Discussion
The principal finding of this study was that a higher level of preoperative peak troponin was associated with increased ICU time and hospital length of stay. However, there was no association with peak troponin level and short- and long-term MACCE or mortality. In sub-analysis, peak troponin >10.00 ng/mL was also not associated with increased hazards for MACCE or mortality.
Previous studies have demonstrated associations with serum troponin elevation with size of myocardial infarction, degree of coronary artery disease, and/or short term morbidity and mortality in patients presenting with acute coronary syndromes5,8,9. Furthermore, preoperative troponin elevation has been associated with increased operative morbidity and mortality, as well as increased rates of complications and/or hospital length of stay following CABG11–15. In our study, patients with high and low peak troponin levels were fairly well-matched with respect to age and preoperative comorbidity. However, we noticed that the high serum troponin group had a lower median preoperative left ventricular ejection fraction. Additionally, they required bridging to CABG with intra-aortic balloon pump counterpulsation more frequently, and also had a higher incidence of preoperative congestive heart failure with advanced New York Heart Association class symptoms (III or IV) in comparison to the low troponin group. Based on these findings, it is possible that higher peak troponin may indicate a larger infarction with higher degree of myocardial stunning, which may account for lower preoperative ejection fractions, higher incidence of heart failure, and need for mechanical support. However, it may be possible that patients with history of preexisting heart failure with decreased ejection fraction are more prone to higher serum troponin leakage after NSTEMI, representing a combination of infarction as well as heart failure. In this theory, these patients with worse baseline cardiac function may require intra-aortic balloon pump bridging more frequently after an equivalent cardiac insult as would a patient with normal baseline function and greater cardiopulmonary reserve.
The timing of CABG following acute coronary syndrome has remained controversial16–19. Patients presenting with STEMI or ongoing ischemia in NSTEMI require immediate revascularization. Often times these patients can present with arrhythmias, hemodynamic instability, and end-organ dysfunction. As a result, emergent immediate revascularization in these scenarios is often met with a high likelihood of morbidity and mortality, especially between 6 and 24 hours of symptom onset16,19. In the setting of NSTEMI or smaller infarctions with no ongoing active ischemia where cardiac function is preserved or mildly depressed, surgical timing can vary between centers. Some groups advocate for prompt revascularization20,21, whereas other centers advocate for a delay prior to surgery22. This delay can allow for the stable patient to undergo preoperative testing and risk-assessment, and also allow any high-dose antiplatelets and/or anticoagulants the patient may have received at time of initial presentation to be metabolized, which may allow for decreased bleeding complications. Other advantages of delaying surgery in some surgeons’ opinion is that tissues and coronary vessels are less edematous and easier to handle, and that there is less myocardial stunning which allows the heart to tolerate the operation better. In our series of NSTEMI patients, we did not observe increasing time from peak troponin level to surgical revascularization to have any significant influence on mortality in multivariable modeling. It is likely that timing of surgery in this patient population can be tailored to the individual circumstances and needs of the patient, and that revascularization does not necessarily need to occur within a specific “surgical window”. There is always a theoretical risk that delaying surgery too long can lead to lethal arrhythmias, cardiac arrest, or worsening cardiac function due to expanding infarction.
In our series, we observed a small but significant increase in intensive care unit time (47.0 hours [IQR 26.0 to 81.6] vs 43.0 hours [IQR 24.6 to 69.0], P=0.004) and hospital length of stay in patients with higher peak troponin levels (10 days [IQR 8 to 13] vs 9 days [IQR 8 to 12], P=0.013). However, in our sub-analysis, with patient cohorts stratified by a troponin level of 10.00 ng/mL, we did not observe differences in intensive care unit (47.1 hours [IQR 25.3 to 85.5] vs 45.0 hours [IQR 25.0 to 72.0], P=0.217) or hospital length of stay (10 days [IQR 8 to 13 vs 10.0 hours [IQR 8 to 12], P=0.124) (Supplemental Table 3 ). Though there is a possibility for increased intensive care and/or hospital times with increased peak troponin, the absolute difference is small and the clinical and logistical relevance are likely limited.
Lastly, our findings support the notion that preoperative peak troponin levels have little predictive value on long-term MACCE or mortality following surgical revascularization. These findings are corroborated by those of Beller and colleagues. In their analysis of 1,272 urgent or emergent CABG procedures, they discovered that presence of preoperative troponin elevation to be associated with higher risk of morbidity, 1-year, and long-term mortality, when compared to no troponin elevation. However, when evaluating the actual troponin levels, increasing levels did not have association with postoperative morbidity or survival following CABG6. Our study did not observe a significant difference in mortality when peak troponin levels were above the median value across patients (> 1.95 ng/mL). Because this median value seemed to be low from a clinical perspective, we elected to perform a post hoc secondary analysis, and evaluate a clinically meaningful cutoff of 10.00 ng/dL. Despite this, high and low troponin cohorts did not have any significant differences in 5-year MACCE or mortality. Similarly, these findings held true when troponin was modeled as a continuous variable. Furthermore, peak troponin >10.00 ng/ML was not associated with increased hazards for MACCE or mortality in our multivariable models. Collectively, these data suggest that the strategy of basing timing of CABG in NSTEMI on normalization of troponin or having the troponin levels downtrend to a certain level does not appear to be warranted.