Using postoperative cardiac Troponin-I (cTi) levels to detect myocardial ischaemia in patients undergoing vascular surgery

Cardiovascular Surgery, Vol. 9, No. 3, pp. 254–265, 2001  2001 The International Society for Cardiovascular Surgery. Published by Elsevier Science Ltd All rights reserved. Printed in Great Britain 0967-2109/01 $20.00

PII: S0967-2109(00)00139-3

www.elsevier.com/locate/cardiosur

Using postoperative cardiac troponin-I (cTi) levels to detect myocardial ischaemia in patients undergoing vascular surgery N. Andrews*, J. Jenkins†, G. Andrews‡ and P. Walker§ *Department of Surgery, Royal Brisbane Hospital, Brisbane, Australia; †Department of Vascular Surgery, Royal Brisbane Hospital, Brisbane, Australia; ‡School of Applied Psychology, Griffith University, Gold Coast Campus, Brisbane, Australia and §Department of Surgery, University of Queensland, Brisbane, Australia Background. Cardiac complications occur commonly in vascular surgery patients. Diagnosis of cardiac complications is difficult because of the inaccuracies associated with traditional cardiac enzyme measurements. CTi, a highly sensitive and specific marker of myocardial injury, may be able to detect cardiac complications with greater ease and accuracy. Methods. The study prospectively examined 100 consecutive patients who underwent major vascular surgery between 6/7/98 and 31/12/98 at the Royal Brisbane Hospital. Daily measurements of cTi, creatine kinase (CK), creatine kinase MB (CKMB), CKMB index, renal function and haemoglobin were taken for three postoperative days. One postoperative electrocardiograph (ECG) was taken. An extensive cardiac history was taken. Intraoperative and postoperative events were recorded. Findings. There were 100 patients. 18 patients (18%) had a cTi elevation. On the basis of classical diagnostic criteria, 15 patients (15%) suffered one or more cardiac complication (either myocardial infarction, congestive cardiac failure, unstable angina or atrial fibrillation). One patient (1%) who had a cTi elevation died. CTi elevation occurred in five patients (5%) who were not diagnosed with cardiac complications based on traditional criteria. Despite not meeting specific diagnostic criteria for cardiac complications, all patients showed signs and symptoms that could be attributed to myocardial ischaemia. Every patient who developed congestive cardiac failure or atrial fibrillation had a cTi elevation. A Chi-square analysis revealed a significant association between cTi elevation and postoperative cardiac complications. Four variables contributed small but significant amounts of unique variance to the prediction of peak cTi on linear regression analysis. These were peak CKMB index, postoperative congestive cardiac failure, postoperative chest pain and postoperative cardiac complications. Conclusions. Routine cTi monitoring of postoperative vascular patients would be an effective and inexpensive way to detect patients with cardiac complications. The relationship between postoperative cTi elevation and significant coronary artery disease remains to be shown.  2001 The International Society for Cardiovascular Surgery. Published by Elsevier Science Ltd. All rights reserved Keywords: cardiac troponin-I, vascular surgery, myocardial ischaemia, cardiac complications

Introduction The 30-day mortality rate for patients undergoing vascular surgery is up to seven percent with myocarCorrespondence to: N. Andrews. Tel.: +61-7-363-68111; fax: +614-021-29058

254

dial infarction contributing to approximately 50% of these postoperative deaths [1]. Myocardial ischaemia, detected on continual postoperative electrocardiographic monitoring, accurately predicts patients that are likely to progress to clinical cardiac complications such as myocardial infarction or congestive cardiac failure [2–4]. However, myocardial ischaemia is usually clinically silent [5–7]. In routine cliniCARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

cal practice, patients with myocardial ischaemia are usually only recognized when they develop adverse cardiac events, either in or out of hospital. Raby et al. [8] showed that of 176 patients with peripheral vascular disease, not undergoing surgery, 32 (18%) had episodes of myocardial ischaemia detected on routine holter monitoring. Over the next 18 months, 12 (38%) of these patients had adverse cardiac events, including six cardiac deaths. The presence of ischaemia was the only independent predictor of outcome for these patients. For patients undergoing vascular surgery, studies have shown that myocardial ischaemia occurs in 20–40% of patients postoperatively [9]. More than 50% of these patients proceed to clinical ischaemic events [10]. With such a high frequency of myocardial ischaemia experienced by vascular patients and the adverse cardiac outcomes that often result, it would seem that continual electrocardiographic monitoring would be warranted for all vascular patients postoperatively. However, this is rarely available at most major hospitals. The only wards fitted with cardiac monitoring devices are the coronary care unit and the intensive care unit. Beds in those units are limited and not usually available for the routine care of vascular patients. In order to detect adverse cardiac events in patients undergoing vascular surgery, routine 12 leads electrocardiograms and routine cardiac enzyme measurements could be performed on all patients. However, there are a number of limitations to this proposal. The majority of myocardial infarctions that occur in vascular patients are non Q-wave infarcts [11]. This creates difficulty in interpreting ECG changes [12]. Traditional cardiac enzymes, creatine kinase (CK) and creatine kinase MB fraction (CKMB) are not sufficiently sensitive and specific to serve this purpose. CK and CKMB are released only when myocardial necrosis occurs [13], not with transient loss of cell membrane integrity as occurs with ischaemia. The CKMB content of normal myocardium is no different from that of slow-twitch skeletal muscle [14]. CK and CKMB are often falsely elevated postoperatively because they are released from skeletal muscle intraoperatively [15,16]. The specificity of CKMB can be enhanced by the calculation of the CKMB/CK ratio (CKMB index). However, this ratio is not very sensitive. The greater the extent of muscle injury, the more likely changes in CKMB due to cardiac injury will be missed [17]. CKMB may also be falsely elevated in patients with renal failure [18]. Routine monitoring of cardiac troponin I levels may be an alternative to the methods just discussed to detect myocardial ischaemia. Cardiac Troponin I (cTi) is a myocardial regulatory protein found only in cardiac tissue [19–21]. Unlike CK-MB, cTi is highly specific for myocardial CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

injury [22,23]. Skeletal muscle does not express cTi at any developmental stage or in response to any pathological stimulus [24]. Thus, cTi is not elevated in patients with chronic muscle disease unless concomitant acute myocardial damage is present [25]. CTi is a more sensitive predictor of minor degrees of myocardial injury than CKMB. This may be because it is 13 times more abundant in the myocardium than CKMB [26,27]. Multiple studies have shown that cTi is not elevated in patients with renal failure, although cardiac troponin T can be [28–30]. Elevations of cTi have been shown to be an independent risk factor for mortality in patients with unstable angina and non Q-wave myocardial infarction [31]. The higher the level of cTi, the greater the mortality rate [32]. CTi remains elevated after myocardial damage for 4–7 days, while CKMB remains elevated for only 2–3 days [33]. Elevations of cTi have been shown to correlate closely with new regional wall motion abnormalities on echocardiography [34]. Given the features of cTi as described above, it would be safe to assume that a cTi elevation is indicative of myocardial damage. The study was designed to 1. Determine the extent of cTi elevation in vascular surgery patients. 2. Examine the association between elevated cTi and cardiac complications diagnosed by traditional criteria. 3. Assess factors that coexist with elevated cTi in vascular surgery patients.

Patients and methods 100 consecutive patients undergoing vascular surgery at the Royal Brisbane Hospital [carotid endarterectomies (CEA), abdominal aortic aneurysm repair (AAA), lower extremity bypass surgery (FPBP) and lower extremity amputation (amp)] between 6/7/1998 and 31/12/1998 were prospectively enrolled in the study. Preoperative data gathered included a full vascular and cardiac history. Routine physical examination was performed, looking specifically for signs of cardiac disease. Preoperative investigations included a 12 lead electrocardiogram (ECG), full blood examination, electrolytes and renal function, coagulation profile, CK, CKMB, CKMB index, cTi and chest radiograph. Intraoperative events such as hypotension were recorded. Postoperatively, the blood tests just mentioned were repeated every day for at least three days. An ECG was taken on the second postoperative day. Any cardiac symptoms were documented. Postoperative cardiac complications include myocardial infarction (MI), unstable angina pectoris (UAP), atrial fibrillation (AF) and congestive cardiac failure (CCF). These cardiac complications are 255

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

defined by classical definitions, irrespective of the cTi level. Myocardial infarction is defined as two out of three of 1. Chest pain 2. New ECG changes of myocardial infarction in at least two consecutive leads — ST elevation >1 mm, Q waves, T wave inversion, Bundle branch block 3. Elevated CKMB and CKMB index Unstable angina is defined as chest pain associated with horizontal ST depression of at least 0.5 mm in two or more consecutive leads. Congestive cardiac failure is defined as dyspnoea associated with classic chest radiograph findings (upper lobe diversion, perihilar shadowing, Kerley B lines, cardiomegaly and diffuse patchy lung shadows). Assays of molecular markers CTi was measured on a Centaur Bayer analyzer. Upper limit of normal was 0.1 µg/l. CK was measured on a Hitachi 747200 analyzer with an upper reference limit of 210 iu/l. CKMB was measured on a Centaur Bayer analyzer. The upper limit of normal was 8 µg/l. CKMB index is the CKMB:CK ratio. The upper limit of normal was 2. For these biochemical markers, an elevation was defined as any value above the upper limit of normal. Creatinine was measured on a Hitachi 747200 analyzer. Upper limit of normal was 0.14 mmol/l. Urea was measured on a Hitachi 747200 analyzer. Upper limit of normal was 9 mmol/l. Postoperative renal dysfunction was defined as a postoperative creatinine more than 50% higher than the upper limit of normal if the preoperative creatinine was normal. Patients were classified as having chronic renal impairment if the preoperative creatinine was more than 50% higher than the upper limit of normal. Acute on chronic renal dysfunction was defined as a postoperative creatinine more than 50% above the preoperative level. Statistical analysis Descriptive statistics were used to examine the extent of cTi elevation in vascular surgery patients. A chi-square analysis was used to evaluate the association between cTi elevation and traditional diagnostic criteria for cardiac complications. Bivariate correlations and multiple regression (with appropriate analysis of residuals) were used to examine the relations between cTi elevation and coexisting factors. CTi was treated as a continuous variable for the bivariate and linear regression analysis. 256

Results One hundred consecutive patients were included in the study. The mean age was 70.5 yr (SD = 10 yr, range 41–98 yr). Thirty-five patients were female. Extent of cTi elevation Elevated cTi levels occurred in 18 patients (18%) postoperatively. Of the 41patients who underwent lower extremity bypass grafting, six (15%) had postoperative cTi elevations. The corresponding figures for emergency repair of ruptured AAA, elective repair of AAA, CEA and amputations were 4/6 (67%), 5/25 (20%), 1/22 (5%) and 2/6 (33%). One patient was excluded from the study because her operation was cancelled She is reported here for interest. The 86-yr-old woman was to have a FPBP. During insertion of an epidural for anaesthesia, the patient became acutely dyspnoeic and hypotensive. ECG revealed anterior T wave inversion. Chest radiograph revealed changes consistent with CCF. The planned procedure was abandoned and the patient returned to the ward. CTi was 0.39 (0–0.1) later that afternoon. CK and CKMB remained in the normal range. On the second postoperative day cTi peaked at 0.92, CK was elevated at 473 (0–140), CKMB was 12.3 (0–8) and CKMB index was 2.6 (0–2). Echocardiograph revealed an ejection fraction of 48% with inferior akinesis. AF developed later that day. The patient continued poorly with symptoms of CCF. There was a second elevation of cTi to 0.12 on the eighth postoperative day. This peaked at 0.38 on the tenth postoperative day. There was no associated CKMB elevation. CTi then returned to normal. On the twenty-second postoperative day the patient developed a further exacerbation of CCF. This was associated with another cTi elevation to 0.85. CKMB was elevated to 9.8 and CKMB index to 18.1. The patient died of CCF on the twentyfourth postoperative day. Association between cTi and cardiac complications diagnosed by traditional criteria On the basis of classical diagnostic criteria, 15 patients (15%) suffered one or more cardiac complication (either MI, CCF, AF or UAP). As reported previously, eighteen patients had elevated cTi levels postoperatively. Patients were cross-classified according to their cTi elevation status and whether the diagnostic criteria for cardiac complications were met. The frequencies are shown in Table 1. A Chisquare analysis revealed a significant association between cTi elevation and postoperative cardiac complications, χ2 (1, N = 100) =62.33, P⬍0.001. Thirteen of the 15 patients who suffered cardiac complications also had a cTi elevation. The number of patients who suffered each particular cardiac comCARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al. Table 1 Chi-square analysis between cTi elevation and postoperative cardiac complicationsa Postoperative cardiac complications based on classical diagnostic criteria Yes No Total Elevated cTi

Yes No Total

‘a’ 13 ‘c’ 2 15

‘b’ 5 ‘d’ 80 85

18 82 100

a

Cell ‘a’ — Postoperative cardiac complication and cTi elevation; Cell ‘b’ — No postoperative cardiac complication but cTi elevation; Cell ‘c’ — Postoperative cardiac complication but no cTi elevation; Cell ‘d’ — No postoperative cardiac complication and no cTi elevation

plication and those patients who suffered cardiac complications but did not have a cTi elevation (patients in cell ‘c’ of Table 1) are discussed below. Twelve patients suffered an acute myocardial infarction. All of these patients except one had an elevated cTi level. This patient (#17, 60-yr-old male) developed central chest pain the second day following a femoral-popliteal bypass. ECG revealed changes consistent with a new anterior infarct. CTi level taken on the first postoperative day was 0.01 (0–0.1). CTi level on the second postoperative day, a few hours before the patient developed chest pain was 0.08. However, due to ward staff error, no further blood tests were taken until the fifth postoperative day. The cTi level on day five was 0.1. The cTi level on day six was 0.05 and then it decreased to 0.01. This trend suggests that a cTi elevation would have been observed if blood tests had been taken. No cardiac investigations were performed and the patient was discharged. One (1%) patient had postoperative UAP. He did not have a cTi elevation. Patient #40 (52-yr-old male) underwent a FPBP. On the second postoperative day he developed central chest pain. ECG revealed anterolateral ST depression. There was no cardiac enzyme leak. Coronary angiography was performed on the sixth postoperative day. This revealed an 80% stenosis of the left main coronary artery, an 80% proximal stenosis of the right coronary artery and a 70% stenosis of the first diagonal. The patient underwent uncomplicated coronary artery bypass grafting on tenth postoperative day. Eleven patients developed CCF. Of these, nine also sustained an acute MI. All patients with CCF had elevated cTi levels. Three patients developed new AF postoperatively. They all had elevated cTi levels. There was one postoperative death (1%). She had an elevated cTi level. Five (5%) patients had elevated cTi levels but were not diagnosed with cardiac complications (patients in cell ‘b’ in Table 1). These patients are discussed here. Despite not meeting diagnostic criteria for specific cardiac complications, they all CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

showed signs and symptoms that could be attributed to myocardial ischaemia. Patient #21 (65-yr-old female) underwent ilealanterior tibial PTFE graft. On the second postoperative day she returned to theatre for a below knee amputation, excision of a popliteal artery false aneurysm and removal of the PTFE graft. From a cardiac point of view, she was asymptomatic postoperatively. However, just prior to her return to theatre, it was noted that her cTi was elevated at 0.71 (0–0.1). CK was 274 (0–140), CKMB was 24.6 (0–8) and CKMB index was 9 (0–2). There were no new changes on the ECG. CTi peaked at 0.75 postoperatively. Postoperative echocardiography revealed severely impaired left ventricular function with an ejection fraction of 20%. The patient went on to rehabilitate after her below knee amputation with no further cardiac investigations. Patient #39 (54-yr-old male) underwent an iliacfemoral bypass graft and femoral endarterectomy. On the third postoperative day he became confused, hypoxemic and his urine output decreased. ECG showed anterior ST segment elevation. CTi peaked at 0.15 (0–0.1). CK peaked at 650 (0–210) on the first postoperative day. CKMB and CKMB index stayed within the normal range. Echocardiography on the tenth postoperative day showed normal left ventricular size and function. Exercise stress test on the twelfth postoperative day was positive for ischaemia in the inferior and lateral leads. The patient received no further cardiac follow up. Patient #85 (61-yr-old male) developed chest pain the day after his elective abdominal aortic aneurysm repair. There were no new changes on his ECG. Peak cTi was 0.16 (0–0.1). CK peaked at 1230 (0– 210) but CKMB and CKMB index remained within the normal range. The patient had no further cardiac investigations. Patient #86 (84-yr-old male) underwent a femoral-peroneal bypass. On the second postoperative day he collapsed. ECG revealed lateral ST depression. CTi rose to 0.16 (0–0.1) that day and peaked at 0.32 on the fourth postoperative day. CK and CKMB remained within the normal range. The patient had no further cardiac investigations. Patient #91 (72-yr-old male) became confused on the second day after his aortobifemoral bypass graft. ECG revealed lateral T wave inversion. Peak CK was 911 (0–210) and CKMB 14.9 (0–8). CKMB index was normal. CTi peaked at 0.16 (0–0.1) on postoperative days two and three. This patient went on to have an uneventful recovery and had no further cardiac investigations. Accuracy of traditional cardiac enzyme elevations with cTi as the gold standard Twelve patients (12%) had an elevation of CKMB and CKMB index. Three of these patients (#59, 257

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

#61, #70) did not have a cTi elevation. These three patients had no other indicators of myocardial ischaemia, symptomatic, radiographic or electrocardiographic. The cardiac enzyme elevations tended to be small and are listed for patients #59, #61 and #70 respectively. CK was 271, 861, 712 (0–140); CKMB was 8.4, 12.5, 9.8 (0–8); CKMB index was 3.1, 2.4, 2.6 (0–2). Sixteen patients had a CKMB elevation. Of these patients, six did not have a cTi elevation. One patient had lateral T wave inversion on ECG but none of the other patients were suspected clinically of having myocardial ischaemia. Fifty-seven patients had elevated CK levels. Forty-two of these patients did not have a cTi elevation. Assuming that cTi is the gold standard biochemical marker for myocardial damage, the sensitivity that CKMB and CKMB index has for detecting myocardial damage is 50%. The specificity is 96.3%. The sensitivity of CKMB alone is 55.5%. The specificity is 92.7%. The sensitivity of CK is 83.3% and the specificity is 48.8%. Frequency of ECG changes compared to cTi elevation Thirty patients (30%) had new postoperative ECG changes. Only 11 of these patients (37%) also had a cTi elevation. Eleven patients (11%) had new ST elevation on their postoperative ECG. Six (55%) of these also had a cTi rise. Sixteen patients had new ST depression on their postoperative ECGs. Eight (50%) of these patients also had a cTi rise. Sixteen patients had new Q waves on the ECG. Six (38%) had a cTi rise. Patients were cross-classified according to their cTi elevation status and the presence of postoperative ECG changes. The frequencies are shown in Table 2. A Chi-square analysis revealed a significant association between cTi elevation and postoperative ECG changes, χ2 (1, N = 100) =15.36, P⬍0.001. Factors that coexist with elevated cTi Simple bivariate correlations between preoperative, intraoperative and postoperative variables and peak cTi were examined using Pearson’s correlation coefficient. The significant bivariate correlations with

Table 2 Chi-square analysis between cTi elevation and postoperative ECG changes Elevated cTi

Postoperative Yes ECG changes No Total

258

Yes

No

Total

11

19

30

7 18

63 82

70 100

r>0.20 are shown in Table 3. The significant variables with r>0.20 were used in a multiple regression analysis in which the dependent variable was peak cTi. Five variables whose Pearson’s correlations with peak cTi were significant were not included in the multiple regression. This was done to maintain the cases:independent variable ratio as discussed below. Preliminary analyses of the residuals indicated an outlier (patient #67 had an extreme score of 41.6 on peak cTi. The remaining subjects had scores between 0.01 and 5.1). This patient was excluded from the analyses, reducing the maximum sample size to 99. The number of cases (99) is small in relation to the number of variables (15) measured. However the cases:independent variable ratio of 6.6 is above the minimum of 5 cases per independent variable as recommended by Tabachnick and Fidell (1989) [35]. Pair wise deletion of missing data was used. The independent variables accounted for 0.75 of total variance in peak cTi level, Multiple R = 0.87, F(15,57) = 11.52, P⬍0.001. Four variables contributed small but significant amounts of unique variance to the prediction of peak cTi. These were peak CKMB index (0.07), postoperative congestive cardiac failure (0.04), postoperative chest pain (0.02) and postoperative cardiac complications (0.02). The remaining variance (0.60) was shared among the 15 independent variables. Table 4 shows the B and β coefficients for the regression equation cTi elevation and postoperative haemoglobin Patients with cTi elevations tended to have lower postoperative haemoglobins than those without cTi elevations. The lowest postoperative haemoglobin was less than 10 (normal range 12–16) in 50% of patients with an elevated cTi level. In those patients without an elevated cTi level, the lowest postoperative haemoglobin was less than 10 in 28%. However, these results did not reach statistical significance on bivariate analysis (r = ⫺0.171, P = 0.09). cTi elevation and renal function Six patients (6%) had preoperative chronic renal impairment. There was no significant association between chronic renal impairment and peak cTi level on bivariate analysis. Three patients (3%) developed acute renal dysfunction postoperatively. Although there was a significant correlation between peak cTi and the presence of postoperative renal dysfunction on bivariate analysis, there was no significant correlation on linear regression analysis. When each patient with renal dysfunction is reviewed separately, it seems that the cTi elevation is primarily related to the cardiac complication rather than the renal dysfunction. These cases are discussed below. Patient #26 (72-yr-old female) underwent emergCARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3 1.00 0.24 0.33 0.27 0.24 ⫺0.21 0.53 0.55 0.24 0.59 0.70 0.65 0.46 0.33 0.23 0.72

Peak cTi(a) NIDDM(b) Post nitrates(c) PostDiuretics(d) Posthepariniv(e) Postheparinsc(f) Chest pain(g) Dypnoea(h) AF postop(i) Card comp(j) CCF postop(k) MI postop(l) Inf comp(m) RF postop(n) High CK(o) HiCKMBindex(p) 1.00 0.09 ⫺0.06 0.17 ⫺0.24 0.32 0.22 0.13 0.27 0.31 0.28 0.13 0.15 ⫺0.07 0.34

b

c

1.00 0.45 0.11 ⫺0.16 0.42 0.32 0.23 0.42 0.39 0.35 0.23 0.27 0.11 0.32 1.00 0.22 ⫺0.25 0.25 0.35 0.21 0.26 0.28 0.25 0.21 0.04 0.08 0.22

d

1.00 ⫺0.55 0.13 0.21 0.17 0.10 0.18 0.07 0.17 0.20 ⫺0.11 0.28

e

f

1.00 ⫺0.19 ⫺0.10 ⫺0.06 ⫺0.14 0.14 ⫺0.11 ⫺0.06 ⫺0.14 0.05 ⫺0.27

g

1.0 0.44 0.37 0.67 0.56 0.62 0.16 0.18 0.33 0.32

h

1.0 0.15 0.64 0.73 0.59 0.35 0.30 0.10 0.37

i

1.00 0.34 0.43 0.41 ⫺0.02 ⫺0.03 0.01 0.27

j

1.00 0.79 0.84 0.14 0.28 0.12 0.57

k

1.00 0.74 0.19 0.38 0.16 0.56

l

1.00 0.18 0.01 0.15 0.61

m

1.00 0.30 0.06 0.48

n

1.00 0.01 0.34

1.00 ⫺0.05

o

a Peak cTi — Highest cTi measured as a continuous variable; NIDDM — Non Insulin Dependent Diabetes Mellitus; Post nitrates — Postoperative treatment with oral or topical nitrate medications; PostDiuretics — Postoperative treatment with diuretic medications; Posthepariniv — Postoperative treatment with intravenous heparin infusion to maintain APTT 60–120s; Postheparinsc — Postoperative treatment with subcutaneous heparin, 5000 u twice daily; Chest pain — Postoperative complaint of chest pain; Dyspnoea — Postoperative complaint of dyspnoea; AF postop — New atrial fibrillation on postoperative ECG; Card comp — Postoperative cardiac complications (AF, MI, CCF, unstable angina); CCF postop — Postoperative congestive cardiac failure; MI postop — Postoperative myocardial infarction; Inf comp — Postoperative local infectious complication such as wound infection, conduit infection; RF postop — Postoperative renal dysfunction or acute on chronic renal dysfunction; High CK — Peak CK level, treated as a continuous variable; HiCKMBindex — Peak CKMB index, treated as a continuous variable

a

Pearson’s correlation for those variables that correlate significantly with peak cTia

Variable

Table 3

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

259

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al. Table 4

B and β coefficients for the regression equationa

Variable

Gender Age NIDDM Preop ACE inhibitor Preop digoxin Postop nitrates Postop diuretics Postop aspirin Postop heparin infusion Postop heparin subcutaneous Postop chest pain Postop dyspnoea Postop AF Postop cardiac complication Postop CCF Postop MI Postop infectious complication Postop renal failure Peak CK Peak CKMB index

Unstandardized coefficient B

Standardized coefficient β

⫺0.030 0.000 ⫺0.157 0.029 ⫺0.017 ⫺0.020 0.009 ⫺0.265 0.111 0.101

⫺0.019 0.007 ⫺0.079 0.019 ⫺0.006 ⫺0.012 0.006 ⫺0.150 0.049 0.064

0.566 0.059 ⫺0.617 ⫺0.657 1.119 0.351 0.546

0.260 0.028 ⫺0.118 ⫺0.321 0.459 0.150 0.105

⫺0.134 0.000 0.088

⫺0.040 0.102 0.485

a

NIDDM — Non Insulin Dependent Diabetes Mellitus; Preop — Preoperative; Postop — Postoperative

ency repair of a ruptured AAA. She had a history of stable angina. Postoperatively she was confused, dyspnoeic and hypotensive. Chest radiograph revealed changes consistent with CCF. ECG on the first postoperative day revealed inferolateral ST depression and anterior ST elevation. ECGs taken on the second postoperative day revealed inferolateral ST depression and new anterior Q waves. There was an intermittent complete heart block and an intermittent junctional rhythm. CTi was elevated to 0.2 (0–0.1) on the first postoperative day. CTi peaked at 0.65 on the third postoperative day. CK peaked at 1000 (0–140) on the first postoperative day and decreased to normal after that. CKMB peaked at 11.6 (0–8) on the second postoperative day. CKMB index remained within the normal range. Creatinine rose from 0.13 (0.05–0.12) preoperatively to peak at 0.29 on the fifth postoperative day. Urea rose from 9 (3.5–9) preoperatively to peak at 35.6 on the fifth postoperative day. Creatinine and urea gradually returned to normal after this. Creatinine and urea remained within the normal range for the first postoperative day when the cTi elevation was first observed. Echocardiograph taken on the third postoperative day revealed mild left ventricular hypertrophy but was otherwise normal. The patient recovered and was discharged without any further cardiac investigations. Patient #32 (90-yr-old female) underwent an emergency femoral-anterior tibial bypass graft. This was complicated by CCF. On the first postoperative 260

day she became hypotensive and her urine output was poor. Later that day she developed central chest pain. The following day she remained hypotensive and became confused and dyspnoeic. Chest radiograph revealed changes consistent with CCF. ECG did not reveal any new changes. CTi was elevated to 0.23 (0–0.1) on the second postoperative day and peaked at 0.58 on the third postoperative day. CK and CKMB remained in the normal range. Preoperative creatinine was 0.11 (0.05–0.12). It peaked at 0.2 on the fourth postoperative day. Urea peaked at 14.5 (3.5–9) on the fourth postoperative day. The patient recovered and had no further cardiac investigations. Patient #67 (66-yr-old woman) underwent an elective repair of AAA. A few hours after her operation she suffered an acute MI and cardiac arrest. She was successfully resuscitated with DC cardioversion and adrenaline. ECG revealed anterolateral ST elevation, widespread T wave inversion and widespread new Q waves. The patient underwent urgent coronary angiography that revealed a complete stenosis of the proximal left anterior descending artery, a 50% proximal stenosis of the left circumflex artery and a 60% proximal stenosis of the right coronary artery. The left anterior descending lesion was successfully angioplastied and stented. Cardiac enzymes taken immediately following the cardiac arrest were all normal. Five hours after the cardiac arrest, cTi was 12.8 (0–0.1) and CK was 2930 (0–140). CTi peaked at 41.9, CKMB peaked at 277 (0–8) and CKMB index peaked at 12.1 (0–2) on the first postoperative day. Cardiac enzymes then gradually returned to normal. Preoperative creatinine was 0.15 (0.05–0.12) and urea was 9.4 (3.5–9). Creatinine peaked at 0.26 and urea peaked at 15 on the third postoperative day. Echocardiography performed one week postoperatively revealed apical akinesis but was otherwise normal. Other patients with cardiac complications and elevated cTi cell ‘a’ Table 1). The other patients with cardiac complications are discussed here for interest. Patient #1 (79-yr-old male) underwent a lower extremity amputation. On the second postoperative day he developed chest pain and increasing dyspnoea that lasted all night. The following morning he collapsed. ECG was difficult to interpret because of a preexisting left bundle branch block. Chest radiograph revealed changes consistent with CCF. CTi rose from 0.02 (0–0.1) prior to the onset of symptoms to 1.9 after the patient collapsed. CTi peaked at 5.1 and gradually returned to normal. CK peaked at 820 (0–210) on the third postoperative day. CKMB peaked at 161 (0–8) and CKMB index peaked at 25.4 (0–2) also on the third postoperative day. The patient was diagnosed with MI and CCF. CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

Echocardiograph taken on the eighth postoperative day showed a mildly dilated left ventricle with severe systolic dysfunction and an ejection fraction of 30%. The whole of the left ventricular wall was severely hypokinetic or akinetic except the basal-middle part of the anterolateral wall. The patient spent 3 days in the coronary care unit. He later returned to the vascular ward for rehabilitation. He had no further cardiac investigations. Patient #3 (72-yr-old female) underwent emergency repair of an expanding AAA. This was complicated by MI, CCF and AF. On the first postoperative day she developed dyspnoea and vague chest pain. Chest radiograph revealed changes consistent with CCF. ECG revealed new AF and inferolateral ST depression and T wave inversion. CTi peaked at 1.4 (0–0.1) on the third postoperative day. CK peaked at 739 (0–140) on the fourth postoperative day. No further cardiac investigations were undertaken. Patient #9 (77-yr-old female) underwent an above knee amputation on. This was complicated by MI and CCF. She had known ischaemic heart disease. Echocardiography performed 6 months prior to the operation revealed moderate left ventricular dysfunction with an ejection fraction of 35%. The anterolateral and inferolateral walls contracted well but the remaining left ventricle was hypokinetic. A stress sestamibi myocardial perfusion scan was also performed 6 months prior to the operation. It revealed no inducible ischaemia. On the day prior to her operation she had two episodes of central chest pain. ECG was difficult to interpret because there was a preexisting left bundle branch block. Just prior to her operation, cTi was noted to be elevated to 0.37 (0–0.1). CK remained in the normal range. The patient proceeded to have her operation. Postoperatively, cTi continued to rise and peaked at 4.5 on the second postoperative day. CK peaked at 2450 (0–140), CKMB peaked at 203 (0–8) and CKMB index peaked at 8.3 (0–2) on the first postoperative day. Echocardiography performed on the third postoperative day revealed a severe left ventricular dysfunction with an ejection fraction of 25%. The whole of the left ventricular wall was severely hypokinetic or akinetic except the basal part of the anterolateral and the basal part of the inferolateral wall. Dipyridamole sestamibi myocardial perfusion scan 2 weeks later revealed no inducible ischaemia but fixed perfusion defects involving the septal, apical, anteroapical and inferior walls. The patient’s ischaemic heart disease was treated medically. She was discharged during the third postoperative week to rehabilitate after her amputation. Patient #28 (62-yr-old male) underwent emergency repair of a ruptured AAA. This was complicated by CCF. On the second postoperative day his CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

gas exchange deteriorated. Chest radiograph revealed changes consistent with CCF. ECG did not reveal any new changes. CTi peaked at 0.48 (0–0.1) and CK peaked at 1400 (0–210) on the second postoperative day. CKMB and CKMB index remained within the normal range. Echocardiography on the seventh postoperative day revealed normal left ventricular size and function with an ejection fraction of 75%. An exercise stress test performed on 23/10/98 did not show any ischaemic ST segment changes but was difficult to interpret because of a preexisting right bundle branch block. Patient #37 (67-yr-old male) underwent elective repair of AAA. This was complicated by MI. On the third postoperative day he became hypotensive. ECG revealed lateral T wave inversion. CTi was elevated to 0.26 (0–0.1). CTi peaked at 0.48 on the fourth postoperative day. CK remained in the normal range. CKMB peaked at 9.8 (0–8) and CKMB index peaked at 18.1 (0–2) on the third postoperative day. No further cardiac investigations were undertaken. Patient #41 (60-yr-old male) underwent emergency repair of an expanding AAA. This was complicated by MI. The patient had known ischaemic heart disease. He had undergone coronary artery bypass grafting 14 years ago. He had complained of recurrent angina for the last four years. Coronary angiography one month preoperatively revealed a 50% ostial stenosis of the left main coronary artery, occluded vein grafts to the left anterior descending artery and right coronary artery, occluded left anterior descending artery and right coronary artery and a small, non dominant left circumflex coronary artery with no significant stenosis. Thallium perfusion scan one month preoperatively revealed perfusion abnormalities in the anterior, apical, septal and inferior walls. The ejection fraction was 25%. On the third postoperative day the patient developed central chest pain. ECG revealed lateral T wave inversion, anterior ST elevation and lateral ST depression. CTi peaked at 0.15 (0–0.1) that day. CK peaked at 898 (0–210) on the second postoperative day. CKMB and CKMB index remained in the normal range. The patient recovered and was referred to the cardiac surgeons for consideration of redo coronary artery bypass grafting. Patient #48 (98-yr-old female) underwent emergency repair of an expanding AAA. This was complicated by CCF and AF. Postoperatively, she developed a low urine output and was fluid loaded multiple times. She developed dyspnoea and chest radiograph revealed changes consistent with CCF. ECG revealed new inferior Q waves. CTi peaked at 1.1 (0–0.1) and CK peaked at 325 (0–140) on the second postoperative day. CKMB peaked at 22.6 (0–8) and CKMB index peaked at 10.6 (0–2) on the first postoperative day. No further cardiac investi261

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

gations were performed. The patient was discharged during the third postoperative week. She died one month later after having internal fixation of a fractured femur. Patient #49 (79-yr-old female) underwent an emergency femoral-popliteal interposition graft. This was complicated by MI, CCF and AF. She developed central chest pain and rapid AF on the second postoperative day. ECG revealed lateral T wave inversion and new AF. On this day cTi was 1.2 (0–0.1). CK at 379 (0–140), CKMB at 45.1 (0–8) and CKMB index at 11.9 (0–2) all peaked on this day. CTi peaked at 1.5 on the fourth postoperative day. She was treated with therapeutic heparin, nitrates, ACE inhibitors and beta-blockers. She recovered and had no further cardiac investigations. Patient #72 (57-yr-old female) underwent an embolectomy, endarterectomy and vein patch to the common femoral artery. This was complicated by MI, CCF and death. She had known ischaemic heart disease, having had three myocardial infarctions in the past and chronic stable angina. Echocardiography taken in 1996 showed an ejection fraction of 25%. On the ninth postoperative day, the patient developed central chest pain. She had another episode of chest pain associated with dyspnoea on the tenth and eleventh postoperative days. Chest radiograph revealed changes consistent with CCF. ECG revealed inferolateral T wave inversion and anterior ST elevation. CTi peaked at 0.14 (0–0.1) on the eleventh postoperative day. CK and CKMB stayed within the normal range. She was treated with ACE inhibitors, nitrates, diuretics and therapeutic clexane. Gated heart pool scan 2 weeks later revealed severe left ventricular dysfunction with an ejection fraction of 18%. She continued with severe CCF that was refractory to treatment. Her leg became ischaemic and she underwent an above knee amputation one month after the first operation. She continued to deteriorate with worsening CCF and died 2 weeks later. Patient #100 (72-yr-old female) underwent a carotid endarterectomy. This was complicated by MI and CCF. She had a history of stable angina. Preoperative echocardiography revealed moderate left ventricular dysfunction with an ejection fraction of 35%, akinesis of the apex and hypokinesis of all the remaining myocardium except the basal-inferior, inferoseptal and anterolateral walls. On the second postoperative day the patient developed central chest pain. On this day, cTi was 2.1 (0–0.1), CK was 910 (0–140), CKMB was 47.8 (0–8) and CKMB index was 5.3 (0–2). ECG did not reveal any new changes. On the third postoperative day the patient became dyspnoeic and chest radiograph revealed changes consistent with CCF. Cardiac enzymes all remained elevated but had started to decline. Echocardiograph on the eighth postoperative day was similar to the 262

preoperative echocardiograph. Dipyridamole thallium stress myocardial perfusion scan with rest redistribution on thirteenth postoperative day revealed reversible perfusion defects in the middle anterior wall, septum and apex. There was a fixed perfusion abnormality in the distal anterior segment. The patient had no further cardiac follow-up. In summary, 20 patients had a cardiac complication or cTi elevation. Nine (45%) of these patients received no cardiac investigations at all. Eight (40%) patients had one or more non-invasive investigation such as echocardiography, stress tests and perfusion scans. Only two (10%) patients had normal noninvasive investigations. Six (30%) patients received echocardiographs. Three (15%) echocardiographs were reported as normal. Three (15%) were abnormal. One (5%) patient had a gated heart pool scan which showed severely impaired left ventricular function. Two (10%) patients had exercise stress tests. One (5%) was positive for ischaemia. The other was normal but difficult to interpret because there was a right bundle branch block on the ECG. Two (10%) patients had myocardial perfusion scans. One (5%) demonstrated reversible ischaemia. The other demonstrated fixed defects only. Three (15%) patients, all of whom had no other cardiac investigations, went directly to coronary angiography. Two of these patients required coronary artery bypass grafting. The other required transluminal coronary angioplasty with stenting. No patients who underwent non-invasive investigations, even if those investigations were abnormal, proceeded to coronary angiography. All patients except three had been referred for cardiology opinion.

Discussion The frequency of cardiac ischaemia, as measured by cTi elevation (18%), was similar to that documented in the literature, for patients who had continual electrocardiographic monitoring postoperatively. Frank et al. [36] reported a 17% rate of myocardial ischaemia. Landesberg et al. [37] reported a 31.8% rate. An interesting point to note is that 4/6 (66.7%) of patients who underwent repair of a ruptured or expanding AAA had a cTi leak. Five (5%) patients had elevated cTi levels but were not diagnosed with cardiac complications. Despite not meeting diagnostic criteria for specific cardiac complications, they all showed signs and symptoms that could be attributed to myocardial ischaemia. By measuring cTi, it was possible to diagnose an additional five patients (5%) with myocardial damage who would otherwise have gone unnoticed. It is likely that most of the patients with specific cardiac complications would also have been missed if it were not for the routine cardiac enzymes and ECGs that were performed as part of the study. CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

CTi elevation occurred in patients with cardiac complications besides myocardial infarction. CTi was elevated in all patients with AF and CCF. One patient with an elevated cTi level died. No one with normal cTi levels died. Routine cTi monitoring would be an easy, effective and inexpensive way to detect patients with cardiac complications. A cTi assay costs $3.00. A hospital bed equipped with electrocardiographic monitoring facilities costs at least $1000 per day. A combination of CKMB and CKMB index together and CKMB alone were specific for myocardial damage (96.3 and 92.7% respectively). However, they were relatively insensitive for detecting myocardial damage (50 and 55.5% respectively). This is not unexpected because it is known that CKMB is only released from cardiac muscle when it is irreversibly damaged [38]. CK elevation was reasonably sensitive for myocardial damage (83.3%), but not very specific (48.8%). This is expected because CK is released from skeletal muscle intraoperatively [39,40]. 30% of patients had postoperative ECG changes that would be consistent with myocardial ischaemia. Only 37% of patients with ECG changes also had a cTi elevation. It is difficult to interpret the importance of postoperative ECG changes on a single twelve lead ECG in patients who have no other clinical indicators of myocardial ischaemia. The changes seen may be transient and meaningless or related to mechanical factors such as lead placement. However, ECG changes are probably more sensitive for detecting myocardial ischaemia than cTi elevations. Patient #40 developed UAP. His ECG revealed ST depression but he had no cTi leak. As reported above he had significant coronary artery disease. To discover the significance of minor changes on the 12 lead ECG, it would be necessary for each patient to also have 24 hour continuous electrocardiographic monitoring. Four variables contributed significant amounts of unique variance to the prediction of peak cTi on linear regression analysis. These were peak CKMB index, postoperative congestive cardiac failure, postoperative chest pain and postoperative cardiac complications. CTi elevation is not limited to patients who have obvious myocardial injury such as myocardial infarction, but is elevated in patients with other cardiac complications such as CCF and AF. Nine patients developed CCF and they all had associated elevations in cTi. Likewise, four patients developed AF and they all had cTi elevations. The release of cTi from damaged myocardium in patients with CCF and AF indicates that the likely aetiology of these cardiac complications is myocardial ischaemia. Operative stress associated with triple vessel coronary artery disease may lead to diffuse ischaemia. This could result in CCF or AF with a small enzyme leak. CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

Operative stress associated with single vessel coronary artery disease is more likely to lead to discrete ischaemia that could result in subendocardial MI or UAP. As would be expected, postoperative chest pain is significantly associated with cTi elevation. Chest pain should be taken seriously in vascular patients and full investigations for myocardial ischaemia and other cardiac complications performed. CKMB index is the cardiac enzyme most associated with cTi elevation. Therefore, if cTi is unavailable, greatest reliance should be placed on CKMB index, rather than CK or CKMB, when trying to interpret cardiac enzyme rises in the postoperative setting. Unlike other studies, the findings failed to demonstrate any preoperative or intraoperative factors that were significantly associated with cTi elevation. Goldman et al. [41] identified nine factors associated with adverse cardiac outcome. Mangano et al. [42] demonstrated that Atenolol could decrease cardiac morbidity following vascular surgery. Landesberg et al. [43] found that ST segment depression and voltage criteria for left ventricular hypertrophy on the preoperative ECG correlated with adverse cardiac outcome on multivariate analysis. Although peak cTi correlated significantly with postoperative acute renal dysfunction on bivariate analysis, there was no significant association on multivariate analysis. In the patients discussed it is likely that the acute renal dysfunction was secondary to or coexistent with the cardiac complications observed. It is extremely unlikely that acute renal dysfunction was the cause of the cTi elevation in those patients. Multiple studies have shown that cTi is not elevated in patients with renal failure, although cardiac troponin T can be [44–46]. Postoperative anaemia seemed to be associated with elevated cTi but it did not reach statistical significance. Other investigators have found anaemia to be related to postoperative cardiac complications. Nelson et al. [47] showed that an haematocrit of ⬍28% following vascular surgery was significantly associated with myocardial ischaemia and morbid cardiac events. Hogue et al. [48] produced similar results in patients following radical prostatectomy. Cardiac investigations were only performed occasionally in patients with cardiac complications or cTi elevations. 9/20 (45%) patients had no cardiac investigations at all. 8/20 (40%) patients had noninvasive investigations. None of these patients proceeded to coronary angiography, even if the noninvasive investigations were abnormal. All three (15%) patients who did have coronary angiography had significant, correctable coronary artery disease. It seems that there is a failure to appreciate the high frequency of coronary artery disease in vascular patients and to investigate cardiac complications 263

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al.

adequately. Hertzer et al. [49] performed coronary angiograms on 1000 patients under consideration for peripheral vascular surgery. He showed that 25% had severe, correctable coronary artery disease. This study was limited because there was no access to angiography and echocardiography to examine myocardial function and coronary arteries of all patients pre and postoperatively. Although large cTi leaks have been shown to be associated with new regional wall motion abnormalities on echocardiography, it is unknown if minor cTi leaks are associated with significant coronary artery disease or new wall motion abnormalities. It remains to be shown whether medical treatment or surgical correction of coronary artery disease would prevent myocardial ischaemia and improve short and long term survival.

4.

5.

6.

7.

8.

9.

Conclusions The high frequency of cardiac complications in vascular surgery patients warrants monitoring to detect myocardial ischaemia and other cardiac complications. CTi elevation occurs in patients with all cardiac complications, not just myocardial infarction. Routine monitoring of cTi would be an effective and inexpensive way to detect cardiac complications. However, it is probably not as sensitive for myocardial ischaemia as continuous electrocardiographic monitoring is. CKMB and CKMB index have a high degree of specificity for cardiac complications, but they are not sensitive enough to detect cardiac complications other than myocardial infarction. ECG changes occur frequently in patients following vascular surgery. Further studies will be required to ascertain the importance of ECG changes following vascular surgery. CTi elevation was predicted by peak CKMB index, postoperative congestive cardiac failure, postoperative chest pain and postoperative cardiac complications on linear regression analysis. Chest pain in vascular patients should be taken seriously and investigated thoroughly. Cardiac complications tend to be under investigated in vascular patients. It remains to be shown whether minor cTi leaks are associated with significant coronary artery disease or new wall motion abnormalities on echocardiography.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19. 20.

References 1. Jamieson, W. R. E., Janusz, M. T., Miyagishima, R. T. et al, Influence of ischaemic heart disease on early and late mortality after surgery for peripheral vascular disease. Circulation, 1982, 66, (Suppl 1) I92–I97. 2. Ouyang, P., Gerstenblith, G., Furman, W. R. et al, Frequency and significance of early postoperative silent myocardial ischaemia in patients having peripheral vascular surgery. American Journal of Cardiology, 1989, 64, 1113–1116. 3. Goldman, L., Caldera, D. L., Nussbaum, S. R. et al, Multifac-

264

21.

22.

torial index of cardiac risk in noncardiac surgical procedures. NEJM, 1977, 297, (16) 845–850. Yeager, R. A., Moneta, G. L., Edwards, J. M. et al, Reducing perioperative myocardial infarction following vascular surgery. The potential role of beta-blockade. Archives of Surgery, 1995, 130, 869–873. Raby, K. E., Goldman, L., Cook, F. et al, Long-term prognosis of myocardial ischaemia detected by holter monitoring in peripheral vascular disease. American Journal of Cardiology, 1990, 66, 1309–1313. Berlatzky, Y., Landesberg, G., Anner, H. et al, Prolonged postoperative myocardial ischaemia and infarction in vascular surgery performed under regional anaesthesia. European Journal of Vascular Surgery, 1994, 8, 413–418. Mangano, D. T., Browner, W. S., Hollenberg, M. et al, Association of perioperative myocardial ischaemia with cardiac morbidity and mortality in men undergoing non cardiac surgery. NEJM, 1990, 323, (26) 1781–1788. Goldman, L., Caldera, D. L., Nussbaum, S. R. et al, Multifactorial index of cardiac risk in noncardiac surgical procedures. NEJM, 1977, 297, (16) 845–850. Landesberg, G., Einav, S., Christopherson, R. et al, Perioperative ischaemia and cardiac complications in major vascular surgery: importance of the preoperative twelve-lead electrocardiogram. Journal of Vascular Surgery, 1997, 26, (4) 570–577. Ouyang, P., Gerstenblith, G., Furman, W. R. et al, Frequency and significance of early postoperative silent myocardial ischaemia in patients having peripheral vascular surgery. American Journal of Cardiology, 1989, 64, 1113–1116. Force, T., Kemper, A. J., Bloomfield, P. et al, Non-Q wave perioperative myocardial infarction: assessment of the incidence and severity of regional dysfunction with quantitative two-dimensional echocardiography. Circulation, 1985, 72, (4) 781–789. Adams, J. E., Sigard, G. A., Allen, B. T. et al, Diagnosis of perioperative myocardial infarction with measurement of cardiac troponin-I. NEJM, 1994, 330, (10) 670–674. Lee, T. H. and Goldman, L., Serum enzyme assays in the diagnosis of acute myocardial infarction. Annals of Internal Medicine, 1986, 105, 221–233. Ingwall, J. S., Kramer, M. F. and Fifer, M. A., The creatine kinase system in normal and diseased human myocardium. NEJM, 1985, 313, 1050–1054. Wu, A. H. B., Feng, Y. J., Contois, J. H. et al, Comparison of myoglobin, creatine kinase-MB and cardiac troponin-I for diagnosis of acute myocardial infarction. Annals of Clinical and Laboratory Science, 1996, 105, 221–233. Lee, T. H. and Goldman, L., Serum enzyme assays in the diagnosis of acute myocardial infarction. Annals of Internal Medicine, 1986, 105, 221–233. Mair, J., Cardiac troponin I and troponin T: are enzymes still relevant as cardiac markers? Clinica Chimica Acta, 1997, 257, 99–115. Lee, T. H. and Goldman, L., Serum enzyme assays in the diagnosis of acute myocardial infarction. Annals of Internal Medicine, 1986, 105, 221–233. Perry, S. V., The regulation of contractile activity in muscle. Biochemistry Society Transactions, 1979, 7, 593–617. Bodor, G. S., Porterfield, D., Voss, E. M. et al, Cardiac Troponin-I is not expressed in fetal and healthy or diseased adult human skeletal muscle tissue. Clinical Chemistry, 1995, 41, (12) 1710–1715. Apple, F. S., Rogers, M. A., Casal, D. C. et al, Creatine kinaseMB isoenzyme adaptations in stressed human skeletal muscle of marathon runners. Journal of Applied Physiology, 1985, 59, (1) 149–153. McLaurin, M. D., Apple, F. S., Voss, E. M. et al, Cardiac troponin-I, cardiac troponin-T and creatine kinase MB in dialysis patients without ischaemic heart disease: evidence of cardiac troponin-T expression in skeletal muscle. Clinical Chemistry, 1997, 43, (6) 976–982.

CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

Postoperative cardiac troponin-I (cTi) detection of ischaemia: N. Andrews et al. 23. Adams, J. E., Bodor, G. S., Davila-Roman, V. G. et al, Cardiac Troponin-I: a marker with high specificity for cardiac injury. Circulation, 1993, 88, 101–106. 24. Wilkinson, J. M. and Grand, R. J. A., Comparison of amino acid sequence of troponin I from different striated muscles. Nature, 1978, 271, 31–35. 25. Adams, J. E., Bodor, G. S., Davila-Roman, V. G. et al, Cardiac Troponin-I: a marker with high specificity for cardiac injury. Circulation, 1993, 88, 101–106. 26. Apple, F. S., Falahati, A., Paulsen, P. R. et al, Improved detection of minor ischaemic myocardial injury with measurement of serum cardiac troponin-I. Clinical Chemistry, 1997, 43, (11) 1342–1349. 27. Antman, E. M., Tanasijevic, M. J., Thompson, B. et al, Cardiac specific troponin-I levels to predict the mortality in patients with acute coronary syndromes. NEJM, 1996, 335, 1342–1349. 28. Trinquier, S., Flecheux, O., Bullenger, M. et al, Highly specific immunoassay for cardiac troponin I assessed in noninfarct patients with chronic renal failure or severe polytrauma. Clinical Chemistry, 1995, 41, (11) 1675–1676. 29. Musso, P., Cox, I., Vidano, E. et al, Cardiac troponin elevations in chronic renal failure: Prevalence and clinical significance. Clinical Biochemistry, 1999, 32, (2) 125–130. 30. Hafner, G., Thome-Kromer, B., Schaube, J. et al, Cardiac troponins in serum in chronic renal failure. Clinical Chemistry, 1994, 40, 1790–1791. 31. Galvani, M., Ottani, F., Ferrini, D. et al, Prognostic influence of elevated levels of cardiac Troponin-I in patients with unstable angina. Circulation, 1996, 95, (8) 2053–2059. 32. Raby, K. E., Goldman, L., Creager, M. A. et al, Correlation between preoperative ischaemia and major cardiac events after peripheral vascular surgery. NEJM, 1989, 321, (19) 1296–1300. 33. Mangano, D. T., Browner, W. S., Hollenberg, M. et al, Association of perioperative myocardial ischaemia with cardiac morbidity and mortality in men undergoing non cardiac surgery. NEJM, 1990, 323, (26) 1781–1788. 34. Adams, J. E., Sigard, G. A., Allen, B. T. et al, Diagnosis of perioperative myocardial infarction with the measurement of cardiac troponin-I. NEJM, 1994, 330, (10) 670–674. 35. Tabachnick, B. G. and Fidell, L. S., Using Multivariate Statistics, 2nd edn, Harper Collins, Northridge, 1989. 36. Frank, S. M., Beattie, C., Christopherson, R. et al, Unintentional hypothermia is associated with postoperative myocardial ischaemia. Anesthesiology, 1993, 78, 468–476. 37. Landesberg, G., Einav, S., Christopherson, R. et al, Perioperative ischaemia and cardiac complications in major vascular sur-

CARDIOVASCULAR SURGERY

JUNE 2001 VOL 9 NO 3

38. 39.

40. 41. 42. 43.

44.

45. 46. 47.

48.

49.

gery: importance of the preoperative twelve-lead electrocardiogram. Journal of Vascular Surgery, 1997, 26, (4) 570–578. Lee, T. H. and Goldman, L., Serum enzyme assays in the diagnosis of acute myocardial infarction. Annals of Internal Medicine, 1986, 105, 221–233. Wu, A. H. B., Feng, Y. J., Contois, J. H. et al, Comparison of myoglobin, creatine kinase-MB and cardiac troponin-I for diagnosis of acute myocardial infarction. Annals of Clinical and Laboratory Science, 1996, 105, 221–233. Lee, T. H. and Goldman, L., Serum enzyme assays in the diagnosis of acute myocardial infarction. Annals of Internal Medicine, 1986, 105, 221–233. Goldman, L., Caldera, D. L., Nussbaum, S. R. et al, Multifactorial index of cardiac risk in noncardiac surgical procedures. NEJM, 1977, 297, (16) 845–850. Mangano, D. T., Layug, E. L., Wallace, A. et al, Effect of atenolol on mortality and cardiovascular morbidity after noncardiac surgery. NEJM, 1996, 335, (23) 1713–1720. Landesberg, G., Einav, S., Christopherson, R. et al, Perioperative ischaemia and cardiac complications in major vascular surgery: importance of the preoperative twelve-lead electrocardiogram. Journal of Vascular Surgery, 1997, 26, (4) 570–577. Trinquier, S., Flecheux, O., Bullenger, M. et al, Highly specific immunoassay for cardiac troponin I assessed in noninfarct patients with chronic renal failure or severe polytrauma. Clinical Chemistry, 1995, 41, (11) 1675–1676. Musso, P., Cox, I., Vidano, E. et al, Cardiac troponin elevations in chronic renal failure: prevalence and clinical significance. Clinical Biochemistry, 1999, 32, (2) 125–130. Hafner, G., Thome-Kromer, B., Schaube, J. et al, Cardiac troponins in serum in chronic renal failure. Clinical Chemistry, 1994, 40, 1790–1791. Nelson, A. H., Fleisher, L. A. and Rosenbaum, S. H., Relationship between postoperative anaemia and cardiac morbidity in high-risk vascular patients in the intensive care unit. Critical Care Medicine, 1993, 21, (6) 860–866. Hogue, C. W., Goodnough, L. T. and Monk, T. G., Perioperative myocardial ischaemic episodes are related to haematocrit level in patients undergoing radical prostatectomy. Transfusion, 1998, 38, 924–930. Hertzer, N. R., Beven, E. G., Young, J. R. et al, Coronary artery disease in peripheral vascular patients. A classification of 1000 coronary angiograms and results of surgical management. Annals of Surgery, 1984, 199, (2) 223–233.

Paper accepted 16 October 2000

265