Research Article, Int J Cardiovas Res Vol: 5 Issue: 3
The Ability of Elevated Troponin I Values following Cardiac Surgery to Predict 30-Day and Long Term MajorAdverse Cardiovascular Events: A Single Center Experience
| Lee JS1, Lochner K1, II Evans MO1, Olsen CH3, Larson C1, Rubal B1, Conner WC2, Slim AM1 and Thomas DM1* | |
| 1Cardiology Service, San Antonio Military Medical Center, 3551 Roger Brooke Drive, San Antonio, Texas 78234-6200, USA | |
| 2Cardiovascular Surgery Service, San Antonio Military Medical Center, 3551 Roger Brooke Drive, San Antonio, Texas 78234-6200, USA | |
| 3Department of Preventive Medicine and Biometrics, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814, USA | |
| Corresponding author : Thomas DM, MD Director, Cardiology Service,San Antonio Military Medical Center, 3551 Roger Brooke Drive,San Antonio, TX 78234-6200 Tel:210) 916-6407 Fax:(210) 916-3051 E-mail:dustin.m.thomas4.mil@mail.mil |
|
| Received: March 26, 2016 Accepted: May 26, 2016 Published: May 31, 2016 | |
| Citation: Lee JS, Lochner K, II Evans MO, Olsen CH, Larson C, et al. (2016) The Ability of Elevated Troponin I Values following Cardiac Surgery to Predict 30-Day and Long Term Major Adverse Cardiovascular Events: A Single Center Experience. Int J Cardiovasc Res 5:4.doi:10.4172/2324-8602.1000272 |
Abstract
Background: The association between elevated troponin levels after coronary artery bypass grafting (CABG) surgery and short term all-cause mortality as well as major adverse cardiovascular events (MACE) is well documented. However, there is paucity of data reported on long term outcomes. Methods: Baseline patient characteristics, perioperative characteristics, and cardiac biomarkers within 24 hours were evaluated prospectively in a consecutive cohort of 419 subjects who underwent cardiac surgery in a single center, from 1 July 2002 through 30March 2007. Death, myocardial infarction, stroke, target vessel revascularization, and composite MACE were abstracted over a median follow-up of 16.8 months. Results: Troponin I values above 4.13 ng/mL were associated with a 45% increased risk for composite MACE (HR 1.45, 95%CI 1.023-2.067, p=0.038) over the first 2 years post cardiac surgery. Additionally, troponin I levels >4.13 n/mL were also associated with worse survival free from MACE (p=0.034). Troponin I values were not predictive of death, myocardial infarction, stroke or target vessel revascularization nor associated with survival free from these events. Multiple pre-, intra-, and periprocedural patient characteristics were associated with increased observed MACE rates to include preoperative ejection fraction <40%, chronic kidney disease, reoperation, and increased cardiopulmonary bypass times. Only EF ≥ 40% was associated with lower observed MACE. Conclusion: A troponin I level above 4.13 ng/mL was associated with a 45% observed increase in composite MACE during the first 2 years following CS. This finding is hypothesis generating only and may suggest measurement of troponin I level 24hrs after the CS may provide valuable prognostic data.
Keywords: Cardiac biomarkers; Troponin; Cardiac surgery; Coronary bypass surgery; Cardiovascular outcomes
Keywords |
|
| Cardiac biomarkers; Troponin; Cardiac surgery; Coronary bypass surgery; Cardiovascular outcomes | |
Abbreviations and Acronyms |
|
| CABG: Coronary Artery Bypass Grafting; CHF: Congestive Heart Failure; CKD: Chronic Kidney Disease; CK: Creatinine Kinase; CKMB: Creatinine Kinase-MB; CMR: Cardiac Magnetic Resonance; CS: Cardiac Surgery; DM: Diabetes Mellitus; DoD: Department of Defense; EMR: Electronic Medical Records; HTN: Hypertension; IABP: Intraaortic Balloon Pump; LBBB: Left Bundle Branch Block; LOS: Length of Stay; LVEF: Left Ventricular Ejection Fraction; MACE: Major Adverse Cardiovascular Events; MI: Myocardial Infarction; PCI: Percutaneous Coronary Intervention; SSDI: Social Security Death Index; STS Score: Society of Thoracic Surgeons Score. | |
Introduction |
|
| Cardiac biomarker elevation is used as part of the universal definition of perioperative myocardial infarction (Type 5 MI) after coronary artery bypass graft (CABG) surgery. Type 5 MI is defined as biomarkers elevation to ten times the upper limit of normal (10 x 99th percentile URL) in subjects with normal baseline values. Additionally, a new pathologic Q waves or new left bundle branch block (LBBB), new angiographic graft or native vessel occlusion, or evidence of new wall motion abnormality or presence of non-viable myocardium must be present [1]. Type 5 MI was found in multiple studies to be associated with adverse outcomes [2-4]. In patients who undergo CABG in the setting of an acute coronary syndrome (ACS) with elevated cardiac biomarkers, the timing of the troponin peak is useful in distinguishing Type 5 MI from the expected trend of the troponin curve [5,6]. The extent of biomarker elevation that is associated with adverse outcomes also varies significantly between studies [7,8]. | |
| We followed a consecutive cohort of subjects who underwent cardiac surgery (CS) to determine troponin I levels associated with adverse outcomes, both in the short term and in long term follow-up. We additionally sought to identify other independent predictors for adverse outcomes that may serve as useful prognostic markers. | |
Methods |
|
| Patient selection | |
| We prospectively followed a cohort of 419 adult subjects (>18 years of age) who consecutively underwent cardiac surgery (CS), defined as on-pump and off-pump coronary artery bypass grafting (CABG) with or without concomitant valve replacement or repair, isolated mitral or aortic valve replacement, and ascending aortic replacement from 1 July 2002 through 30 March 2007 at a single center, tertiary referral hospital (San Antonio Military Medical Center, Joint Base San Antonio-Fort Sam Houston, Texas). Subjects were excluded if they underwent congenital repair, atrial and ventricular septal defect repair, or tumor resection. Institutional review board approval for outcome data abstraction and data presentation was obtained locally. | |
| Patient demographics and clinical data were abstracted manually utilizing the Department of Defense (DoD) electronic medical records (EMR). Surgical and post-operative details were ascertained using both the local cardiothoracic surgical database as well as EMR to abstract the type of cardiac surgery, operative time, time on cardiopulmonary bypass (when applicable), aortic cross clamp time (when applicable), use of intra-aortic balloon pump (IABP), hospital length of stay (LOS) as well as intensive care unit (ICU) LOS, duration of intubation postoperatively, and duration on intravenous vasoactive medications (defined as epinephrine, norepinephrine, phenylephrine, dopamine, dobutamine, or milrinone) post-operatively. | |
| Measurement of cardiac biomarker concentration | |
| CK-MB (CK-MB STAT, Roche Diagnostics, Indianapolis IN) and Troponin I (Troponin I, Ortho Clinical Diagnostics, Raritan, NJ) concentrations were measured routinely within 24 hours of postoperative day in all patients. Analyses were performed by the central practice-certified hospital laboratory. Serum cTnI concentration was measured on a VITROS ECi/ECiQ Immunodiagnostic System analyzer (Ortho Clinical Diagnostics, Raritan, NJ). Serum CK-MB concentration was analyzed by using a biotinylated monoclonal anti-CK-MN antibody and a monoclonal CK-MB-specific antibody labelled with a ruthenium complex on Roche Elecsys e 601 analyzer (Roche Diagnostics, Indianapolis IN) by electrochemluminescence immunoassay. The 99th percentile of upper limit of normal is 0.08ng/ ml for Troponin I and 5ng/ml for CK-MB. | |
| Clinical data end points | |
| Serum troponin I levels were obtained at 24hrs on all subjects as part of their routine postoperative care. A daily electrocardiogram (ECG) was obtained on all subjects postoperatively. ECG’s were reviewed and adjudicated by a cardiovascular expert for the presence of ischemia, defined as 0.5 mm ST segment depression in ≥ 2 contiguous leads and in the absence of voltage criteria for left ventricular hypertrophy (LVH). Perioperative myocardial infarction (Type V MI) was defined as troponin I value>10 times the 99th percentile upper range limit (URL) in addition to new pathologic Q waves or new left bundle branch block (LBBB), new angiographic graft or native vessel occlusion, evidence of new wall motion abnormality, or presence of non-viable myocardium. The DoD EMR as well as the social security death index (SSDI) were reviewed systemically for all end points that were not collected prospectively to include: need for reoperation within 30 days, myocardial infarction (MI), target vessel revascularization, cerebrovascular accident (CVA), MI greater than 30 days post-operative period, or hospitalization for congestive heart failure (CHF) for all subjects who underwent cardiothoracic surgery and mortality. The primary endpoint was composite major adverse cardiovascular events (MACE), defined as all-cause mortality (ICD- 9 code 798.1, 798.2, 798.9, and V12.53), non-fatal MI (410.0-410.9), CVA (433.0-2, 434.0-01, 434.10-11, and 435), hospital readmission within 30days, and target vessel revascularization (PCI: 92980, 92981, 92982, 92995, and 92996 or CABG: 33510-33514, 33516, and 33533- 33536). All endpoints were abstracted using a query of ICD-9 codes, manual EMR chart review, and the SSDI. All events identified by ICD-9 code were adjudicated manually. | |
| Statistical analysis | |
| Statistical analysis was performed using IBM SPSS version 19.0 (IBM, Armonk, New York). Continuous variables are presented as means ± standard deviation and medians with interquartile range, as appropriate. Categorical variables are presented as frequencies with percentages. Comparison of means was performed using oneway ANOVA with post-hoc Bonferroni correction. Comparison of medians was performed using a Mann-Whitney U test. Survival analysis was performed using a Kaplan Meier analysis and Wilcoxon test or log rank test as appropriate based on Schoenfeld residuals. Receiver operator characteristics curve was performed to determine troponin I cutoff values. In this study, the independent variables are troponin I level, ECG, age, time on vasoactive medications, hospital and ICU LOS, operative time, body mass index (BMI), diabetes mellitus (DM), hypertension (HTN), smoking, creatinine, ejection fraction, number of grafts, and aortic cross clamp time. The dependent variable is the major adverse cardiovascular events (MACE). Multiple linear regression analysis was performed to determine the magnitude of MACE variation attributable to specific independent variables. Cox regression analysis was used to determine the relationship between the independent variables and MACE. P-value<0.05 was considered statistically significant. | |
Results |
|
| A total of 419 consecutive patients underwent cardiac surgery (CS) over nearly a 5 year period and were followed for a median of 16.8 months (range 8 days to 9.4 years). During the total duration of follow up, total of 244 MACE events occurred. Troponin I values and median creatinine kinase-MB fragment (CK-MB) at 24 hours post-CS (Tables 1 and 2) were significantly increased in patients with MACE compared with patients without MACE. No difference in total creatinine kinase (CK) values was seen between two cohorts of patients. | |
 |
Table 1: Perioperative statistics in patients undergoing and following CS. Continuous variables are depicted as median values with ranges or interquartile range as annotated. †Valve replacement and ascending aortic replacement ONLY subjects were excluded. ‡ Operations were excluded in instances when cardiopulmonary bypass (CPB) and/or aortic cross clamping were not used. Major adverse cardiovascular events (MACE) defined as a composite of death, myocardial infarction (MI), stroke (CVA), target vessel revascularization, and readmission to the hospital within 30 days of CS. CABG=coronary artery bypass graft surgery; ACS=acute coronary syndrome; IABP=intra-aortic balloon pump |
|
Table 2: Baseline patient demographic data. Age and body mass index (BMI), in lbs/in2, are depicted as mean values ± a standard deviation. MACE defined as a composite of death, myocardial infarction (MI), stroke (CVA), target vessel revascularization, and readmission to the hospital within 30 days of CS. CHF=congestive heart failure. |
| Receiver operator characteristics (ROC) curve of troponin I with respect to predicting MACE is shown in Figure 1. Troponin I was observed to be a very weak, though not statistically significant predictor of MACE in isolation with an area under the curve (AUC) of 0.5681. When corrected for the procedural variables of time spent on a ventilator, time on vasopressors, CPB time, and aortic cross clamp time, the AUC decreased slightly to 0.5215 (p=0.391). | |
|
Figure 1: Receiver operator characteristics (ROC) curve of troponin I in predicting major adverse cardiovascular events (MACE) following cardiac surgery. |
| A troponin I cutoff of 4.13 ng/mL derived from the ROC curve with a sensitivity of 55.9% and specificity of 56% was used in a Cox regression model. This analysis demonstrated that troponin I values >4.13 ng/mL following cardiac surgery were associated with a 45% increase in observed MACE in the 2 years following cardiac surgery (HR 1.45, 95%CI 1.023-2.067, p=0.038) after adjusting for the same procedural variables described above. Uncorrected hazard ratios for the other periprocedural and periprocedural variables are depicted graphically in Figure 2. An ejection fraction ≥ 40% was the only variable associated with lower observed MACE (Table 3). Preprocedural variables associated with increased observed MACE rates include a history of CKD, DM2, patients undergoing a reoperation, and a history of CHF. Intraprocedural variables that were associated with increased observe MACE include patients undergoing a concomitant valve replacement in addition to CABG, cardiopulmonary bypass (CPB) times exceeding 154 minutes, increasing numbers of coronary anastomoses, and the need for intraaortic balloon pump (IABP) placement prior to or during cardiac surgery. The strongest co-variable associated with increased observed MACE was the development of a post-operative sternal wound infection (HR 2.45, 95%CI 1.40- 4.0, p=0.002). Other post-operative factors associated with increased observed MACE include time on vasoactive medications, time spent intubated following surgery, ICU LOS, and total hospital LOS. | |
|
Table 3: Uncorrected hazard ratios for preprocedural and periprocedural variables with respect to observed major cardiovascular events (MACE) using a Cox proportional hazard model. |
|
Figure 2: Forrest plot depicting hazard ratios (HR) derived from a cox proportional hazard model looking at preprocedural and periprocedural parameters and the association with observed major cardiovascular events (MACE). HTN=hypertension; DM=diabetes mellitus; HLD=hyperlipidemia; BMI=body mass index; CKD=chronic kidney disease; CHF=congestive heart failure; EF=ejection fraction; CABG=coronary artery bypass graft surgery; CS=cardiac surgery; ACS=acute coronary syndrome; IABP=intraaortic balloon pump; ICU=intensive care unit; LOS=length of stay. *=Hazard ratios for these risk factors varied with time, thus the reported HR correspond to the 1st 2 years of follow-up. |
| Kaplan-Meier survival analysis utilizing a Wilcoxon test with the patient population dichotomized into patients with troponin value >4.13 ng/mL and <4.13 ng/mL following cardiac surgery is shown in Figure 3. Survival free from composite MACE (Figure 3A) was improved in patients with troponin I values<4.13 ng/mL (p=0.034). The observed magnitude of this effect was greatest in the first 2 years following cardiac surgery with the curves starting to converge in the remote post-operative period followed by gradual decrease in their correlation over time. No difference in event-free survival (Figure 3B-E) was observed in death, stroke, myocardial infarction (excluding type V MI), and target vessel revascularization. | |
|
Figure 3: Kaplan-Meier survival curves utilizing Wilcoxon test demonstrating long-term event-free survival rates observed in patients with post-cardiac surgery troponin I values above and below 4.13 ng/mL with respect to composite MACE (A), death (B), stroke (C), myocardial infarction (D), and target vessel revascularization (E). ACE=major adverse cardiovascular events. |
| Table 2 summarizes the baseline demographic data of the total cohort. Patients with MACE during follow-up were significantly more likely to be older (67.3 ± 10.1 vs 60.9 ± 11.8, p<0.001 male (p=0.015), diabetic (p=0.011), have a history of CHF (p<0.001), CKD (p<0.001), and ejection fraction <40% (16.3% vs 3.4%, p<0.001). In contrast, patients in whom MACE was not observed were more likely to have a normal pre-operative EF ≥ 55% (p=0.004). The remaining baseline characteristics did not differ significantly between the groups. | |
| Surgical and postoperative details are summarized in Table 1. The majority of cohort had isolated CABG (n= 318, 75.9%), while 50 patients (11.9%) had combination valve replacement/repair and CABG. Off-pump CABG was performed in 58 subjects (13.8%), emergent CABG (<24 hours) was performed in 9 subjects (2.1%), and total of 107 subjects (25.5%) underwent CABG within 7 days of ACS. Patients in whom MACE was observed were more likely to have had a combination valve replacement/repair and CABG (p=0.036) and undergoing off-pump CABG (p=0.026). | |
| Longer operative times (393 min (range 140-981) vs 357 min (range 171-840), p<0.001), cardiopulmonary bypass (CPB) times (159 min (range 15-385) vs 146 min (range 1-344), p=0.018), and aortic cross clamp times (126 min (range 28-374) vs 117 min (range 30-247), p=0.049) were observed in the patients who experienced MACE. Post-operatively, longer median post-operative ventilation times (p<0.001), time on vasoactive medications (p<0.001), intensive care unit (ICU) length of stay (LOS) (p=0.001), and hospital LOS (p<0.001) were all longer in patients with subsequent MACE. The need for perioperative intra-aortic balloon pump (IABP) placement (p=0.005) and the incidence of post-operative sternal wound infection (p=0.034) was more common in patients with MACE. | |
| Troponin I values at 24 hours post-CS (Table 2) were significantly increased in patients with MACE (median: 4.78 ng/mL (IQR 2.41, 8.54) compared with patients without MACE (median: 3.62 ng/mL (IQR 2.03, 6.66, p=0.009). Median creatinine kinase-MB fragment (CK-MB) 24 hours following CS was also significantly higher in the patients with MACE (median: 13.4 mg/mL (IQR 8.2, 23.9) vs 11.3 mg/ mL (IQR 6.6, 19), p=0.018). No difference in total creatinine kinase (CK) values was seen between MACE compared with no MACE patients. Receiver operator characteristics (ROC) curve of troponin I with respect to predicting MACE is shown in Figure 1. Troponin I was observed to be a very weak, though not statistically significant predictor of MACE in isolation with an area under the curve (AUC) of 0.5681. When corrected for the procedural variables of time spent on a ventilator, time on vasopressors, CPB time, and aortic cross clamp time, the AUC decreased slightly to 0.5215 (p=0.391). | |
Discussion |
|
| Cardiac Troponins are regulatory proteins that control the interaction between actin and myosin. Cardiac troponin T (cTnT) and cardiac troponin I (cTnI) are two of three types of troponin that are used in clinical setting to determine myocardial necrosis. Although elevated troponin indicates myocardial necrosis in majority of time, pathophysiologic causes can be either due to ischemic cause and nonischemic causes [1]. In our study, we attempted to answer question of whether elevated troponin level within 24 hour perioperative time, regardless of its causes, has short and long term prognostic value. | |
| A troponin I level above 4.13 ng/mL was associated with a 45% observed increase in composite MACE during the first 2 years in our patient population following CS. Additionally, multiple preprocedural, intraprocedural and periprocedural factors were observed to increase the risk for composite MACE. Only EF ≥ 40% was associated with lower observed MACE following cardiac surgery. | |
| The correlation between elevated troponin level in subjects who underwent cardiac surgery and all-cause mortality has been shown in multiple studies [2,3,6,7,9,10]. Lurati Buse et al. [9] prospectively followed1559 subjects who underwent on-pump CABG in a singlecenter demonstrating that elevated troponin level (defined as cTnT with “threshold” of 0.8 ug/L) within 48 hours of post-operative period was associated with significantly elevated 12 months mortality or major adverse cardiac event (MACE) (HR 2.13, 95% CI 1.47 -3.15). Furthermore, multivariate analysis illustrated elevated troponin T was independent predictor of mortality and the composite end point of MACE at one year. | |
| All-cause mortality association with elevated troponin I level (ng/ ml) at 24 hour post-operative cardiac surgery was assessed by Croal et al in 1365 subjects who underwent a range of cardiac surgeries. In this study, elevated troponin I level was also associated with 30 day (OR 1.14 per 10ug/L, CI 1.05 to 1.24), 1 year (OR 1.10 per 10ug/L CI 1.03 to 1.18 ), and 3 year mortality (OR 1.07 per 10ug/L, CI 1.00 to 1.15). The sustained higher rates of all-cause mortality, whether short or long term, were observed in subjects with troponin I range of 8.49 to 350.81ug/L, revealing the proportional association between the intensity of troponin elevation and mortality. | |
| This negative prognostic value of post-operative troponin elevation can be partially explained by the association of troponin elevation with irreversible myocardial necrosis found in cine and delayed enhancement of cardiac magnetic resonance (CMR) [5]. This single-center randomized controlled trial had serially examined with CMR and cardiac biomarkers of forty subjects 24 hours after CABG. This study demonstrated that certain level of cardiac biomarker elevation at 24 hours post-surgery alone without supplemental evidence is sensitive for the detection of type V MI. Importantly; this study also confirmed the correlation of level of troponin elevation with mass of myocardial necrosis that was previously known [11]. | |
| Limitations | |
| Our cohort was limited by its small size; however, we feel the high rates of short and long term mortality and MACE, provide sufficient events from which to comment. While the event rate is high, this may be reflective of the ability to have complete patient follow-up in nearly 100% of patients undergoing CS in our single payer, closed system with expansive EMR coverage. The high MACE are not reflective of the surgeon skill or approach as all patients’ data points and outcomes were reported to the STS database without variance from the national average. This is a single center experience at a center with intermediate volume of CS, thus may not be applicable to large, high volume centers. In previous studies, EuroScore II has been used to assess long term perioperative mortality risk for cardiac surgeries. Although our study did not use EuroScore II, multivariate analysis was powered to show troponin was independent risk factor for MACE at 2 years. | |
Conclusion |
|
| A troponin I level above 4.13 ng/mL was associated with a 45% observed increase in composite MACE during the first 2 years following CS. This finding is hypothesis generating only and may suggest measurement of troponin I level 24hrs after the CS may provide valuable prognostic data. | |
References |
|
|
|
