Available online at Received 10 March 2009

Transcription

1 Available online at Journal of Electrocardiology 42 (2009) Significance of T-wave amplitude and dynamics at the time of reperfusion in patients with acute ST-segment elevation myocardial infarction treated with primary percutaneous coronary intervention Jacob Thorsted Sørensen, MD, a, Marc Aaron Murinson, b Anne Kjer Kaltoft, MD, PhD, a Kjell Christer Nikus, MD, c Galen Strohm Wagner, MD, b Christian Juhl Terkelsen, MD, PhD a a Department of Cardiology B, Aarhus University Hospital, Skejby, Aarhus N, Denmark b Duke Clinical Research Institute, Duke University, Durham, NC, USA c Department of Cardiology, Heart Center, Tampere University Hospital, Tampere, Finland Received 10 March 2009 Abstract Keywords: Background: Peri-interventional T-wave changes may reflect the microvascular reperfusion status and potentially carry early independent, prognostic information in patients with ST-elevation myocardial infarction (STEMI) treated with primary percutaneous coronary intervention (PCI). Methods: The first available electrocardiogram (ECG) (index ECG) and the ECG recorded immediately post-pci were analyzed for T-wave morphology in 207 patients with STEMI. Absolute T-wave amplitude was recorded and any change in T-wave amplitude from index ECG to post-pci ECG was calculated. Continuous ST monitoring was performed from hospital arrival until 90 minutes after PCI. Maximum troponin level and left ventricular ejection fraction were evaluated before discharge. Final infarct size was assessed by myocardial perfusion imaging after 1 month. Results: Large, positive T-wave amplitude in the index ECG and the post-pci ECG was associated with delayed ST resolution after PCI. In the post-pci ECG, T-wave amplitude was positively associated with troponin-t value (P b.001) and final infarct size (P =.036), and inversely associated with left ventricular ejection fraction (P b.001). However, T-wave amplitude in the post-pci ECG was also associated with procedural increase in ST elevation (P b.001) and inversely associated with spontaneous ST resolution (P b.017). A net decrease in T-wave amplitude during reperfusion therapy was associated with faster microvascular reperfusion as evaluated by time to ST resolution. Conclusion: Large T-wave amplitudes in static pre- and post-pci ECGs are associated with delayed microvascular reperfusion, whereas the dynamic development of more negative T waves during PCI is associated with earlier microvascular reperfusion. However, in the acute setting, T waves provide little incremental information when compared to ST parameters available in the perinterventional phase Elsevier Inc. All rights reserved. Acute myocardial infarction; T-waves; Primary percutaneous coronary intervention; Electrocardiography Background It is recommended that patients with acute ST-segment elevation myocardial infarction (STEMI) receive reperfusion therapy with fibrinolysis 1 or, preferably, primary percutaneous coronary intervention (PPCI) 2 as soon as possible after symptom onset. Over the years, several large-scale trials have established significant prognostic indicators for the outcome in these patients. 3-6 Corresponding author. Department of Cardiology B, Aarhus University Hospital, Skejby, DK-8200 Aarhus N, Denmark. address: jacobthorsted@gmail.com Today, the diagnosis of patients with STEMI is established very early, frequently already in the prehospital setting. 7 This enables rapid triage and transport to highvolume PPCI centers. 8 Primary percutaneous coronary intervention can be performed with a choice of different intracoronary stents (drug-eluting or bare metal) or possibly just with balloon angioplasty alone in the acute setting. Also available are several adjunctive pharmacologic therapeutic options as well as mechanical devices such as intra-aortic balloon pumps and other assist systems for increased cardiac output and coronary perfusion. With these complex and diverse treatment options follow the possibility of tailored /$ see front matter 2009 Elsevier Inc. All rights reserved. doi: /j.jelectrocard

2 678 J.T. Sørensen et al. / Journal of Electrocardiology 42 (2009) therapy for the individual patient. This, however, requires clinical and para-clinical prognostic indicators to be available before or during the PCI procedure to ensure optimal decision-making at the time of the intervention. Previously established outcome parameters such as coronary biomarkers, left ventricular ejection fraction (LVEF), infarct size estimated by myocardial perfusion imaging (MPI), and evaluation of microvascular perfusion status by ST-resolution (STR) analysis are not available until after the procedure, that is, at a time when irreversible myocardial damage has occurred. We need pre- and periinterventional prognostic markers to help us choose the optimal therapy at the time of PPCI for the individual patient. It is known that the T wave reflects the myocardial repolarization, and changes in the T-wave configuration are among the first to indicate an AMI. Several studies from the thrombolytic era have investigated the prognostic significance of early T-wave changes in patients with AMI It has been demonstrated that post-thrombolytic T-wave changes may reflect microvascular reperfusion status. 12 However, data on the value of T-wave changes in the setting of primary PCI are limited. The purpose of the present study was to investigate if (a) static T-wave amplitude in pre-pci and post-pci electrocardiograms (ECGs) and (b) dynamic T-wave changes during PCI predict microvascular reperfusion status and biochemical and echocardiographic outcome in STEMIpatients at follow-up. Methods Patient population The study population consisted of 215 patients with STEMI enrolled in a previously published randomized, controlled trial evaluating the routine use of a thrombectomy device (Rescue catheter, Boston Scientific/Scimed Inc., Maple Grove, MN) during PPCI. 13 The study period was from January 1, 2004, to February 28, Three patients were excluded as they presented with true posterior infarction (ST-depression in leads V 1 -V 4 and no ST elevation). Five patients were excluded because of insignificant ST-elevation in the index ECG (n = 1); missing post-pci ECG (n = 1); or because T-wave analysis was impossible because of bundle-branch block appearance (n = 1), ventricular pacing (n = 1), or excessive ECG noise (n = 1). The remaining 207 were included in the present substudy. Electrocardiographic measurements All ECGs were previously compiled as part of the MONAMI registry. 14 The index ECG was acquired in the prehospital phase (n = 90), at a local hospital (n = 111), or at the PCI center (n = 6). Prehospital ECGs and interhospital ( transport ) ECGs were stored electronically upon arrival at the catheterization laboratory. Additional ECGs were acquired on arrival at the catheterization laboratory, at time of first wire, first balloon, and first stent, at the end of the procedure (post-pci ECG), and 90 minutes after first balloon. Twelve-lead continuous ST monitoring was initiated on arrival at the PCI center and continued until 90 minutes after first balloon as previously described. 15 Electrocardiograms were analyzed without knowledge of the clinical data and outcome. The electronic ECGs were analyzed by the Code Stat Suite software package (Medtronic Emergency Response Systems, Redmond, WA). Remaining ECGs were analyzed manually. The lead with the maximum ST-elevation pre-pci was chosen for ST-deviation analysis. ST-deviation was measured at the ST-M point (the point of reference for the Fig. 1. The end of the ST segment was determined in leads without ST deviation, and T-wave deflection was recorded from the end of the ST-segment. The time interval measured from the beginning of the QRS complex to the start of the T wave was then applied to the lead with maximal ST-elevation to determine the beginning of the T-wave. The vertical dashed line indicates the onset of the T-wave in a lead without ST-deviation.

3 J.T. Sørensen et al. / Journal of Electrocardiology 42 (2009) determined in leads without ST-deviation, and T-wave deflection was recorded from the end of the ST-segment. The time interval measured from the beginning of the QRS complex to the start of the T wave was then applied to the lead with maximal ST-elevation to determine the beginning of the T wave. If there were no leads without ST deviation, the start of the T wave was established from the lead with the smallest ST deviation. For the index ECG and the post-pci ECG, the T-wave morphology was categorized as (Fig. 1): Fig. 2. In biphasic leads: deflection 1 + deflection 2 = net T-wave amplitude (in millivolts) (eg, an initial downward deflection of 0.2 mv succeeded by a positive deflection of 0.5 mv results in a net T-wave amplitude of 0.3 mv). computerized algorithm in the ECG recorders [Lifepak 12 Medtronic Emergency Response Systems, Redmond, WA]). ST-segment deviation is automatically recorded every 30 seconds, and an increase in ST elevation of more than 0.1 mv for more than 2.5 minutes was defined as an ST peak. T-wave morphology was also studied in the lead with the most pronounced ST-segment elevation before PCI, based on the assumption that this lead most accurately reflects depolarization and repolarization changes in the infarct region. As shown in Fig. 1, the end of the ST segment was a. Pure positive: only positive deflection. In leads with ST-depression, the T-wave is also categorized as positive if the T wave begins below the TP baseline but the direction is never negative. b. Pure negative: only negative deflection. In leads with ST elevation, the T wave is also categorized as negative if the T wave begins above the TP baseline but the direction is never positive. c. Biphasic: initial negative deflection, followed by positive deflection or initial positive deflection, followed by negative deflection. d. Isoelectric: deflection of less than mv in any direction. For pure positive and pure negative T-waves, the positive deflection from the baseline (TP line) and negative deflection Table 1 Baseline characteristics, peri-interventional data, and outcome data according to T-wave amplitude in the index ECG T-wave amplitude b0 mv 0tob0.5 mv 0.5 to b1.0 mv 1 mv P n = 5 n = 63 n = 86 n = 53 Absolute T-wave amplitude in mv (median) 0.20 ( 0.33 to 0.15) 0.30 ( ) 0.70 ( ) 1.2 ( ) b.001 Baseline characteristics Male sex 5 (100%) 52 (83%) 63 (73%) 42 (79%).42 History of Hypertension 0 (0%) 13 (21%) 26 (30%) 13 (25%).39 Diabetes 0 (0%) 5 (8%) 8 (9%) 2 (4%).69 Hyperlipidemia 0 (0%) 7 (11%) 7 (8%) 6 (11%).84 Previous MI 1 (20%) 9 (14%) 6 (7%) 9 (17%).17 Previous PCI 0 (0%) 2 (3%) 2 (2%) 7 (13%).058 Smoking (current) 2 (40%) 39 (62%) 57 (67%) 28 (53%).33 Age (y) 62 (58-73) 64 (57-75) 64 (55-73) 65 (55-73).94 Peri-interventional data Killip class I on admission 5 (100%) 58 (95%) 82 (97%) 48 (91%).63 Systolic blood pressure (mm Hg) 120 ( ) 133 ( ) 135 ( ) 130 ( ).50 Anterior myocardial infarction 2 (40%) 16 (25%) 37 (43%) 39 (74%) b.001 Max. cumulated STE before PCI (mv) 0.5 ( ) 0.8 ( ) 1.6 ( ) 2.3 ( ) b.001 Spontaneous STR before PCI 2 (40%) 22 (37%) 18 (22%) 10 (20%).11 ST peaks during PCI procedure 1 (20%) 8 (14%) 21 (26%) 14 (29%).20 Time from symptom onset to first balloon (min) 708 ( ) 256 ( ) 201 ( ) 192 ( ) b.001 No. of patients with 1-vessel disease 4 (80%) 23 (42%) 42 (55%) 27 (57%).193 TIMI flow 2-3 before PCI 1 (20%) 16 (25%) 29 (35%) 18 (35%).63 TIMI flow 2-3 after PCI 5 (100%) 61 (97%) 77 (95%) 51 (100%).39 Outcome STE 30 min after PCI (mv) 0.17 ( ) 0.08 ( ) 0.16 ( ) 0.22 ( ) b.001 Time to 100% STR a (min) 32 (0-54) 13 (0-55) 33 (7-90) 65 (11-90).011 Max. troponin T (μg/l) 3.0 ( ) 3.6 ( ) 5.0 ( ) 6.1 (3.4-10).054 LVEF 45 (39-58) 50 (40-60) 52 (47-59) 42 (35-53).091 FIS by MPI after 1 mo (% of LV) 9 (0-19) 12 (4-23) 10 (2-23) 11 (4-23).65 Categorical variables are shown as absolute numbers. Continuous variables are shown as median values with interquartile range. ECG indicates electrocardiogram; MI, myocardial infarction; PCI, primary percutaneous coronary intervention; STE, ST-elevation; STR, ST-resolution; TIMI, thrombolysis in myocardial infarction; LVEF, left ventricular ejection fraction; FIS, final infarct size; MPI, myocardial perfusion imaging; LV, left ventricle. a Set to 90 if STR is not achieved.

4 680 J.T. Sørensen et al. / Journal of Electrocardiology 42 (2009) from the baseline (TP line) were measured as a continuous parameter (millivolts). For biphasic T-waves, the initial deflection (negative/positive) and terminal deflection (positive/negative) were measured from the baseline (TP line). To assign the net amplitude (absolute T-wave amplitude) in mv for biphasic T-waves, amplitudes from the initial and terminal deflection were summed (Fig. 2). T-wave amplitudes for all nondigital ECGs were recorded in increments of mv. For digital ECGs, all T-wave measurements provided by the Code Stat Suite software were rounded to the nearest mv unless such values were deemed inaccurate. MAM, in consultation with JTS, CJT, and GSW, made measurements manually from the digital ECGs when such discrepancies were evident. Electrocardiographic analyses for T-wave changes were performed in collaboration with the ECG core laboratory at the Duke Clinical Research Institute. To control for interobserver variability and to test the use of the Marquette algorithm against expert manual ECG interpretation, KCN manually reviewed the index ECG and post-pci ECG for 20 consecutive patient cases for whom ECGs was digitally available. All T-wave amplitudes were categorized into 4 groups: negative (net amplitude, to b0 mv), slightly positive (net amplitude, 0 to b0.50 mv), moderately positive (net amplitude, 0.5 to b1.00 mv), and highly positive (net amplitude, 1 mv) (Tables 1 and 2). Dynamic T-wave changes from index to post-pci ECG were also divided into 4 groups: negative change of more than 1 mv, negative change between 0.5 and 1 mv, negative change up to 0.5 mv, or positive change (Table 3). Outcome parameters Patients underwent echocardiography for evaluation of LVEF before discharge. Troponin-T levels were measured in all patients. Continuous ST-segment monitoring (as described in the Methods section) provided data for resolution analysis. Myocardial perfusion imaging was performed at a 1-month follow-up visit to evaluate final infarct size (FIS). The MPI was performed as gated single photon-emission computed tomography, using 99m Tc-Sestamibi as a tracer. Myocardial perfusion was analyzed by the QPS and QGS software (Hermes Medical Solutions AB, Stockholm, Sweden) and the perfusion defect was reported as a percentage of the left ventricular volume. Statistical analysis Dichotomous data are presented as percentage (number/ valid cases). Continuous variables are presented as medians (25th-75th percentiles; valid cases). Fisher exact test and Kruskal-Wallis test were used as appropriate for compar- Table 2 Baseline characteristics, peri-interventional data, and outcome data according to T-wave amplitude in the post-pci ECG T-wave amplitude b0 mv mv mv N1 mv P n=34 n=118 n=43 n=12 Absolute T-wave amplitude in mv (median) 0.15 ( 0.25 to 0.10) 0.2 ( ) 0.65 ( ) 1.20 ( ) b.001 Baseline characteristics Male sex 26 (77%) 93 (79%) 31 (72%) 12 (100%).20 History of Hypertension 7 (21%) 27 (23%) 15 (35%) 3 (25%).41 Diabetes 4 (12%) 6 (5%) 4 (9%) 1 (8%).37 Hyperlipidemia 5 (15%) 10 (9%) 5 (12%) 0 (0%).49 Previous MI 5 (15%) 14 (12%) 4 (9%) 2 (17%).78 Previous PCI 1 (3%) 7 (6%) 3 (7%) 0 (0%).87 Smoking (current) 24 (71%) 75 (64%) 23 (54%) 4 (33%).10 Age (y) 60 (55-74) 64 (56-73) 67 (55-76) 72 (58-78).42 Peri-interventional data Killip class I on admission 30 (91%) 112 (97%) 39 (91%) 12 (100%).38 Systolic blood pressure (mm Hg) 130 ( ) 134 ( ) 133 ( ) 150 ( ).14 Anterior myocardial infarction 9 (27%) 43 (39%) 32 (74%) 10 (83%) b.001 Max. cumulated ST elevation before PCI (mv) 0.9 ( ) 1.3 ( ) 1.9 ( ) 2.9 ( ) b.001 Spontaneous STR before PCI 12 (36%) 34 (31%) 6 (15%) 0 (0%).017 ST peaks during PCI procedure 5 (15%) 16 (14%) 17 (44%) 6 (50%) b.001 Time from symptom onset to first balloon (min) 251 ( ) 223 ( ) 198 ( ) 185 ( ).20 No. of patients with 1-vessel disease 16 (52%) 55 (52%) 16 (44%) 9 (82%).192 TIMI flow 2-3 before PCI 12 (36%) 40 (35%) 10 (24%) 2 (16%).41 TIMI flow 2-3 after PCI 32 (94%) 112 (99%) 39 (95%) 11 (92%).085 Outcome STE 30 min after PCI (mv) 0.05 ( ) 0.10 ( ) 0.30 ( ) 0.59 ( ) b.001 Time to 100% STR a (min) 10 (0-48) 19 (0-89) 90 (30-90) 90 (90-90) b.001 Max. troponin T (μg/l) 3.1 ( ) 4.5 ( ) 6.4 (2.8-12) 8.6 (6.3-14) b.001 LVEF 46 (39-56) 50 (42-60) 40 (35-50) 38 (35-45) b.001 FIS by MPI after 1 mo (% of LV) 8 (2-22) 10 (2-21) 15 (3-29) 22 (15-29).036 Categorical variables are shown as absolute numbers. Continuous variables are shown as median values with interquartile range. ECG indicates electrocardiogram; MI, myocardial infarction; PCI, primary percutaneous coronary intervention; STE, ST-elevation; STR, ST-resolution; TIMI, thrombolysis in myocardial infarction; LVEF, left ventricular ejection fraction; FIS, final infarct size; MPI, myocardial perfusion imaging; LV, left ventricle. a Set to 90 if STR is not achieved.

5 J.T. Sørensen et al. / Journal of Electrocardiology 42 (2009) Table 3 Outcome according to T-wave dynamic from baseline ECG to ECG immediately post-pci (amplitude) Change in T-amplitude from index ECG to post-pci ECG Less than 1 mv 1 to 0.5 mv 0.5 to 0 mv 0 mv P n = 16 n = 59 n = 105 n = 27 Absolute dynamic change in T-wave 1.25 ( 1.40 to 1.13) 0.68 ( 0.75 to 0.60) 0.30 ( 0.40 to 0.15) 0.15 ( ) b.001 amplitude mv (median) STE 30 min after PCI (mv) 0.14 ( ) 0.13 ( ) 0.11 ( ) 0.26 ( ).01 Time to 100% STR a (min) 19 (0-62) 19 (1-90) 23 (0-90) 90 (32-90).008 Max. troponin T (μg/l) 3.3 ( ) 6.2 ( ) 4.0 ( ) 6.2 (2.8-11).25 LVEF 44 (35-53) 50 (40-60) 47 (40-60) 45 (35-55).24 FIS by MPI after 1 month (% of LV) 6 (1-11) 10 (1-23) 13 (3-23) 16 (7-24).16 Continuous variables are shown as median values with interquartile range. ECG indicates electrocardiogram; PCI, primary percutaneous coronary intervention; STE, ST-elevation; STR, ST-resolution; LVEF, left ventricular ejection fraction; FIS, final infarct size; MPI, myocardial perfusion imaging; LV, left ventricle. a Set to 90 if STR is not achieved. ison of the groups. The statistical significance level was P b.05 (2-sided test). Univariable analysis was performed to determine covariates that predicted FIS. Bivariate correlation analysis was performed to evaluate association between covariates that predicted FIS. The software package STATA 10.0 (StataCorp LP, College Station, TX) was used for statistical analysis. Results The baseline characteristics according to the categorical T-wave amplitude of the index ECG are shown in Table 1.A trend was seen toward a lower percentage of patients with positive T-amplitude (groups 3 and 4) achieving spontaneous STR before PCI. A significant difference between the 4 groups was found in maximum cumulated ST elevation (P b.001) and time to 100% STR, (P =.011). The difference in maximum troponin-t level (P =.054) was borderline significant. Less favorable outcomes were seen with more positive T-waves. The T waves in the ECG recorded immediately after PCI were also stratified in 4 groups according to amplitude (Table 2). The median (IQR) time from the ECG recorded before first wire until the ECG recorded after the procedure was 25 (19-36) minutes. A shift between the 4 groups was seen, as 34 patients exhibited negative T waves in the post- PCI ECG compared to 5 patients in the index ECG. More positive T waves in the post-pci ECG were associated with a lesser degree of spontaneous STR before PCI (P b.001), presence of ST peaks during the procedure (P b.001), longer time to 100% STR (P b.001), a higher troponin-t value (P b.001), lower LVEF (P b.001), and larger infarct size by MPI (P =.036) (Table 2). Net T-amplitude in the post-pci ECG was more pronounced in anterior compared to non-anterior infarctions (median, 0.4 vs 0.1 mv; P b.001), less pronounced in patients achieving spontaneous STR compared to patients not achieving spontaneous STR before PCI (median 0.1 vs 0.3 mv, P =.002), and more pronounced in patients with ST peaks during PCI compared to patients without ST peaks during PCI (median, 0.5 vs 0.2 mv; P b.001). Furthermore, net T-amplitude in the post-pci ECG was associated with maximum cumulated ST-elevation before PCI (Pearson correlation, 0.41; P b.001) and with ST-elevation 30 minutes after PCI (Pearson correlation, 0.64; P b.001). In examining the dynamic changes of the T wave from index ECG to the post-pci ECG (Table 3), a significant difference in time to 100% STR (P b.001) was found among the groups. Less favorable outcomes were seen for patients with an absolute increase in T amplitude. Dynamic T-amplitude change was associated with T amplitude in the post-pci ECG (Pearson correlation, 0.43; P b.001), maximum cumulated ST elevation before PCI (Pearson correlation, 0.16; P =.03), and ST elevation in the post-pci ECG (Pearson correlation, 0.22; P =.003). Covariates associated with FIS are shown in Table 4. In validating the T-wave classification introduced, we found a high correlation between T amplitude evaluated by MAM with the use of the Marquette algorithm compared to T- Table 4 Pre- and peri-interventional variables associated with FIS in univariable analysis Covariate B T P History of Hypertension Hyperlipidemia Previous MI Age (y) Anterior infarct location Maximum cumulated ST-elevation before PCI (mv) Spontaneous STR before PCI b.001 ST peaks during the procedure b.001 ST elevation 30 min after PCI (mv) b.001 T-amplitude post-pci ECG (mv) Dynamic T-amplitude change index ECG to post-pci ECG (mv) TIMI flow before PCI b.001 TIMI flow after PCI Only variables where P b.20 is shown. n = 183 (patients where FIS was available). FIS indicates final infarct size; MI, myocardial infarction; PCI, primary percutaneous coronary intervention; STR, ST-resolution; TIMI, thrombolysis in myocardial infarction.

6 682 J.T. Sørensen et al. / Journal of Electrocardiology 42 (2009) amplitude evaluated manually by an expert (KCN) (Pearson correlation, [P b.001] in the index ECG T amplitude; [P b.001] in the post-pci ECG). Discussion The present study is the first to address the impact of static T-wave amplitude and dynamic T-wave changes at the time of reperfusion on outcome defined by FIS and microvascular perfusion status. The main findings were that (1) large, positive T-waves in both the index ECG and the post-pci ECG are associated with delayed microvascular reperfusion, (2) T-wave amplitude in the post-pci ECG is positively associated with FIS, and (3) an increase in T-wave amplitude during reperfusion therapy is associated with delayed microvascular reperfusion. Previously, Sgarbossa et al 12 found that a negative T- wave in an ECG recorded shortly after reperfusion therapy with thrombolytics was associated with a significantly improved outcome. This has also been shown by Corbalan et al 16 who demonstrated that after thrombolysis the T-wave yielded significant, independent prognostic information. The studies mentioned above were conducted in the thrombolytic era and focused on markers of outcome available after the intervention rather than at the time of intervention (eg, in the catheterization laboratory at presentation or at the bedside during thrombolysis). This study, to a certain degree, corroborates these previous findings. T-wave morphology at presentation and the dynamic T-wave changes (eg, from positive to less positive or even negative) contribute with prognostic information even in the early hours of infarction in patients treated with primary PCI. However, T-wave amplitude in the post-pci ECG was associated with most of the other covariates predicting FIS in the per-interventional phase. Large, positive T-wave amplitudes in the post-pci ECG therefore in part is surrogate marker of anterior infarction, absence of spontaneous STR, and presence of ST peaks during the procedure. Furthermore, T-wave amplitude in the post-pci ECG is associated with the level of STR and residual ST deviation after PCI. Accordingly, much of the information derived from the T-wave amplitude may be a surrogate of ST deviation. The previous studies by Sgarbossa and Corbalan did not focus on peri-interventional T-wave changes as the ECGs in these studies were obtained at days 5 to 7 and 24 hours after admission, respectively. This suggests that T-wave changes may carry significant prognostic information if evaluated after the ST-segment changes has faded. However, if the T-waves are evaluated in the hyper-acute setting, the information they contribute seems to be washed out by other major ECG parameters (presence of ST peaks, persistent ST elevation, and spontaneous STR). Doevendans et al 17 examined early markers of successful reperfusion in thrombolytic-treated STEMI patients. They found that early T-wave inversion and ST-segment resolution were predictors for patency of the infarct-related vessel; STsegment resolution had the highest likelihood ratio. These findings support our results, although the 2 studies have quite different aims and methodology. In a previous study, Herz et al 18 found that mortality in patients presenting with negative T waves was related to the time of symptom onset. Patients with short symptom duration (b2 hours) and negative T waves experienced a better outcome than patients with longer symptom duration and negative T waves. This study did not look specifically at T-wave morphology according to time from symptom onset to treatment. As seen in Table 1, there is a significant difference in symptom onset to treatment time between the 4 groups; however, the group with a very long treatment delay comprises only 5 patients. Furthermore, presence of spontaneous STR and thrombolysis in myocardial infarction (TIMI) 2-3 flow before procedure was not different between groups. The present study is too small to provide any further information on this matter. A smaller study by Wong and colleagues 19 comparing ECG characteristics of the presenting ECG to angiographic findings found that TIMI 3 flow was seen more often in patients with positive T waves. In our study, there was no significant association between TIMI flow and T-wave morphology in the presenting ECG. However, time from symptom onset to angiography was shorter in our study and we did not look at TIMI flow 3 isolated. In 1994, Matetzky et al compared the presence of early (within 24 hours) T-wave inversion to angiographic findings in 94 fibrinolytic-treated STEMI patients. 20 They found a clear association between early T-wave inversion and patency of the infarct-related artery. It may therefore be hypothesized that patients presenting very early with negative T-waves are the ones who experience spontaneous ST-resolution and accordingly have a higher probability for an optimal outcome. 21 This study did not focus on timing intervals from symptom onset to T-wave inversion or how the T waves changed over the first hours of admission; rather, it investigated if T-wave behavior in a population of STEMI patients with differing duration of ischemia could be a reliable and independent predictor of outcome. The method for T-wave measurement could be challenged as being simplistic and unphysiologic. The intent of the study, however, was to find an operational and straightforward method of ECG analysis for use in the catheterization laboratory. Multiple regression analysis was initially performed but abandoned because of colinearity, evident from the strong correlation between T-wave appearance and STsegment parameters. Based on the present study, further investigations of the prognostic significance of per-interventional T-wave changes and potential role of T-wave analysis for triage of therapy at the catheterization laboratory seems to be of little relevance because much of the T-wave information is lost in the major ST-segment changes that most of these patients experience in the acute phase of their infarction.

7 J.T. Sørensen et al. / Journal of Electrocardiology 42 (2009) Study limitations In the present study, the ECGs from 207 patients were analyzed. A larger cohort would have enabled subgroup analysis of, for example, anterior vs non-anterior infarctions. T-wave analysis is complex and there is no consensus on the methodology used to classify and measure T-wave appearance. The method introduced here, nevertheless, proved reproducible and robust. References 1. Antman EM, Hand M, Armstrong PW, et al focused update of the ACC/AHA 2004 guidelines for the management of patients with STelevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2008;51: Boersma E. Does time matter? A pooled analysis of randomized clinical trials comparing primary percutaneous coronary intervention and inhospital fibrinolysis in acute myocardial infarction patients. Eur Heart J 2006;27: Granger CB, Goldberg RJ, Dabbous O, et al. Predictors of hospital mortality in the global registry of acute coronary events. Arch Intern Med 2003;163: Fox KA, Dabbous OH, Goldberg RJ, et al. Prediction of risk of death and myocardial infarction in the six months after presentation with acute coronary syndrome: prospective multinational observational study (GRACE). BMJ 2006;333: Steg PG, Goldberg RJ, Gore JM, et al. Baseline characteristics, management practices, and in-hospital outcomes of patients hospitalized with acute coronary syndromes in the Global Registry of Acute Coronary Events (GRACE). Am J Cardiol 2002;90: Westerhout CM, Fu Y, Lauer MS, et al. Short- and long-term risk stratification in acute coronary syndromes: the added value of quantitative ST-segment depression and multiple biomarkers. J Am Coll Cardiol 2006;48: Terkelsen CJ, Nørgaard BL, Lassen JF, et al. Telemedicine used for remote prehospital diagnosing in patients suspected of acute myocardial infarction. J Intern Med 2002;252: Terkelsen CJ, Lassen JF, Norgaard BL, et al. Reduction of treatment delay in patients with ST-elevation myocardial infarction: impact of prehospital diagnosis and direct referral to primary percutaneous coronary intervention. Eur Heart J 2005;26: Jacobsen MD, Wagner GS, Holmvang L, et al. Quantitative T-wave analysis predicts 1 year prognosis and benefit from early invasive treatment in the FRISC II study population. Eur Heart J 2005;26: Hochrein J, Sun F, Pieper KS, et al. Higher T-wave amplitude associated with better prognosis in patients receiving thrombolytic therapy for acute myocardial infarction (a GUSTO-I substudy). Global utilization of streptokinase and tissue plasminogen activator for occluded coronary arteries. Am J Cardiol 1998;81: Adler Y, Zafrir N, Ben-Gal T, et al. Relation between evolutionary ST segment and T-wave direction and electrocardiographic prediction of myocardial infarct size and left ventricular function among patients with anterior wall Q-wave acute myocardial infarction who received reperfusion therapy. Am J Cardiol 2000;85: Sgarbossa EB, Meyer PM, Pinski SL, et al. Negative T waves shortly after ST-elevation acute myocardial infarction are a powerful marker for improved survival rate. Am Heart J 2000;140: Kaltoft A, Bottcher M, Nielsen SS, et al. Routine thrombectomy in percutaneous coronary intervention for acute ST-segment-elevation myocardial infarction: a randomized, controlled trial. Circulation 2006; 114: Terkelsen CJ, Norgaard BL, Lassen JF, et al. Potential significance of spontaneous and interventional ST-changes in patients transferred for primary percutaneous coronary intervention: observations from the ST- MONitoring in Acute Myocardial Infarction study (The MONAMI study). Eur Heart J 2006;27: Terkelsen CJ, Kaltoft AK, Norgaard BL, et al. ST changes before and during primary percutaneous coronary intervention predict final infarct size in patients with ST elevation myocardial infarction. J Electrocardiol 2009;42: Corbalan R, Prieto JC, Chavez E, Nazzal C, Cumsille F, Krucoff M. Bedside markers of coronary artery patency and short-term prognosis of patients with acute myocardial infarction and thrombolysis. Am Heart J 1999;138: Doevendans PA, Gorgels AP, van der Zee R, et al. Electrocardiographic diagnosis of reperfusion during thrombolytic therapy in acute myocardial infarction. Am J Cardiol 1995;75: Herz I, Birnbaum Y, Zlotikamien B, et al. The prognostic implications of negative T waves in the leads with ST segment elevation on admission in acute myocardial infarction. Cardiology 1999;92: Wong CK, French JK, Aylward PE, Frey MJ, Adgey AA, White HD. Usefulness of the presenting electrocardiogram in predicting successful reperfusion with streptokinase in acute myocardial infarction. Am J Cardiol 1999;83: Matetzky S, Barabash GI, Shahar A, et al. Early T wave inversion after thrombolytic therapy predicts better coronary perfusion: clinical and angiographic study. J Am Coll Cardiol 1994;24: Rimar D, Crystal E, Battler A, et al. Improved prognosis of patients presenting with clinical markers of spontaneous reperfusion during acute myocardial infarction. Heart 2002;88:352.