level and median lactate level after 24 hours of ICU admission were significantly higher at the reoperation group. e median peak troponin I level was 795 (IQR 630-1200) vs 4190 (IQR 3700-6300) (P < 0.001) in the PCI and reoperation groups, respectively (Table 5).

Predictors of Hospital
Mortality. Logistic regression analysis was performed to determine the predictors of hospital mortality after postcardiotomy myocardial infarction.

Discussion
e rapid deterioration of haemodynamics or presence of signs of myocardial ischemia early after cardiac surgical operations is a complex life-threatening condition where rapid and precise diagnosis and management are of fundamental importance. Criteria for diagnosing perioperative MI after cardiac surgeries other than CABG are still controversial and have not been collected in the Fourth Universal Definition of MI [16,17]. e diagnosis of perioperative MI is challenging, especially during the early postoperative period. In fact, ECG, cardiac biomarkers, and cardiac imaging can be misleading in the early postoperative hours [18]. However, it is generally accepted that the presence of new and persistent ECG alterations, new regional wall motion abnormalities on echocardiography, significant elevations in cardiac enzyme levels and hemodynamic instability despite high doses of inotropic and vasopressor drugs, or frequent ventricular arrhythmias are sufficient criteria to suspect perioperative MI [19].
Few studies described the benefits of emergency coronary angiography and interventions after CABG operations for acute graft failure. [6,[20][21][22] Gaudino et al. [23] described the effectiveness of emergency coronary angiography in diagnosis and management for myocardial ischemia after different cardiac surgeries. We found in our study that emergency coronary angiography had a role in early diagnosis of the aetiology of haemodynamics deterioration and myocardial ischemia. About 18% of patients with perioperative MI showed insignificant coronary lesions. Also, coronary angiography helped in management of 57.4% of cases by coronary interventions and stenting with fewer Data are presented as mean ± SD, median (IQR), or N (%).  Cardiology Research and Practice postoperative complications compared to the patients transferred to operation room for revascularization. We found that the presence of echocardiographic new RWMAs had higher sensitivity but lower specificity than the presence of new ST segment elevation for predicting angiographic coronary lesions in perioperative MI, while ST segment depression had lowest sensitivity and specificity for prediction of angiographic lesions. e benefits of echocardiography during and after cardiac surgeries include its ability to evaluate the cardiac contractility, valves function, detection of cardiac tamponade, and detection of regional wall motion abnormalities. [24][25][26].
We used the peak cardiac troponin I as a biomarker to detect the degree of myocardial damage as this study was retrospective with unknown baseline troponin and absence of standardized postoperative measurements patterns. In our cardiac surgical critical care unit, we do not depend on the value or pattern of troponin in patients with lifethreatening conditions after cardiotomy. Interestingly, the median troponin I was significantly higher in the nonsurvivors compared to the survivors and in the reoperation group compared to the PCI group. is may explain more extensive myocardial damage and worse clinical outcome in these groups. However, troponin was not significant in our multivariable regression analysis for mortality. Troponin was linked to mortality and worse outcomes after cardiac surgeries in several studies [27][28][29].
We evaluated our critically ill patients with the SOFA score and blood lactate levels, which reflect the haemodynamic profile and organs perfusion and functions. We used the SOFA score in assessment due to its simplicity and validity to predict mortality [30,31]. e mortality group had a significantly higher mean SOFA score upon ICU admission with an increased trend after 48 hours reflecting the multiorgan functions deterioration. e PCI group had a significantly lower mean SOFA score with a decreasing trend after 48 hours as compared to the reoperation group. We used the peak lactate level and the level after 24 hours of ICU admission to get data about organ perfusion. Hyperlactatemia has been linked to increased mortality among critically ill patients including those postcardiotomy [32]. e nonsurvivors had a higher median lactate peak level with delayed clearance after 24 hours compared to the survivors group. Also, the PCI group had a lower median peak lactate level with a lower median lactate level after 24 hours compared to the reoperation group. e hyperlactatemia with delayed clearance could reflect the impaired different organ perfusion and development of organ dysfunction and increased mortality after postcardiotomy MI especially in the reoperation group. Moreover, our regression analysis revealed that reoperation and progressive hyperlactataemia were the independent predictors of hospital mortality after perioperative MI.
In our study, the nonsurvivors group of perioperative MI had significantly preoperative CKD with GFR less than 60 (mL/min/1.73 m 2 ) and postoperative AKI and haemodialysis as compared to the survivors group. Our results were similar to other studies that confirmed the greater postcardiotomy morbidity and mortality with decreased renal functions [33,34]. Interestingly, we found that the frequency of AKI and use of haemodialysis were significantly greater in the reoperation group than in the PCI group despite the risk of contrast-induced nephropathy in those unstable patients. Sef et al. [22] reported the safety of PCI for perioperative MI after CABG without increased risk of acute renal failure.
Despite dual antiplatelet therapy, the PCI group did not have significantly different bleeding complications than the reoperation group, and we could not provide an explanation for this finding. Our nonsurvivors group had a significantly longer CPB time compared to the survivors group. e CPB time is well known as an independent risk factor of postcardiotomy morbidity and mortality [35]. Cubero-Gallego et al. [34] found that the CPB time was longer in patients with perioperative MI after valve surgery than those without perioperative MI.
Finally, rapid and precise diagnosis of postcardiotomy MI and possible revascularization should be done to avoid the bad consequences. According to our results, emergency coronary angiography was helpful in rapid diagnosis of coronary lesions and allowed interventions, which were associated with lesser myocardial damage, better haemodynamics, and organ functions compared to emergency reoperation for revascularization.

Conclusions
Perioperative MI is associated with significant morbidities and hospital mortality. Reoperation for revascularization and progressive hyperlactataemia are independent predictors of hospital mortality. Emergency coronary angiography is helpful in diagnosis and management of perioperative MI.

AKI:
Acute kidney injury aPTT: Activated partial thromboplastin time BBB: Bundle branch block BMI: Body mass index CABG: Coronary artery bypass graft CKD: Chronic kidney disease CI: Confidence interval CPB time: Cardiopulmonary bypass time CX: Circumflex artery DM: Diabetes mellitus IABP: Intra-aortic balloon pump ICH: Intracerebral haemorrhage LAD: Left anterior descending artery LM: Left main artery LVEF: Left ventricle ejection fraction MI: Myocardial infarction OR: Odds ratio PCI: Percutaneous coronary intervention RCA: Right coronary artery RWMAs: Regional wall motion abnormalities 8 Cardiology Research and Practice SOFA: Sequential organ failure assessment STEMI: ST segment elevation myocardial infarction TEE: Transesophageal echocardiography VA-ECMO: Venoarterial extracorporeal membrane oxygenation VT/VF: Ventricular tachycardia/fibrillation.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Additional Points
Our work was a retrospective study with possible selection bias without randomization because the choice of modality of revascularization (PCI versus reoperation) was individualized according to the clinical condition and circumstances. Even though the PCI and reoperation groups had apparently similar baseline characteristics and the PCI group had better outcomes and lesser mortality, we cannot conclude that PCI is better than reoperation as a recommendation. We just concluded that PCI was helpful in management.

Conflicts of Interest
e authors declare that there are no conflicts of interest.