The relationship between ASAT, CKMB, Troponin-T and mortality after cardiac surgery

Abstract

Objective. To investigate the relationship between ASAT, CKMB, Troponin-T and mortality after cardiac surgery. Design. ASAT, CKMB and TnT were analysed in 116 patients. Correlation, sensitivity, specificity and predictive values for permanent myocardial injury (defined as TnT ≥2.0 µg/l postoperative day four) were calculated. In the second part our clinical protocol (ASAT on postoperative day 1 and TnT on day 3–4 in patients with ASAT above 2.5 µkat/l) was evaluated. Mortality was compared between patients with ASAT <2.5 µkat/l (ASAT-), 2. ASAT ≥2.5 µkat/l and TnT <2.0 µg/l (ASAT+/TnT-) and 3. ASAT ≥2.5 µkat/l and TnT ≥2.0 µg/l (ASAT+/TnT+). Results. Both ASAT and CKMB had irrespectively of cut-off level, low positive and high negative predictive value of permanent myocardial injury. Early and mid-term mortality did not differ significantly between ASAT- and ASAT+/TnT- patients. Conclusions. ASAT and CKMB can be used to exclude but not to diagnose permanent myocardial injury after cardiac surgery. Increased postoperative ASAT in the absence of increased TnT is not associated with worse clinical outcome than after normal postoperative ASAT.

• Myocardial injury
• cardiac surgery
• biomarkers

Perioperative myocardial injury (PMI) is a major cause of morbidity and mortality after cardiac surgery 1. The reported incidence varies between 2% and 80% depending on diagnostic criteria 1, 2. There is no consensus of definition for PMI 2, 3 in contrast to general acute myocardial infarction (AMI), where a combination of symptoms, electrocardiogram (ECG), and cardiac biomarkers, is used for diagnosis 4, 5.

Early detection of PMI is essential to give the patients appropriate postoperative treatment 6, 7. ECG changes alone, is not a reliable criterion of PMI and needs to be combined with biomarkers 1. However, the operative trauma causes an unspecific release of cardiac biomarkers, interfering with sensitivity and specificity early after surgery 1, 8. Aspartate aminotransferase (ASAT) and creatine kinase isoenzyme MB (CKMB) are traditional biomarkers of both AMI and PMI 7. They act as catalysers, ASAT in the forming of aspartate, a source of nitrogen in the urea cycle, and CKMB in the delivery of phosphate to ATP 9. The traditional biomarkers reach their peak value on the first day after ischemia and remains in plasma for 24–36 hours 10. The new generation of cardiac biomarkers, the troponins (TnT and TnI), which regulate calcium mediated muscle contraction 2, have a later peak value with markedly elevated levels for several days after PMI 7. With the troponins, sensitivity and specificity of PMI have improved 2, 4–7, 10–12. However, it should be noted that renal function influence troponin levels, especially TnT, with false high levels in patients with impaired renal function 11.

If troponin level 3 – 4 days after surgery is used as a marker for permanent PMI, as previously suggested 1, 11, 12, it might be combined with an early screening marker for earlier detection. Consequently, the first aim was to assess the correlation between two early biomarkers of myocardial injury (ASAT and CKMB) and TnT measured day 3–4 after surgery. In addition we wanted to investigate whether an increase in an early screening marker (ASAT), in the abscence of permanent myocardial injury, is associated with impaired outcome after cardiac surgery.

Material and methods

Patients

In the first part of the study, 116 consecutive patients (77% men, mean age 68±9 years) undergoing various elective first-time cardiac surgery procedures at our institution were included in a prospective study. Patients with known renal disease were excluded.

In the second part of the study, 2 751 patients (74% men, mean age 66±10 years) undergoing first time elective cardiac surgery were included in a retrospective study. Patient characteristics are presented in Table I. Pre- and perioperative data was prospectively collected in an institutional database.

Table I.  Patient characteristics for all patients included in the two studies. Number (%) or mean±SD.

Variable	Substudy 1	Substudy 2
Number of patients	116	2 751
Male sex	77%	74%
Age (years)	68±9	66±10
Body mass index	27.0±3.7	26.8±3.9
Higgins risk score	2.5±2.3	2.3±2.2
Diabetes mellitus	14%	19%
Left ventricular ejection fraction (%)	55±12%	55±10%
Extracorporal circulation time (minutes)	89±51	84±42
Aorta clamping time (minutes)	57±40	55±12
CABG	66%	72%
Valve surgery	12%	13%
CABG + valve surgery	9%	7%
Other cardiac surgery	13%	8%

Key: CABG = Coronary Artery Bypass Grafting, SD = standard deviation.

Study design

In the first substudy, ASAT and CKMB were analysed the first postoperative day and TnT was analysed the fourth postoperative day. Permanent PMI was defined as TnT > 2.0 µg/l on postoperative day 3–4. The correlations between the different biomarkers were determined and sensitivity, specificity, positive and negative predictive values were calculated with different cut-off levels.

In the second substudy our present clinical protocol for detection of permanent PMI was evaluated. The protocol includes ASAT measurement on the first postoperative day in all patients and TnT measurement on the third or fourth postoperative in all patients with ASAT above 2.5 µkat/l on day 1. The association between ASAT and TnT was calculated. To analyze clinical outcome in substudy 2, the patients were divided into three groups; those with postoperative ASAT below a cut-off level of 2.5 µkat/l (ASAT-, n = 2566), those with ASAT above ≥2.5 µkat/l and normal TnT <2.0 µg/l (ASAT + /TnT-, n = 84), and those with ASAT ≥2.5 µkat/l and TnT ≥2.0 µg/l (ASAT + /TnT+, n = 101). Thirty-day mortality and cumulative mid-term survival were compared between the three groups. Survival data was collected from the Swedish Civil Register. Mean follow-up was 17 months (range 0–32 months) and 100% complete.

Biochemical analyses

The biomarkers were analysed with clinical standard methods. ASAT activity was measured with photometry, while TnT and CKMB were analysed with electrochemiluminiscence immunoassay (ECLIA).

Statistics

The data is generally presented as the mean and standard deviation. Spearman's rank sum test was used to calculate correlation between the different biomarkers of myocardial injury. Student's T-test was used to compare continuous data between two groups and ANOVA for three-group comparisons. χ² test was used to compare categorical data. Cumulative midterm survival was calculated according to Kaplan-Meier and compared between the groups with log-rank test. A p-value of <0.05 was considered significant.

Results

Substudy 1

Incidence of permanent myocardial injury

Nine of the 116 patients (7.8%) had a permanent myocardial injury, defined as a TnT >2.0 µg/l on the fourth postoperatively day.

Correlations between biomarkers

There was a significant correlation between the different biomarkers; CKMB vs. ASAT r = 0.79, p < 0.001; CKMB vs. TnT r = 0.84, p < 0.001 and ASAT vs. TnT r = 0.75, p < 0.001.

Sensitivity, specificity, positive and negative predictive values

Sensitivity, specificity, positive and negative predictive values were comparable for ASAT and CKMB (Table II). The highest positive predictive value of ASAT (60%) was achieved with a cut-off level of 3.0 µkat/l. However with this cut-off level, the sensitivity obtained was only 67%, while a cut-off level of 2.5 µkat/l, presented a similar positive predictive value (56%), together with a maximal sensitivity. The highest positive predictive value of CKMB (70%) was received with a cut-off level of 100 µg/l, with a sensibility equal to a cut-off level of 75 µg/l.

Table II.  Sensitivity, specificity, positive and negative predictive values for permanent myocardial injury (defined as Troponin-T ≥2.0 µg/l postoperative day 4) of ASAT and CKMB with different cut-off levels.

Biomarker/ Cut-off level	Sensitivity	Specificity	Positive predictive value	Negative predictive value
ASAT
2.0 µkat/l	100%	82%	32%	100%
2.5 µkat/l	100%	93%	56%	100%
3.0 µkat/l	67%	96%	60%	97%
CKMB
50 µg/l	89%	84%	32%	99%
75 µg/l	78%	91%	41%	98%
100 µg/l	78%	97%	70%	98%

Key: ASAT = aspartate aminotransferase, CKMB = creatine kinase isoenzyme MB, PMI = peri-operative myocardial injury.

Substudy 2

ASAT vs. TnT

There was a significant correlation between postoperative ASAT and TnT (r = 0.36, p < 0.001). Only 55% (101/185) of the patients with ASAT ≥2.5 µkat/l had TnT ≥2.0 µg/l.

Outcome

Baseline characteristics of the three groups are given in Table III. Thirty-day mortality was significantly higher in the ASAT + /TnT+ group (9.0%) than in the ASAT- group (0.8%) and the ASAT + /TnT- group (1.2%), p < 0.001 and p = 0.012 respectively (Figure 1). ASAT- and ASAT + /TnT- groups did not differ significantly (p = 0.68). Mid-term survival (Figure 2) was inferior in the ASAT + /TnT+ group with 86% at 2 years, compared to 96% in the ASAT- and 97% in the ASAT + /TNT- group, p < 0.001 and p = 0.010, respectively. Again, the ASAT- and ASAT + /TnT- groups did not differ significantly (p = 0.87).

Figure 1.  Thirty-day mortality in the ASAT- group, ASAT + /TnT- group and ASAT + /TnT+ group. There were significant differences between ASAT + /TnT+ and ASAT- (p < 0.001) and between ASAT + /TnT+ and ASAT + /TnT- (p = 0.012).

Figure 2.  Cumulative mid-term survival (Kaplan-Meier) in the three ASAT groups (ASAT- solid line, ASAT + /TnT- dashed line and ASAT + /TnT+ dotted line). There were significant differences between ASAT + /TnT+ and ASAT- (p < 0.001) and between ASAT + /TnT+ and ASAT + /TnT- (p = 0.010).

Table III.  Patient characteristics for the three ASAT groups in substudy 2. Number (%) or mean±SD.

Variable	ASAT-	ASAT + /TnT-	ASAT + /TnT +	p-value
Number of patients	2566	84	101
Male sex	75%**	61%	67%	0.003
Age (years)	66±10^†	66±10^†	67±10	0.04
Body mass index	27±4	27±4	26±3	0.62
Higgins risk score	2.2±2.2*^,†††	2.7±2.1^†††	4.5±3.1	<0.001
Diabetes mellitus	19%^††	12%	9%	0.010
Left ventricular ejection fraction (%)	56±12	57±14	55±12	0.33
Extracorporal circulation time (minutes)	82±41***^,†††	112±44	119±54	<0.001
Aorta clamping time (minutes)	53±43***^,†††	75±39	72±39	<0.001
CABG	74%***^,†††	37%^††	59%	<0.001
Valve surgery	12%*	20%	17%	0.026
CABG + valve surgery	6%***	26%^†	11%	<0.001
Other cardiac surgery	8%**	17%	13%	0.003

Key: ASAT = aspartate aminotransferase, CABG = Coronary Artery Bypass Grafting, SD = standard deviation, Asterisks denote statistically significant differences between the ASAT- and ASAT + /TnT- groups: * = p < 0.05 vs. ASAT + /TnT- ** = p < 0.01 vs. ASAT + /TnT-, *** = p < 0.001 vs. ASAT + /TnT-; Crosses denote statistically significant differences between the ASAT- and ASAT + /TnT- groups:^†=p< 0.05 vs. ASAT + /TnT+, ^††=p < 0.01 vs. ASAT + /TnT+, ^†††=p < 0.001 vs. ASAT + /TnT+.

Discussion

The main findings in the present study are as follows: First, ASAT and CKMB are comparable early screening markers for permanent myocardial injury after cardiac surgery. Both markers have high negative but low positive predictive value for permanent PMI, defined as TnT > 2.0 µg/l. Secondly, increased ASAT in the absence of definitive myocardial injury, is not associated with inferior outcome than after normal ASAT.

In substudy 1, we compared two early screening biomarkers for PMI after cardiac surgery. An optimal biomarker for PMI needs to be inexpensive, cardiospecific and released in proportion to the magnitude of the permanent myocardial injury. Furthermore, the biomarker needs to be released rapidly after the injury and then persist in plasma in a diagnostic time window 2. Traditional markers for myocardial injury, such as ASAT and CKMB, have been used after surgery but have low specificity due to early unspecific release, i.e. diagnostic noise 8. More recently studies have been published which suggest that troponins may improve sensibility and specificity in the diagnosis of permanent myocardial injury 2, 4–7, 10. The troponins have later peak values than traditional biomarkers, with increased plasma levels for several days 7. However, troponins are not spared from unspecific release early after surgery and are more expensive to analyze 8. Thus the troponins are suitable as a late marker, preferably analysed on the third or fourth postoperative day, when interfering diagnostic noise can be avoided 1, 11, 12. Accordingly, in the first part of the study we investigated the association between two traditional inexpensive markers and permanent myocardial injury after cardiac surgery, defined as TnT > 2.0 µg/l on the fourth postoperative day. ASAT and CKMB were comparable early screening markers, which confirms previous observations 3. We found that irrespectively of marker and cut-off level there was a high negative predictive value and a low positive predictive value of ASAT and CKMB for permanent myocardial injury. Therefore, an increase in ASAT or CKMB above the cut-off levels is not conclusive for a permanent myocardial injury and there will be a considerable amount of false positives. On the other hand, the absence of an increased ASAT or CKMB virtually excludes the presence of a permanent myocardial injury, which may be regarded as the primary objective of an early screening marker.

According to substudy 1, ASAT was a somewhat better screening marker of PMI than CKMB. ASAT with a cut-off level of 2.5 µkat/l presented the most favourable combination of sensitivity and specificity, 100% and 93% respectively (Table II), compared to the best combination of sensitivity and specificity of CKMB, obtained with a cut-off level of 100 µkat/l, with a sensitivity of only 78%, and a specificity of 97%. Consequently, substudy 1 resulted at our institution in changed routines concerning PMI diagnostics at our institution. Now ASAT is used as screening marker, analysed on postoperative day one in all patients and TnT as late marker, analysed postoperative day three or four on patients with suspected PMI (ASAT > 2.5 µkat/l).

In substudy 2 we evaluated this clinical routine in a substantially larger patient population. A similar positive predictive value for ASAT (with a cut-off level of 2.5 µkat/l) as in substudy 1 was obtained (55% vs. 56%, respectively) but negative predictive value, sensitivity and specificity could not be calculated because TnT was analysed only in patients where the early screening marker indicated a suspected perioperative myocardial injury (ASAT > 2.5 µkat/l). However, since the results in substudy 1 indicated a 100% negative predictable value with ASAT > 2.5 µkat/l, it is plausible that no perioperative myocardial injury was missed.

A secondary objective with substudy 2 was to investigate outcome in patients with increased ASAT in the absence of permanent PMI. If outcome in this subgroup of patients were inferior to the patients with normal postoperative ASAT it would be important to identify risk factors and potential treatment options. However, elevated ASAT in the absence of elevated TnT was not associated with significantly impaired early and mid-term mortality (Figures 1 and 2) despite differences in risk profile. Interestingly, the increased midterm mortality in the ASAT + /TnT+ group was almost entirely an effect of an increased early mortality as demonstrated in Figure 2. This indicates that if patients with a permanent PMI survive the immediate postoperative period is mid-term survival comparable to the one achieved in patients without PMI.

It is obvious from Table III that the patients with increased ASAT (but normal TnT) postoperatively differed in several important baseline variables from the patients with normal postoperative ASAT. The proportion of patients undergoing valve surgery and combined valve surgery and CABG was larger in the increased ASAT group. In addition, patients with increased ASAT had higher Cleveland Clinic Risk score, longer extracorporeal circulation (ECC) time and aortic cross clamp time. The importance of operation time is illustrated by a positive correlation between postoperative ASAT concentration and both ECC time (r = 0.44, p < 0.001) and aortic clamp time (r = 0.37, p < 0.001). Furthermore, in the subgroup of patients undergoing isolated valve surgery was the positive predictable value of ASAT >2.5 µkat/l only 39%, demonstrating that almost 2/3 of valve patients with ASAT above the cut-off level did not had a permanent myocardial injury.

The present study has important limitations. The optimal cut-off level of TnT and the optimal time point after surgery for detection of permanent myocardial injury is controversial. Different cut-off levels have been used in different studies; Metzler et al. used a TnT cut-off level of 0.6 µg/l day one after surgery 11 while Carrier et al. used a cut-off level of 3.4 µg/l 2 days after surgery 12. However, in an addendum to Carrier et al.'s study, a cut-off level of 1.0 µg/l 36–48 h after surgery was suggested by the same authors 12. In our studies we chose after a literature review to use 2.0 µg/l on day three or four as cut-off level. However, this approach can be debated and it would be desirable to have a consensus on cut-off levels and timepoints regarding TnT after cardiac surgery. Another limitation is that in substudy 2, we only analyzed TnT levels in patients with ASAT above 2.5 µkat/l. Hence, we could not calculate sensitivity, specificity and predictive values in this considerably larger patient population.

In conclusion, ASAT and CKMB can be used to exclude but not to diagnose permanent myocardial injury. Elevated ASAT postoperatively in the absence of increase in troponins, is not associated with worse clinical outcome than after a normal postoperative ASAT.

The abstract was presented at the Scandinavian Association of Thoracic Surgery's Annual Meeting in Reykjavik, August 17–19, 2006