Author + information
- Joseph B. Selvanayagam, MBBS, FRACP,
- David Pigott, MD,
- Lognathen Balacumaraswami, FRCS,
- Steffen E. Petersen, MD,
- Stefan Neubauer, MD, FRCP and
- David P. Taggart, MD, PhD* ()
- ↵*Department of Cardiothoracic Surgery, John Radcliffe Hospital, Oxford, OX3 9DU, United Kingdom
To the Editor:Elevation of cardiac biochemical markers after coronary artery bypass grafting (CABG) is relatively common, and large rises are prognostically significant (1,2). Current European Society of Cardiology/American College of Cardiology guidelines, however, do not include a specific troponin-based criterion for the diagnosis of myocardial infarction (MI) occurring after CABG (2), and uncertainties remain regarding the functional significance of small elevations in biochemical markers of post-CABG.
Delayed enhancement magnetic resonance imaging (DE-MRI) can quantify irreversible myocardial injury, and is superior to single-photon emission computed tomography imaging for the identification of subendocardial MI (3). We have recently reported that, in patients randomized to either on-pump coronary artery bypass grafting (ONCABG) or off-pump coronary artery bypass grafting (OPCABG) techniques, there was no significant difference between the groups in extent of surgery-related irreversible myocardial injury when assessed by DE-MRI (4). Taking advantage of this group of CABG patients, we set out to determine the optimal cardiac biochemical parameter that would have the highest sensitivity and specificity for the detection of perioperative irreversible myocardial injury as determined by DE-MRI. A secondary goal was to see whether large rises in postoperative biochemical markers will impact on late functional improvement as assessed by cine MRI.
The study was approved by our institutional ethics committee. Both the patient population characteristics and cardiovascular MRI protocol have been described elsewhere (4). A total of 60 patients were enrolled into the study—30 randomized to off-pump surgery and 30 to on-pump surgery. Cardiac troponin I (cTnI)/creatine kinase-MB (CK-MB) were both measured preoperatively, and at 1, 6, 12, 24, 48, and 120 h postoperatively. The upper limit of normal as defined by our laboratory was 1.0 μg/l for cTnI and 3.7 ng/ml for CK-MB.
Continuous variables that were not distributed normally were compared using the Mann-Whitney Utest, and correlation between such variables was made using the Spearman Rank test. A p value of <0.05 was considered statistically significant. Biochemical data was divided into five categories for troponin I and four categories for CK-MB. Receiver-operator characteristic (ROC) analyses were used in each biochemical category to determine the diagnostic performance of each biochemical parameter for detecting new irreversible injury.
Both DE-MRI and biochemical results were available for 53 of 60 patients. As previously reported, the median (interquartile range) area under the curve values for troponin I release were significantly higher in the ONCABG group (182.0 [226.3] μg/l) compared with the OPCABG group (135.0 [125.0] μg/l; p = 0.02) (4). The median troponin at 48 h was 1.3 (2.3) μg/l in OPCABG patients and 1.5 (2.5) μg/l in ONCABG patients (p > 0.05). Furthermore, the mean time at which the highest troponin I was recorded was significantly longer in the group in which new irreversible injury was documented by DE-MRI (20.3 ± 11.9 h vs. 11.3 ± 11.0 h; p = 0.04).
Similar to the findings for troponin I, the median cumulative CK-MB measurement in the first five days after surgery was significantly higher in the ONCABG group: 596 (680) ng/ml versus 462 (429) ng/ml (p = 0.009). Not surprisingly, peak CK-MB was significantly higher in the group with new hyperenhancement (11.3 [9.0] ng/ml vs. 18.3 [35.0] ng/ml; p = 0.02). The cumulative troponin I rise early (0 to 24 h) after surgery was moderately correlated to the size of new irreversible injury, although not as strongly as a late (24 to 120 h) elevation in troponin I (r2= 0.32 for early vs. r2= 0.47 for late; p < 0.05). Furthermore, when considering all biochemical parameters, a single troponin I measurement at 48 h after surgery was most strongly related to new myocardial hyperenhancement (r2= 0.54; p = 0.009).
Receiver-operating characteristic (ROC) analysis showed the best univariate predictive values for troponin I and CK-MB parameters to detect new irreversible injury as reflected by larger ROC areas (Table 1).The optimal cutoff point (sensitivity/specificity) to differentiate the presence of new postprocedure irreversible injury was 81.5 μg/l (76%/62%) for troponin release 24 to 120 h, 1.15 ug/l (82%/52%) for troponin I value at 48 h, and 9.2 ng/ml (72%/60%) for peak CK-MB. Combining both a late peak (>18 h) and an absolute value of >1.15 μg/l in troponin I at 48 h resulted in sensitivity 79%, specificity 81%, and accuracy 80% for the detection of new irreversible injury.
We correlated the effect on late left ventricular function according to highest (“peak”) CK-MB or troponin I recorded in the first five days. Overall, there was no correlation between peak CK-MB/troponin I and change in global function at six months (delta ejection fraction [EF]) (Fig. 1A).However, the change in EF at six months after surgery was significantly different between patients stratified according to highest CK-MB/troponin I recorded, as either > or <5× the upper limit of normal (Fig. 1B).
Our findings show that when using new myocardial hyperenhancement on DE-MRI as a reference standard, the combination of a late peak in troponin I (>18 h) and an absolute troponin I value of 1.15 μg/l at 48 h had the best predictive accuracy for perioperative irreversible myocardial injury. On a practical, clinical level, this would mean sampling serum at three to four time points postoperatively in the first two days after surgery (for example, at 6, 12, 24, and 48 h), to detect new postprocedure MI. Our findings have important implications for the definition of CABG-related MI, and for understanding the relationship between biochemical release and irreversible myocardial injury in the setting of bypass surgery.
There was a stronger relationship between cumulative troponin I release in the postoperative period of 24 to 120 h and new irreversible injury, when compared with the relationship for troponin I release in the first 24 h. There are a number of potential factors that could explain these findings. Work by Remmpis et al. (5) and others has shown that protein release from nonstructurally bound cytosolic pools accounts for up to 6% of the total cTnI content of the cardiomyocyte. “Washout” of this cytosolic component of troponin (reflecting surgery-induced increased cell permeability without cellular death) could result in one seeing “troponin blips” in the first few hours after surgery without corresponding irreversible myocardial injury on imaging. In addition, in some ONCABG patients, there is a possibility that inadequate cardioplegia might have resulted in global intermittent hypoperfusion, which, in turn, might have caused a global, mainly subendocardial ischemic injury below the sensitivity of DE-MRI.
Overall, we found no direct relationship between the degree of CK-MB/troponin rise postoperatively and the late improvement in left ventricular function. This is not surprising, however, given the multifactorial etiology of postoperative biochemical elevation, compounding the effect of revascularization on global left ventricular function (likely improvement of hibernating segments), and the small volume of new irreversible injury noted in our patients (average of 2% of left ventricular mass). Notwithstanding this, in the small number of patients who did demonstrate large rises in CK-MB or troponin I (>5× upper limit of normal), there was an inverse linear relationship between the rise in biochemical parameters and lack of improvement in global left ventricular function. This observation provides one potential explanation for the effect on late mortality seen in patients demonstrating such large biochemical marker elevation (1).
We conclude that combining both a late peak in cTnI and an absolute value >1.15 μg/l at 48 h had the best accuracy for the detection of post-CABG MI, and that large postoperative rises in CK-MB or cTnI adversely predict late ventricular function.
Please note: this work was supported by the British Heart Foundation and the Medical Research Council, United Kingdom. The Guidant and Medtronic companies provided funding for the analysis of the biochemical specimens. Dr. Selvanayagam is funded by the Wellcome Trust.
- American College of Cardiology Foundation
- Costa M.A.,
- Carere R.G.,
- Lichtenstein S.V.,
- et al.
- French J.K.,
- White H.D.
- Selvanayagam J.B.,
- Petersen S.E.,
- Francis J.M.,
- et al.