Serial and single time-point measurements of cardiac troponin T for prediction of clinical outcomes in patients with acute ST-segment elevation myocardial infarction
- First Online:
- Cite this article as:
- Kurz, K., Schild, C., Isfort, P. et al. Clin Res Cardiol (2009) 98: 94. doi:10.1007/s00392-008-0727-9
- 98 Views
Cardiac troponins are the preferred biomarkers to predict infarct size in patients (pts) after acute myocardial infarction (AMI). Less information is currently available to verify the prognostic value of such a biomarker surrogate.
We included 82 pts with acute STEMI and compared all single time point and serial cardiac troponin T (cTnT) values (peak and area-under-the-curve) from admission until day 4 to predict future major adverse cardiac events (MACE).
Pts who had suffered any MACE during follow-up had higher cTnT values (median (25th/75th percentiles) on day 4 (3.16 µg/l (2.71/5.20) Vs. 2.1 µg/l (1.19/3.96), P = 0.0304), and higher peak cTnT values (5.11 µg/l (3.31/9.47) Vs. 2.92 µg/l (1.81/5.63), P = 0.0234). The likelihood to develop a composite of MACE was twofold higher in the intermediate cTnT tertile (1.66–3.04 µg/l, n = 23), and in the upper cTnT tertile (3.35–20.68 µg/l, n = 23) for cTnT on day 4. For cTnT peak the risk was 1.7-fold higher in the intermediate cTnT peak tertile (2.55–5.01 µg/l, n = 28) and 2.4-fold in the upper cTnT peak tertile (5.11–18.93 µg/l, n = 27). The optimal ROC cutoff for cTnT to predict the composite of MACE was 2.69 µg/l measured on day 4 and 2.85 µg/l for the cTnT peak.
A single measurement of cTnT after STEMI is an independent predictor for MACE, performs as effective as serial cTnT sampling and may be useful to assess future events.
Keywordscardiac troponin TST-segment elevation myocardial infarctionoutcome
After acute myocardial infarction (AMI) infarct size is a strong determinant of prognosis . Several imaging techniques such as cardiac magnetic resonance imaging, thallium sestamibi, or echocardiography are commonly used to describe infarct size and left ventricular function [6, 12, 15]. However, the use of these imaging techniques is limited due to restricted availability and high costs. Therefore, estimating infarct size from concentrations or activities of cardiac proteins in peripheral blood, as has been used for years, is a convenient alternative . As undoubtedly substantiated, cardiac troponins are the preferred biomarkers to predict infarct size in patients with AMI . However, less information is currently available to verify the prognostic value of such a biomarker surrogate.
Therefore, this study sought to determine the best sampling method and timing, i.e. single time-point versus serial sampling (single time-point, peak and area-under-the-curve) of cardiac troponin T (cTnT), to predict development of major adverse cardiac events (MACE) in patients after acute ST-segment elevation myocardial infarction (STEMI). Additionally, results were compared with the conventionally measured risk markers CK, CKMB, CRP and NT-pro BNP.
In the Chest Pain Unit of the University of Heidelberg, 156 consecutive patients with AMI who received primary PCI within 12 h after the onset of symptoms were screened. Only patients with STEMI were included while, based on previous findings patients with non-STEMI were excluded from further analysis due to inferior performance of serologic estimation of infarct size owing to heterogeneous spectrum of non-STEMI (NSTEMI) including large strictly posterior MI and microinfarction . STEMI was defined by the presence of ST segment elevation ≥0.1 mV in unipolar leads or elevation ≥0.2 mV in anterior wall leads and was retrospectively confirmed by a rise of cTnT above 0.03 µg/l measured on at least one occasion within 24 h after the ischemic index event with a subsequent fall.
Blood samples were taken at admission and daily for 96 h. After collection, blood samples were centrifuged immediately and serum stored at −20°C until analysis. cTnT was measured on the same Elecsys analyzer using the 4th generation assay. The detection limit of the assay is 0.01 µg/l. The decision limit used for diagnosis of AMI is 0.03 µg/l, with an imprecision of less than 10%.
The laboratory staff responsible for measurements was blinded to the patient data. We measured single-point cTnT at 24, 48, 72 and 96 h. Peak and cumulative cTnT over 96 h (area-under-the-curve, cTnTAUC) were determined retrospectively from serial samples. A value was defined as peak if it was the highest in the concentration time course and if there was at least one lower value before and after this maximum value. Routinely, CK (activity), CKMB (mass), CRP and NT-pro BNP was measured and peak values for CK and CKMB were calculated analogue to cTnT.
All decisions regarding acute reperfusion therapy and clinical management of patients were left at the discretion of the attending cardiologist who was not involved in the study. The study protocol was approved by the local ethical committee of the University of Heidelberg and all patients gave informed consent before study enrolment.
Follow-up was done by telephone contact and hospital records. The major adverse cardiac events include cardiac and non-cardiac death, nonfatal myocardial infarction (MI) and target vessel revascularization (TVR), and the composite of MACE. Reperfusion success was defined as postprocedural TIMI 2 or 3 flow together with a residual stenosis <30%.
Means with standard deviation or medians with 25th and 75th percentiles were calculated to describe continuous variables. The Kolmogorov–Smirnov test was used to test for normal distribution. Correlations were calculated using Spearman rank correlation for continuous variables and Mann–Whitney U test was used for comparisons of non-parametric data. Serial measurements (peak value cTnT, cTnTAUC) have been performed to evaluate the differences between subgroups. MACE-free survival was plotted by Kaplan-Meier survival curves. ROC curves were constructed to test the relationship between biomarker values and cardiovascular events. For ROC analysis, patients were split into two groups, those with biomarker values above median and those below. A logistic regression analysis was performed to test for independent predictors. For all analyses, a P-value <0.05 was regarded as statistically significant. All statistical analyses were performed using MedCalc Version 9.2 for Windows (MedCalc Software, Mariakerke, Belgium).
Entire cohort (n = 82)
w/t MACE (n = 69)
With MACE (n = 13)
History of cardiac events
CAD risk factors
Entire cohort (n = 82)
w/t MACE (n = 69)
With MACE (n = 13)
Number of PCI
TIMI flow pre PCI
TIMI flow post PCI
A total of 314 samples were obtained from the 82 patients yielding a mean of 3.8 samples per patient. Median cTnT values with corresponding 25th/75th percentiles were 0.11 µg/l (0.01/0.55) at admission, 2.65 µg/l (1.42/5.63) on day 1, 2.87 µg/l (1.62/4.59) on day 2, 2.39 µg/l (1.59/4.59) on day 3, 2.34 µg/l (1.37/4.11) on day 4, 3.46 µg/l (2.02/5.79) for peak cTnT value, and 340.2 µg/l/96 h (222.5/615.3) for cTnTAUC.
cTnT in anterior versus non-anterior STEMI
Patients with anterior AMI had significantly higher cTnT levels than non-anterior AMI at admission (0.36 µg/l (0.05/0.95) Vs. 0.04 µg/l (0.01/0.33), P = 0.0148), on day 1 (3.10 µg/l (1.79/7.73) Vs. 2.41 µg/l (1.13/3.43), P = 0.0863), on day 2 (3.94 µg/l (2.41/5.83) Vs. 2.01 µg/l (0.86/3.48), P = 0.0023), on day 3 (3.65 µg/l (2.00/5.79) Vs. 2.02 µg/l (1.42/3.42), P = 0.0111), and on day 4 (2.71 µg/l (1.87/6.35) Vs. 2.05 µg/l (0.88/2.98), P = 0.0219). Likewise cTnT peak (4.45 µg/l (2.66/9.36) Vs. 2.80 µg/l (1.09/4.78), P = 0.0042), and cTnTAUC (492.11 (300.93/920.54) Vs. 282.16 (141.70/474.68), P = 0.0035) were higher in patients with anterior versus non-anterior AMI.
cTnT after successful versus unsuccessful or non-attempted PCI
Patients with successful reperfusion had statistically significant higher cTnT levels on day 3 (2.42 µg/l (1.72/5.39) Vs. 1.35 µg/l (0.08/3.29), P = 0.0343) and day 4 (2.41 µg/l (1.49/4.38) Vs. 0.91 µg/l (0.06/2.90), P = 0.0375) than those with non-attempted or unsuccessful PCI. No significant differences between successful and unsuccessful PCI were found at admission (0.11 µg/l (0.01/0.55) Vs. 0.155 µg/l (0.05/0.69), P = 0.5367), on day 1 (2.66 µg/l (1.48/5.24) Vs. 2.19 µg/l (0.80/6.32), P = 0.6591), day 2 (2.93 µg/l (1.68/5.13) Vs. 1.81 µg/l (0.49/3.49), P = 0.1218), peak cTnT (3.54 µg/l (2.37/5.78) Vs. 1.04 µg/l (0.21/6.32), P = 0.0586) and cTnTAUC (350.81 (245.57/643.36) Vs. 266.98 (13.16/527.14), P = 0.1551).
cTnT in early (<3 h) versus late reperfusion
Patients with very early reperfusion (<3 h) had statistically significant lower cTnT levels at admission (0.01 µg/l (0.01/0.08) Vs. 0.71 µg/l (0.14/2.57), P = 0.0005) than patients with late reperfusion (>12 h). However, no significant differences were found for cTnT on day 1 (2.42 µg/l (1.63/4.25) Vs. 1.14 µg/l (0.15/3.03), P = 0.1029), on day 2 (2.99 µg/l (1.72/5.35) Vs. 2.92 µg/l (0.47/3.94), P = 0.6278), on day 3 (2.52 µg/l (1.60/5.39) Vs. 2.63 µg/l (1.04/4.31), P = 0.5267), and on day 4 (1.51 µg/l (0.97/2.39) Vs. 2.69 µg/l (0.81/4.83), P = 0.2182). In addition cTnT peak (3.13 µg/l (1.78/5.40) Vs. 3.46 µg/l (1.14/6.00), P = 0.9249), and cTnTAUC (281.52 (208.02/396.62) Vs. 388.25 (35.54/477.88), P = 0.7867) were comparable in both groups.
Mean follow-up was 205 ± 157 days. During follow-up four patients died (4.9%), 3 patients (3.7%) suffered a non-fatal re-infarction, and another nine patients (11%) required TVR. A composite of any MACE occurred in 13 cases. First event occurred on day 6, last MACE during follow up period on day 280, mean 78.3 (±75.8) days. Patients who had suffered any cardiovascular event during follow-up had statistically significant higher cTnT values on day 4 (3.16 µg/l (2.71/5.20) Vs. 2.1 µg/l (1.19/3.96), P = 0.0304), and higher peak cTnT values (5.11 µg/l (3.31/9.47) Vs. 2.92 µg/l (1.81/5.63), P = 0.0234), and there was a trend to higher cTnTAUC (614.67 (337.10/662.06) Vs. 329.63 (205.97/541.37), P = 0.0677). No significant differences were found for cTnT at admission (0.07 µg/l (0.01/0.65) Vs. 0.11 µg/l (0.01/0.54), P = 0.8928), cTnT on day 1 (3.43 µg/l (1.78/6.07) Vs. 2.42 µg/l (1.38/5.44), P = 0.5109), on day 2 (3.45 µg/l (2.47/4.61) Vs. 2.85 µg/l (1.05/4.65), P = 0.2753) on day 3 (2.42 µg/l (1.96/5.49) Vs. 2.34 µg/l (1.57/4.45), P = 0.4109).
Logistic regression analysis for prediction of MACE-free survival
cTnT day 4≥ vs. <2.69 µg/l
Age ≥ vs. <median (60.4 years)
Reperfusion ≥ vs. <12 h
Successful PCI yes/no
There was a tendency towards impaired renal function measured by creatinine clearance in patients with MACE (76.56 (±31.26) ml/min × 1.73qm) vs. patients without MACE (97.30 (±37.47) ml/min × 1.73qm, P = 0.0758). There was no significant correlation of cTnT day 4, cTnT peak and cTnTAUC and creatinine clearance (P = 0.2157, P = 0.0625, P = 0.2669) and creatinine clearance was no independent risk factor for developing MACE (P = 0.3201, odds ratio 1.0115).
Prediction of MACE by early conventional biomarkers, markers of inflammation and cardiac function
Cardiac troponins (cTn) have an established role for diagnosis and risk stratification in acute coronary syndromes without ST-segment elevation, while the usefulness of cTn is less well established in STEMI. Previous reports have accumulated some evidence that a cTn already measurable at admission might carry a higher risk for unsuccessful reperfusion with either thrombolytic therapy or primary PCI . Moreover, an elevated cTn at admission has been related to adverse long-term outcomes even after successful primary PCI . More recently, cTnT measured 24, 48, 72 or 96 h after primary PCI for STEMI was found to closely correlate with infarct size that was quantified precisely by contrast-enhanced MRI . Given that prognosis after AMI is closely linked to infarct size it was tempting to speculate that the cTnT biomarker surrogate measured after AMI might carry prognostic information.
The present study supports the hypothesis that a single cTnT measured 96 h after onset of symptoms represents an independent predictor of long-term prognosis. A cTnT on day 4 above 2.69 µg/l was associated with a twofold higher risk for an adverse CV event. A second important finding of our study is that a single point measurement is as effective for prediction of long-term prognosis as the peak cTnT or the cumulative cTnT concentration obtained from serial sampling over 4 days. Comparable with other reports in the literature, in our cohort, a single measurement of cTnT on day 4 is superior to less specific commonly used markers of myocardial inflammation and necrosis as CK, CKMB, NT-pro BNP and CRP .
Experimental work has shown that the increases in serum troponin parallel the loss in tissue concentrations of troponin , and that blood values of cTnT correlate with histochemical infarct size in the non-reperfused experimental AMI [1, 10]. Interestingly, single-point measurements at 96 hours after LAD ligation were as effective as measurement of cumulative release requiring tight serial sampling over 120 h, or peak concentrations with retrospective location of the peaks, requiring at least three samples . Consistently, an increasing number of imaging studies has confirmed that the more convenient single-point cardiac troponin T or I values perform as well as cumulative release in humans [8, 9, 13]. SPECT or CE-MRI infarct size has been found to correlate well with single-point values of cTnT at approximately 72 h [8, 9] and 96 h  and with single-point values of cTnI between 24 and 72 h .
The prognostic hazard associated with higher cTnT values is most likely due to a larger infarct size and more severely depressed LV function as demonstrated in two previous MRI controlled trials [3, 13]. Both studies could demonstrate a direct relationship between the magnitude of cTnT and infarct size as well as LV function as measured by CE-MRI [3, 13].
Several variables have been reported to potentially influence infarct size and thus cTnT concentrations. Consistently, higher single and serial cTnT concentrations suggesting larger infarct size were measured in anterior AMI than in non-anterior infarction. Furthermore, patients with impaired renal function tend to more cardiovascular events. Similarly, in our cohort, there was a tendency towards worse renal function in patients developing MACE, however, there was no significant correlation of cTnT and renal function and creatinine clearance was no independent risk factor for developing MACE.
As expected, we were able to demonstrate that cTnT was higher in patients with successful compared to those without successful reperfusion at day 3 and 4, which could be explained by higher myocardial salvage in successful reperfusion. Still, median cTnT on day 4 in patients with successful reperfusion is lower than the ROC optimized cut off for cTnT day 4 to predict MACE and, likewise, median cTnT value on day 4 in any patient who had suffered a cardiovascular event during follow-up was higher. Successful reperfusion was not associated with higher MACE risk during follow-up period. Thus, a single cTnT measured preferentially on day 3 or 4 after AMI may be used as a surrogate endpoint that allows both an estimation of infarct size and prognosis. Such a surrogate biomarker endpoint is readily available, easy to measure and inexpensive and thus may help to circumvent the necessity of expensive cardiac imaging.
The finding that a single cTnT was at least as effective as peak cTnT or cumulative cTnT over 4 days that necessitate multiple serial measurements may improve the acceptance of such a biomarker endpoint in clinical routine. However, it should be noted that our results were obtained in patients with STEMI and must not be generalized to patients with NSTEMI. As we reported earlier , estimation of infarct size with cTnT is less accurate in NSTEMI most likely due to heterogeneity of infarct size, inclusion of strictly posterior infarction, and lower spatial resolution of CE-MRI at the very low end of infarct size . Moreover, it must be acknowledged that collection of cTnT was restricted to the first 4 days and that complete recovery of cTnT over 14–21 days may have resulted in better performance of serial cTnT measurements. Nevertheless, a single point-cTnT is easy and inexpensive and is more likely to gain acceptance into clinical practice.
A single measurement of cTnT after STEMI is an independent predictor for major clinical events and performs as effective as serial cTnT sampling. Therefore, a single cTnT measured on day 4 after STEMI may be useful for estimation of infarct size and assessment of future prognosis.