Cardiac troponin T is an important predictor of mortality after cardiac surgery

Machine learning-based model development for predicting risk factors of prolonged intra-aortic balloon pump therapy in patients with coronary artery bypass grafting

Machine learning algorithms are frequently used to clinical risk prediction. Our study was designed to predict risk factors of prolonged intra-aortic balloon pump (IABP) use in patients with coronary artery bypass grafting (CABG) through developing machine learning-based models. Patients who received perioperative IABP therapy were divided into two groups based on their length of IABP implantation longer than the 75th percentile for the whole cohort: normal (≤ 10 days) and prolonged (> 10 days) groups. Seven machine learning-based models were created and evaluated, and then the Shapley Additive exPlanations (SHAP) method was employed to further illustrate the influence of the features on model. In our study, a total of 143 patients were included, comprising 56 cases (38.16%) in the prolonged group. The logistic regression model was considered the final prediction model according to its most excellent performance. Furthermore, feature important analysis identified left ventricular end-systolic or diastolic diameter, preoperative IABP use, diabetes, and cardiac troponin T as the top five risk variables for prolonged IABP implantation in patients. The SHAP analysis further explained the features attributed to the model. Machine learning models were successfully developed and used to predict risk variables of prolonged IABP implantation in patients with CABG. This may help early identification for prolonged IABP use and initiate clinical interventions.

Risk factors for postoperative myocardial injury-related cardiogenic shock in patients undergoing cardiac surgery

BACKGROUND

Myocardial injury-related cardiogenic shock (MICS) is significantly associated with poor outcomes in patients after cardiac surgery. Herein, we aimed to investigate the risk factor for postoperative MICS.

METHODS

We performed a case-control study on 792 patients undergoing cardiac surgery from 2016 to 2019, including 172 patients with postoperative MICS and 620 age- and sex-matched controls. MICS was defined as composite criteria: a cardiac index of < 2.2 L/m²/min, arterial lactate levels of > 5 mmol/L at the end of the surgery, a vasoactive-inotropic score of > 40 at the end of the surgery, and a cardiac troponin T (cTnT) level of > 0.8 µg/L on postoperative day 1 (POD1) with an increase of > 10% on POD 2.

RESULTS

A total of 4671 patients who underwent cardiac surgery in our hospital between 2016 and 2019 were included; of these, 172 (3.68%) had MICS and the remaining 4499 did not. For investigating the risk factors, we selected 620 age- and sex-matched controls. In the univariate analysis, MICS was significantly associated with death (P < 0.05), extracorporeal membrane oxygenation (P < 0.05), continuous renal replacement therapy (P < 0.01), and ventricular arrhythmias (P < 0.05). Multivariable logistic regression analysis revealed that diabetes mellitus (OR:8.11, 95% CI: 3.52-18.66, P < 0.05) and a cardiopulmonary bypass (CPB) time of > 2 h (OR: 3.16, 95% CI: 1.94-5.15, P < 0.05) were associated with postoperative MICS. Moreover, long-time administration of preoperative calcium channel blocker (CCB) was associated with a less incidence of MICS (OR: 0.11, 95% CI: 0.05-0.27, P < 0.05).

CONCLUSIONS

Postoperative MICS is significantly associated with poor outcomes. Diabetes mellitus and long CPB time are associated with MICS. Preoperative CCB administration is associated with less incidence of MICS.

Procedure-Specific Relationships Between Postoperative Troponin T and a Composite of Mortality and Low Cardiac Output Syndrome: A Retrospective Cohort Analysis

BACKGROUND

Myocardial injury after coronary artery bypass grafting (CABG) is defined as troponin concentrations >10 times 99th percentile upper reference limit (URL) according to the Fourth Universal Definition. However, troponin concentrations after non-CABG cardiac surgery which indicate greater-than-expected myocardial injury and increased risk for complications remain unclear. Our goal was to assess procedure-specific relationships between troponin T and a composite outcome of low cardiac output syndrome and in-hospital mortality in cardiac surgical patients.

METHODS

Patients having cardiac surgery between January 2010 and December 2017 were categorized into 4 groups by procedure: (1) CABG; (2) mitral valve repair; (3) aortic valve repair/replacement (AVR); (4) mitral valve replacement (MVR) or CABG + valve surgeries. Exclusion criteria were elevated preoperative troponin T, preoperative kidney failure, circulatory arrest, or preoperative/planned mechanical circulatory support. Logistic regression was used to assess the association between troponin T and composite outcome, both overall and by procedure, including assessment of the interaction between procedure and troponin T on outcome.

RESULTS

Among 10,253 patients, 37 (0.4%) died and 393 (3.8%) developed the primary outcome. Troponin T concentrations differed by procedure (P < .001). Compared to CABG, AVR had 0.53 (99.2% confidence interval [CI], 0.50-0.56; unadjusted P < .001) times lower troponin T concentrations, while MVR/CABG + valve were 1.54 (99.2% CI, 1.45-1.62, unadjusted P < .001) times higher. There were linear relationships between log2 troponin T concentration and log odds mortality/low cardiac output syndrome. The (unadjusted) relationships were parallel for various types of surgery (interaction P = .59), but at different levels of the outcome.

CONCLUSIONS

The relative increase in odds for mortality/low cardiac output syndrome per a similar increase in troponin T concentrations did not differ among cardiac surgical procedures, but the absolute troponin T concentrations did. Troponin concentrations should thus be interpreted in context of surgical procedure.

Dynamic observation of 5-fluorouracil-induced myocardial injury and mitochondrial autophagy in aging rats

Patients treated with 5-fluorouracil (5-FU) can develop rare but potentially severe cardiac effects, including cardiomyopathy, angina pectoris, heart failure and cardiogenic shock. The specific pathologies and underlying mechanisms are yet to be fully understood. The results of previous studies have indicated that mitochondrial autophagy is widely detected in many angiocardiopathies. In the present study, the dynamic changes in the homeostasis of mitochondrial injury and autophagy were observed in rats treated with 5-FU for different durations. A corresponding control group and a 5-FU model group were established in groups of Sprague-Dawley rats aged 2 and 18 months, and the myocardial enzyme levels were determined at different time points. At 2 weeks post-model establishment, cardiac ultrasound and myocardial histological staining were performed, cardiomyocyte apoptosis and myocardial mitochondrial function were assessed, and mitochondrial ultrastructure was examined. In addition, the expression levels of autophagy-related proteins were evaluated in the 18-month-old rats on days 7 and 14 of 5-FU administration. The experimental results demonstrated that 5-FU induced an elevation in the levels of myocardial enzymes, as well as changes in the cardiac structure and function, and that these changes were more prominent over longer drug durations. In addition, 5-FU decreased the levels of myocardial mitochondrial ATP and mitochondrial membrane potential, and aggravated myocardial fibrosis and cardiomyocyte apoptosis compared with those observed in the untreated control group, treated with the same volume of saline as 5-FU in the 5-FU group. These injuries were particularly evident in aging rats. Notably, 5-FU increased the expression levels of myocardial mitochondrial autophagy-related proteins, and electron microscopy revealed a more severe autophagic state in the model groups compared with that in the control groups. In conclusion, 5-FU induced myocardial mitochondrial damage, the degree of which was more severe in aging rats compared with that in young rats. The mitochondrial autophagy induced by 5-FU was excessive, and the degree of autophagy was aggravated with increased 5-FU administration time.

The Future of Cardioprotection-Pointing Toward Patients at Elevated Risk as the Target Populations

Translation of the cardioprotective effect by pharmacological and mechanical conditioning therapies into improvement of clinical outcome for the patients has been disappointing. Confounding factors like comorbidity and comedications may explain some of the loss in translation. However, the substantial improvement of outcome in disease states involving ischemia-reperfusion injury, that is, planned cardiac surgery, elective percutaneous coronary intervention, and even primary percutaneous coronary intervention for ST-segment myocardial infarction (STEMI), is the most plausible explanation for the missed demonstration of a clinical benefit. Remote ischemic conditioning has demonstrated consistent cardioprotective effect in experimental and in clinical proof-of-concept studies. As an adjunctive cardioprotective treatment beyond reperfusion, remote ischemic conditioning should address target populations at risk of extensive tissue damage, including patients who experience complications, which may induce profound myocardial ischemia in relation to cardiac surgery or elective percutaneous coronary intervention. Moreover, patients with STEMI and predictable impaired clinical outcome due to delayed hospital admission, high Killip class, cardiogenic shock, and cardiac arrest remain target groups. For high-risk patients, daily remote ischemic conditioning or the corollary of blood flow-restricted exercise may be alternative cardioprotective options during postoperative and post-myocardial infarct rehabilitation.

Prognostic value of high-sensitivity troponin I after cardiac surgery according to preoperative renal function

Cardiac troponin levels can be elevated without myocardial injury in patients with renal impairment. However, the prognostic value of elevated troponin levels after cardiac surgery has not been well evaluated in patients with renal impairment. We evaluated the relationship between postoperative troponin levels and mortality following cardiac surgery according to preoperative renal function.Among 3661 patients underwent cardiac surgery between March 2005 and December 2015, 1909 patients were analyzed after excluding those with insufficient laboratory data, preoperative myocardial infarction, underwent Cox-Maze or redo surgery, or with a follow-up period <30 days. The primary outcome was risk of 30-day mortality according to elevated postoperative high-sensitivity cardiac troponin I (hs-cTnI) levels in varying degrees of renal function. Secondary outcomes included long-term cardiac-cause and all-cause mortality during the median follow-up of 52 months.After adjustment for risk factors, elevated peak postoperative hs-cTnI was associated with 30-day mortality [adjusted odds ratio 1.028, 95% confidence interval (CI) 1.013-1.043, P < .001], long-term cardiac-cause [adjusted hazard ratio (HR) 1.013, 95% CI 1.009-1.017, P < .001] and all-cause mortality (adjusted HR 1.013, 95% CI 1.009-1.016, P < .001), in patients with preoperative normal renal function [estimated glomerular filtration rate (eGFR) ≥60 ml/minute/1.73 m]. However, in patients with renal impairment (eGFR < 60 ml/minute/1.73 m), hs-cTnI levels were not associated with mortality following cardiac surgery.Elevated hs-cTnI levels following cardiac surgery did not predict short- and long-term mortality in patients with preoperative renal impairment.

Heparin-binding protein measurement improves the prediction of myocardial injury-related cardiogenic shock

Background

Heparin-binding protein (HBP), a potent inducer of increased vascular permeability, is a potentially useful biomarker for predicting outcomes in patients with postoperative myocardial injury-related cardiogenic shock (MIRCS). We aimed to evaluate and validate HBP as a prognostic biomarker for postoperative MIRCS.

Methods

We performed a case-control study in 792 patients undergoing cardiac surgery from January 1, 2016, to August 1, 2019, including 172 patients with postoperative MIRCS and 620 age- and sex-matched controls. The association between HBP and MIRCS was determined by multivariate logistic regression analysis. Receiver operating characteristic curves (ROCs) with area under the curve (AUC) were performed to calculate the cut-off value, sensitivity and specificity. The association between HBP and cardiac troponin T (cTnT) was determined by multivariable linear regression analysis. Blood samples were drawn from the coronary sinus and arterial line of the cardiopulmonary bypass (CPB) before aortic cross-clamping (time point 1) and 5 min after aortic declamping (time point 2).

Results

Before aortic cross-clamping, coronary sinus HBP (HBP_CS1) showed no differences between the two groups. However, after declamping, the MIRCS group had a significantly higher sinus HBP level (HBP_CS2) than did the control group. HBP_CS2 predicted MIRCS with an AUC of 0.85 (95% CI: 0.81–0.89, cut-off: 220 ng/ml, sensitivity: 92% and specificity: 70%). After adjusting for confounding factors, we found that HBP was an independent risk factor for MIRCS (OR: 7.65, 95% CI: 4.86–12.06, P < 0.01) and was positively associated with cTnT (β > 0, P < 0.01).

Conclusions

Elevated levels of coronary sinus HBP were useful biomarkers for predicting MIRCS after cardiac surgery.

Cardiac Myosin-Binding Protein C Release Profile After Cardiac Surgery in Intensive Care Unit

BACKGROUND

Cardiac surgical procedures produce iatrogenic myocardial cell injury with necrosis that result in an obligatory release of biomarkers. Cardiac myosin binding protein C (cMyBP-C) has recently emerged as a specific and sensitive biomarker in patients with acute myocardial injury. We therefore aimed to investigate the release profiles of cMyBP-C after cardiac surgical procedures.

METHODS

Enzyme-linked immunosorbent assay to detect blood cMyBP-C was established by using two monoclonal antibodies against N-terminus of human cMyBP-C. Consecutive patients undergoing cardiac operations (N = 151) were recruited in this study. Blood cMyBP-C was assayed preoperatively, at intensive care unit arrival (0 hour after the operation), at 2 to 48 hours, and before discharge. The characteristics and detailed surgical procedure were recorded.

RESULTS

The established immunoassay was capable of detecting human cMyBP-C (0 to 1000 ng/L). The released cMyBP-C peaked immediately after cardiac surgery (0 h), attaining 3.8-fold higher than before the operation, dropped abruptly within 24 hours, and stayed at a higher level until discharge. Postoperative cMyBP-C levels correlated positively with high-sensitivity cardiac troponin T (hs-cTnT), creatine kinase, myoglobin, and creatine kinase MB isoenzyme. Different cardiac surgical procedures were characterized by different levels of release of cardiac biomarkers. Isolated off-pump coronary artery bypass grafting was associated with the smaller amount of cMyBP-C release, whereas valve replacement/plasty surgery produced higher release, in particular the multiple-valve surgery. Both cMyBP-C and hs-cTnT correlated with surgical techniques, postoperative intensive care unit stay, and hospital stay.

CONCLUSIONS

Circulating cMyBP-C is a promising novel biomarker for evaluating cardiac surgical trauma in patients undergoing a cardiac operation.

Urinary Titin Is Increased in Patients After Cardiac Surgery

Background: Few non-invasive biomarkers have been used to detect myocardial injury in patients with heart diseases. Recently, the N-terminal fragment (N-titin) of titin, a giant sarcomeric protein, which is involved in muscular passive tension and viscoelasticity, has been reported to detect muscle damage in patients with cardiomyopathy as well as in patients with skeletal muscle dystrophy and in healthy volunteers with endurance exercise. In the present study, we evaluated whether urinary N-titin is changed during a perioperative period and whether its increase reflects myocardial damage.

Materials and Methods: In 18 patients who underwent cardiac surgery, blood and urine samples were obtained before and after surgery. We measured the urinary levels of N-titin with a highly sensitive ELISA system.

Results: Urinary N-titin to creatinine (N-titin/Cr) was significantly increased in all patients postoperatively (43.3 ± 39.5 pmol/mg/dL on the day of operation) and remained significantly high for at least 4 days postoperatively. Urinary N-titin/Cr was positively correlated with serum cardiac troponin T (r = 0.36, p = 0.0006, n = 90) but not creatine kinase-MB (CK-MB). We also found that urinary N-titin/Cr in patients after a coronary artery bypass grafting operation was higher by day 2 postoperatively than in patients following open cardiac surgeries.

Conclusion: The cleaved N-titin was significantly increased in urine after cardiac surgery. Urinary N-titin may be useful for detecting the risk of latent postoperative cardiac damage.

Albumin, a marker for post-operative myocardial damage in cardiac surgery

PURPOSE

Low serum albumin (SA) is a prognostic factor for poor outcome after cardiac surgery. The aim of this study was to estimate the association between pre-operative SA, early post-operative SA and postoperative myocardial injury.

METHODS

This single center cohort study included adult patients undergoing cardiac surgery during 4 consecutive years. Postoperative myocardial damage was defined by calculating the area under the curve (AUC) of troponin (Tn) values during the first 72 h after surgery and its association with SA analyzed using linear regression and with multivariable linear regression to account for patient related and procedural confounders. The association between SA and the secondary outcomes (peri-operative myocardial infarction [PMI], requiring ventilation >24 h, rhythm disturbances, 30-day mortality) was studied using (multivariable) log binomial regression analysis.

RESULTS

In total 2757 patients were included. The mean pre-operative SA was 29 ± 13 g/l and the mean post-operative SA was 26 ± 6 g/l. Post-operative SA levels (on average 26 min after surgery) were inversely associated with postoperative myocardial damage in both univariable analysis (regression coefficient - 0.019, 95%CI -0.022/-0.015, p < 0.005) and after adjustment for patient related and surgical confounders (regression coefficient - 0.014 [95% CI -0.020/-0.008], p < 0.0005).

CONCLUSIONS

Post-operative albumin levels were significantly correlated with the amount of postoperative myocardial damage in patients undergoing cardiac surgery independent of typical confounders.

Changes in Circulating Monocyte Subsets (CD16 Expression) and Neutrophil-to-Lymphocyte Ratio Observed in Patients Undergoing Cardiac Surgery

BACKGROUND

The characteristics of circulating inflammatory cells (leukocytes) in patients undergoing heart surgery remains poorly understood. Recently, neutrophil-to-lymphocyte ratio (NLR) and specific monocyte subsets (based on CD14/CD16 expression) have been suggested as markers of inflammation and predictors of outcomes. The present study aims to characterize the influence cardiac surgery with cardiopulmonary bypass has on specific circulating leukocytes.

METHODS

All enrolled patients had blood samples taken pre- (0 days), early post- (5 days), and late post- (90 days) surgery. Complete blood counts were performed and whole leukocyte isolations were obtained from blood samples and analyzed with flow cytometry. Fluorophore-linked antibodies (CD45, CD11b, CD14, and CD16) were added to the blood cell isolations and later assessed by flow cytometry.

RESULTS

Seventeen patients were enrolled and samples obtained at 0, 5, and 90 days. We demonstrated a significant increase in NLR (2.2-fold; p = 0.0028) and CD16 mean fluorescence index (MFI-measure fluorescence intensity shift of CD16 in a gated cell population) early at day 5 (2.0-fold; p = 0.0051). Both NLR and CD16 MFI levels generally returned to normal by day 90. There was a significant positive correlation between NLR and CD16 MFI (r² = 0.29; p = 0.0064). Adverse cardiovascular event (AE) was defined as prolonged length of hospitalization or readmission to hospital for cardiac reasons after discharge was seen in 59% of patients (no deaths occurred). In an unadjusted analysis of AE, we identified NLR as a likely predictor of AE, which meant that patients developing AE had a significantly higher baseline NLR (p = 0.0065), something that was not observed with CD16 MFI (p = 0.2541).

CONCLUSION

Cardiac surgery is associated with a significant increase in NLR and CD16 MFI (non-classical monocytes) early after surgery corresponding to the early inflammatory phase after surgery. Furthermore, we have, for the first time, identified a significant correlation between NLR and CD16 MFI. While the mechanism for this relationship remains unclear, our findings support the use of a simple test of NLR as a biomarker of inflammation for predicting outcomes in cardiac surgery patients.