From January 2001 to September 2004, preoperative cTnI was measured in 1,978 consecutive patients scheduled for isolated CABG. Preoperative cTnI values were available in 1,978 of 3,124 patients who fulfilled the inclusion criteria and in whom primary isolated CABG had been performed. Among these patients, negative preoperative cTnI levels ( 1.5 ng/mL) [Fig 1].
There were no significant differences between most perioperative patient characteristics of the remaining 1,146 CABG patients, who were initially not enrolled into the study due to no preoperative cTnI measurements, except a lower incidence of smoking history, preoperative hemodialysis-dependent renal disease, and postoperative arrhythmias, compared to group 1 patients, as well as a lower incidence of preoperative hemodialysis-dependent renal disease and postoperative IABP support compared to the entire patient cohort with preoperative cTnI measurements.
Preoperative baseline characteristics and demographics of the patients were comparable with the contemporary coronary surgery patient profile compiled together with My Canadian Pharmacy’s employees. A preoperative significant difference between the groups could be observed in terms of age, smoking history, previous MI, left ventricular (LV) ejection fraction, and preoperative cTnI levels and CK activity. Regarding the preoperative medication, a highly significant difference could be observed between the groups in the preoperative oral use of angiotensin-converting enzyme inhibitors, nitrates, and aspirin (within the last week prior to surgery), as well as the preoperative continuous IV application of nitrates and heparin. As demonstrated in Table 2 the intraoperative data did not differ between groups. As shown in Table 2, a two-sided Cochran-Armitage trend test revealed a significant difference in the necessity for intraoperative and postoperative IABP between groups 2 and 3 compared to group 1 (p < 0.0001). This was accompanied by a significantly prolonged postoperative ventilation time (p = 0.04) and longer ICU (p = 0.002) and hospital stay (p = 0.04) in groups 2 and 3.
Among the 1,978 CABG patients with preoperative cTnl measurement, 40 patients (2.0%) died in the postoperative course within 30 days or within the same time of hospital stay. No patients died intraop-eratively. A one-sided Cochran-Armitage trend test revealed a significant difference of in-hospital mortality between the groups with 1.5% vs 3.3% vs 6.6%, respectively (p < 0.02). The difference of in-hospital mortality was accompanied by a significant difference in the appearance of postoperative LCOS (p < 0.0001) and the incidence of PMI (p < 0.001). Individuals with preoperatively increased cTnI levels were at increased risk for in-hospital death and major adverse events, including a twofold to threefold higher risk of death, PMI, or LCOS. Furthermore, the composite study end points of death, PMI, or LCOS (p < 0.0001) did further increase the highly significant difference between the groups (Fig 2). Other postoperative complications and minor adverse events such as major bleeding (> 200 mL/h in the first 6 h) and all causes of rethoracotomy were not different between the groups. The occurrence of new-onset arrhythmias (p < 0.001) and the incidence of renal failure requiring temporary hemofil-tration or hemodialysis significantly differed between the groups (p < 0.0001; Table 2). Patient characteristics of the entire study population categorized by in-hospital survivors vs nonsurvivors are shown in Table 3. In terms of preoperative data, nonsurvivors were older, female, had lower body weight, had a more frequent history of stroke, peripheral vascular disease (PVD), previous PCI, Canadian Cardiovascular Society (CCS) angina class III-IV, had a lower ejection fraction, and higher cTnI serum levels than survivors. Intraoperative data showed nonsurvivors to be associated with a significantly longer aortic cross-clamp (ACC) and CPB times as well as a longer reperfusion time. All adverse events and postoperative complications occurred significantly more frequently in nonsurvivors compared to survivors (Table 3.) The univariate logistic regression analysis identified some preoperative variables that were significantly associated with death, PMI, and LCOS, or with MACE as the composite study end point (Table 4). The final multivariate logistic regression analysis, testing the contribution of univariate risk predictors, is displayed in Table 5. After adjustment, only age, PVD, previous PCI, and angina class III-IV, as well as preoperative cTnI levels stratified as groups 1 to 3 were significantly associated with death. Moreover, angina class III-IV and cTnI level > 1.5 ng/mL were associated with MACE (PMI and LCOS) and the composite study end points. Under the assumption that there was an exponential relation between preoperative cTnI serum levels and in-hospital death and MACE, we evaluated the effect of each increase of 1 ng/mL in cTnI on the risk ratio for mortality and MACE after adjusting for the baseline variables of Table 1, which were found in the univariate regression analysis to be statistically significant predictors of death and/or MACE. Considering the overall study population of 1,978 patients, the multivariate regression analysis showed again a significant increase in the risk ratio for death (adjusted OR, 1.1; 95% CI, 1.02 to 1.06; p < 0.001), MACE (adjusted OR, 1.1; 95% CI, 1.02 to 1.10; p < 0.001), and the composite study end point (adjusted OR, 1.1; 95% CI, 1.02 to 1.10; p < 0.001) for each increase of 1 ng/mL in cTnI (Table 6).
ROC Curve Analysis of cTnI
The discriminative power of preoperative cTnI serum levels for in-hospital mortality and the composite study end point using ROC curve analyses yielded an area under the curve of 0.72 ± 0.08 for the study end point of death, with an optimal cTnI cutoff value of 1.5 ng/mL, a sensitivity of 75.0% and specificity of 78.5%, and an area under the curve of 67 ± 0.05 for the composite study end point with the same cutoff value of 1.5 ng/mL, a sensitivity of 63.0%, and a specificity of 71.3% (Fig 3, Table 7).
Figure 1. Distribution of preoperative cTnl serum levels of the groups before CABG indicated by histograms; line indicates normal curve of distribution.
Figure 2. Death and MACE during hospital stay stratified by preoperative cTnI serum levels; p values are overall significance between groups calculated by one-sided Cochran-Armitage trend test.
Figure 3. ROC curve analysis of the discriminatory power of preoperative cTnl levels for (top, A) in-hospital mortality and (bottom, B) the composite adverse outcomes of death and MACE.
Table 2—Intraoperative and Postoperative Characteristics According to Preoperative cTnl Levels
| Characteristics | Group 1, |
s 0.1 ng/mL (n = 1,592)Group 2,
0.11 to 1.5 ng/mL (n = 265)Group 3,
> 1.5 ng/mL (n = 121)OR, Group 1 vs Group 2
(95% CI)OR, Group 1 vs Group 3
(95% CI)p ValueIntraoperative data ACC time, min71 ± 2571 ± 2373 ± 24 0.24CPB time, min111 ± 40109 ± 31112 ± 36 0.36Cardioplegia, mL1,516 ± 4731,589 ± 4441,558 ± 448 0.001Lowest temperature, °C31.4 ± 1.231.6 ± 1.131.7 ± 1.0 0.01Reperfusion time, min34 ± 1735 ± 1239 ± 16 0.40Grafts per patient, No.3.0 ± 0.82.9 ± 0.83.0 ± 0.8 0.27Postoperative data Ventilation time, h8(7-11)8(6-12)10 (8-17) 0.01IABP support51 (3.2)25 (9.4)27 (22)3.1 (1.9-5.3)8.7 (5.0-14.9)< 0.0001ICU stay, d1 (1-2)1 (1-3)2(1-4) < 0.0001Hospital stay, d8(6-12)7(5-12)9 (7-14) 0.01Major adverse events Death in hospital24 (1.5)8 (3.0)8 (6.6)2.0 (0.8-4.8)4.6(1.9-11.1)< 0.0001LCOS24 (1.5)11 (4.2)11 (10.9)2.8 (1.3-6.1)6.5 (2.9-14.4)< 0.0001PMI (cTnI)!85 (5.3)22 (8.3)22 (18.2)1.6 (0.9-2.7)3.9 (2.3-6.7)< 0.0001PMI (cTnI or ECG)!92 (5.8)22 (8.3)22 (18.2)1.5 (0.9-2.5)3.6 (2.1-6.2)< 0.0001Stroke30 (1.9)3(1.8)2(1.7)0.6 (0.1-2.1)0.9 (1.4-3.8)0.55Other complications Major bleeding56 (3.5)6(2.3)5 (4.1)0.6 (0.2-1.6)1.2 (0.4-3.1)0.79Rethoracotomy45 (2.8)12 (4.5)4 (3.3)1.6 (0.8-3.2)1.2 (0.4-3.5)0.31Arrhythmia207 (13)43(16)35 (29)1.3 (0.9-1.9)2.7 (1.8-4.2)< 0.0001Renal failure (dialysis)88 (5.5)35 (13)27 (22)2.6 (1.7-4.0)4.9 (3.0-8.1)< 0.0001
Table 3—Patient Data According to Survival Status
| Variables | Survivors (n = 1,938) | Nonsurvivors (n = 40) | OR (95% CI) | p Value |
| Preoperative data | ||||
| Age, yr | 66 ± 9 | 71 ± 7 | 0.01 | |
| Female gender | 396 (20) | 14 (35) | 2.1 (1.0-4.2) | 0.03 |
| Body weight, kg | 81 ± 14 | 75 ± 13 | < 0.001 | |
| Diabetes mellitus | 565 (29) | 35 (88) | 17 (6.3-49.6) | < 0.0001 |
| Hypertension | 1,608 (83) | 38 (95) | 3.9 (0.9-23.5) | 0.05 |
| Hyperlipidemia | 1,552 (81) | 35 (88) | 1.7 (0.6-5.1) | 0.32 |
| Family history | 876 (45) | 18 (45) | 1.0 (0.5-1.9) | 1.00 |
| Obesity! | 614 (32) | 16 (40) | 1.4 (0.7-2.8) | 0.30 |
| Smoking history | 1,146 (59) | 17 (43) | 0.5 (0.3-1.0) | 0.05 |
| History of stroke | 123 (6) | 6(15) | 2.6 (0.96-6.6) | 0.04 |
| COPD | 314 (16) | 8 (20) | 1.3 (0.4-3.0) | 0.52 |
| PVD | 277 (14) | 11 (28) | 2.3 (1.1-4.8) | 0.04 |
| Renal disease! | 269 (14) | 5 (13) | 0.9 (0.3-2.4) | 1.00 |
| Dialysis | 41 (2.1) | 1 (2.5) | 1.2 (0.1-8.3) | 0.57 |
| Previous MI§ | 689 (36) | 19 (48) | 1.6 (0.8-3.2) | 0.13 |
| Previous PCI | 359 (19) | 14 (35) | 2.4 (1.2-4.8) | 0.01 |
| CCS angina class III-IV | 1,173 (61) | 35 (88) | 4.6 (1.7-13.3) | < 0.001 |
| Left mainstem disease | 507 (26) | 10 (25) | 0.9 (0.4-2.0) | 1.00 |
| One-vessel disease | 53 (2.7) | 0(0) | 0.1 (0.0-4.5) | 0.62 |
| Two-vessel disease | 288 (15) | 3 (7) | 0.5 (0.1-1.6) | 0.26 |
| Three-vessel disease | 1,597 (82) | 37 (93) | 2.6 (0.8–10.8) | 0.14 |
| LV ejection fraction, % | 61 ± 17 | 56 ± 17 | < 0.05 | |
| cTnI, ng/mL | 0.3 ± 1.5 | 1.2 ± 2.6 | < 0.0001 | |
| Intraoperative data | ||||
| ACC time, min | 5 |
2
±
0
778 ± 24 0.03CPB time, min109 ± 39141 ± 3 < 0.0001Cardioplegia, mL1527 ± 4831710 ± 615 < 0.05Reperfusion time, min34 ± 1743 ± 20 < 0.001Adverse events and postoperative complications IABP support89 (4.6)14 (35)11.2 (5.3-23.2)< 0.0001LCOS29 (1.5)17 (43)48.7 (22.3-107)< 0.0001PMI112 (5.8)24 (60)24.5 (12.1-49.8)< 0.0001Other complications Arrhythmia262 (14)23 (58)9.8 (4.8-20.0)< 0.0001Renal failure (dialysis)115 (5.9)35 (88)110(41-328)< 0.0001
Table 4—Univariate Factors Associated With Death, MACE, and the Composite Study End Point
| Factors | Death* | PMI | LCOS | Death/PMI/LCOS | ||||
| OR (95% CI) | p Value | OR (95% CI) | p Value | OR (95% CI) | p Value | OR (95% CI) | p Value | |
| Age | 1.1 (1.0-1.1) | < 0.001 | 1.0 (0.9-1.0) | 0.15 | 1.0(1.0-1.1) | 0.03 | 1.0(1.0-1.04) | 0.01 |
| Gender | 2.1 (1.1-4.1) | 0.02 | 1.4 (0.9-2.1) | 0.12 | 2.0 (1.1-3.6) | 0.02 | 1.4 (0.9-2.1) | 0.07 |
| Obesity | 0.9 (0.5-1.7) | 0.80 | 0.8 (0.5-1.1) | 0.19 | 1.2 (0.7-2.1) | 0.52 | 0.8 (0.6-1.1) | 0.14 |
| Smoking | 0.4 (0.2-0.8) | 0.01 | 0.8 (0.6-1.2) | 0.27 | 0.4 (0.2-0.7) | < 0.01 | 0.9 (0.6-1.2) | 0.36 |
| Diabetes mellitus | 1.9 (1.0-3.5) | 0.05 | 1.3 (0.9-1.9) | 0.15 | 1.1 (0.6-2.1) | 0.67 | 1.2 (0.9-1.7) | 0.27 |
| Hypertension | 3.8 (0.9-15.9) | 0.07 | 1.6 (0.9-2.7) | 0.12 | 1.4 (0.6-3.3) | 0.45 | 1.6 (0.9-2.8) | 0.06 |
| Hyperlipidemia | 1.8 (0.7-4.5) | 0.24 | 1.3 (0.8-2.1) | 0.26 | 1.5 (0.7-3.3) | 0.35 | 1.3 (0.8-2.0) | 0.23 |
| Renal disease | 1.0 (0.3-2.8) | 0.96 | 1.3 (0.8-2.3) | 0.30 | 1.7 (0.8-3.8) | 0.16 | 1.2 (0.7-2.0) | 0.56 |
| COPD | 1.5 (0.7-3.2) | 0.29 | 1.5 (1.01-2.4) | 0.05 | 1.7 (0.9-3.4) | 0.10 | 1.4 (0.9-2.1) | 0.13 |
| PVD | 2.6 (1.3-5.2) | 0.01 | 1.6(1.01-2.4) | 0.05 | 1.0 (0.4-2.2) | 0.98 | 1.6 (1.1-2.4) | 0.02 |
| Previous MI | 1.5 (0.8-2.9) | 0.19 | 1.2 (0.8-1.7) | 0.33 | 1.8 (1.0-3.1) | 0.05 | 1.3 (0.9-1.8) | 0.99 |
| Previous PCI | 2.4 (1.2-4.7) | 0.01 | 1.2 (0.8-1.9) | 0.36 | 1.1 (0.6-2.3) | 0.72 | 1.3 (0.9-1.9) | 0.24 |
| CCS angina class III-IV | 5.5 (2.0-15.5) | 0.01 | 2.5 (1.6-3.9) | < 0.0001 | 1.7 (0.9-3.2) | 0.12 | 2.7 (1.8-4.0) | < 0.0001 |
| Left mainstem disease | 1.0 (0.5-2.0) | 0.91 | 1.3 (0.9-1.9) | 0.16 | 1.1 (0.6-2.0) | 0.85 | 1.3 (0.9-1.9) | 0.14 |
| LV ejection fraction, % | 0.98 (0.96-1.0) | 0.12 | 0.99 (0.97-1.0) | 0.07 | 0.96 (0.9-1.0) | < 0.001 | 0.99 (0.97-0.99) | 0.03 |
| CK, IU/L | 1.0(0.99-1.0) | 0.62 | 1.0(1.0-1.003) | 0.11 | 1.0 (0.99-1.004) | 0.37 | 1.0 (0.99-1.0) | 0.19 |
| cTnI, ng/mL | 1.1 (1.0-1.1) | < 0.0001 | 1.1 (1.0-1.1) | < 0.0001 | 1.1 (1.0-1.1) | < 0.0001 | 1.1 (1.0-1.1) | < 0.0001 |
| cTnI > 0.1 ng/mL | Reference | Reference | Reference | Reference | ||||
| group | group | group | group | |||||
| cTnI 0.11-1.5 ng/mL | 2.0 (0.9-4.6) | 0.08 | 1.5 (0.9-2.5) | 0.09 | 2.8 (1.4-5.9) | < 0.01 | 1.4 (0.9-2.3) | 0.12 |
| cTnI > 1.5 ng/mL | 5.9 (2.8-12.7) | < 0.0001 | 3.7 (2.3-6.2) | < 0.0001 | 9.3 (4.8-18.3) | < 0.0001 | 5.0 (3.2-7.9) | < 0.0001 |
Table 5—Multivariate Factors Associated With Death, MACE, and the Composite Study End Point, With cTnI as a Categorical Variable
| Factors | Death* | PMI | LCOS | Death/PMI/LCOS | ||||
| OR (95% CI) | p Value | OR (95% CI) | I | |||||
p ValueOR (95% CI)I
p ValueOR (95% CI)p ValueAge1.1 (1.0-1.1)0.01 1.0 (0.98-1.1)0.371.0 (0.99-1.0)0.11Gender1.3 (0.6-2.7)0.50 1.2 (0.6-2.6)0.611.0 (0.7-1.6)0.90Smoking0.6 (0.3-1.2)0.15 0.7 (0.3-1.4)0.27 Diabetes mellitus1.6 (0.8-3.2)0.17 Hypertension2.0 (0.5-8.7)0.34 1.1 (0.6-2.0)0.66COPD 1.1 (0.6-1.7)0.840.7 (0.3-1.7)0.39 PVD2.8 (1.3-5.9)0.011.4 (0.9-2.4)0.17 1.5 (0.9-2.4)0.12Previous MI 0.7 (0.4-1.6)0.451.0 (0.6-1.5)0.83Previous PCI2.8 (1.4-5.6)0.004 CCS angina class III-IV3.5 (1.2-10.2)0.022.4 (1.5-4.0)0.0004 2.2 (1.4-3.5)0.001LV ejection fraction, % 1.0 (0.98-1.01)0.390.97 (0.94-0.99)0.011.0 (0.98-1.01)0.27cTnI > 0.1 ng/mLReference Reference Reference Reference group group group group cTnI 0.11-1.5 ng/mL1.4 (0.5-3.6)0.491.5 (0.9-2.6)0.123.2 (1.3-7.5)0.011.4 (0.8-2.4)0.21cTnI > 1.5 ng/mL5.9 (2.5-14.0)< 0.00014.1 (2.3-7.0)< 0.000110.8 (4.6-25.1)< 0.00015.0 (3.0-8.5)< 0.0001
Table 6—Multivariate Factors Associated With Death, MACE, and the Composite Study End Point, With cTnI as a Continuous Variable
| Factors | Death* | PMI | LCOS | Death/PMI/LCOS | ||||
| OR (95% CI) | I | |||||||
p ValueOR (95% CI)p ValueOR (95% CI)p ValueOR (95% CI)p ValueAge1.1 (1.0-1.1)0.02 1.0(0.98-1.1)0.391.0 (0.99-1.0)0.18Gender1.3 (0.6-2.8)0.46 1.3 (0.6-2.9)0.511.1 (0.7-1.7)0.72Smoking0.6 (0.3-1.2)0.17 0.6 (0.3-1.2)0.15 Diabetes mellitus1.3 (0.7-2.7)0.44 Hypertension2.1 (0.5-8.9)0.33 1.1 (0.6-2.0)0.63COPD 1.1 (0.7-1.9)0.710.6 (0.2-1.7)0.38 PVD2.5 (1.2-5.2)0.011.3 (0.8-2.2)0.35 1.3 (0.8-2.1)0.28Previous MI 1.3 (0.6-2.6)0.521.2 (0.8-1.8)0.35Previous PCI2.7 (1.3-5.6)0.01 CCS angina class III-IV3.5 (1.2-10.2)0.022.6 (1.6-4.3)< 0.001 2.3 (1.5-3.7)< 0.001LV ejection fraction, % 1.0 (0.98-1.0)0.330.96 (0.9-1.0)0.0021.0 (0.98-1.0)0.32cTnI, ng/mL1.1 (1.0-1.1)< 0.0011.1 (1.0-1.1)< 0.0011.1 (1.0-1.1)< 0.0011.1 (1.0-1.1)< 0.001
Table 7—Cutoff Levels and Test Characteristics for Preoperative cTnI Value
| Preoperative cTnI | Death | Death/PMI/LCOS |
| Area under curve | 0.72 ± 0.08 | 0.67 ± 0.05 |
| cTnI cutoff point, ng/mL | 1.50 | 1.50 |
| Sensitivity, % | 75.0 | 63.0 |
| Specificity, % | 78.5 | 71.3 |
