Although the prognosis of patients with stress (takotsubo) cardiomyopathy is relatively favorable, serious complications occur in some patients. It is generally accepted that electrocardiography is an essential tool for the diagnosis of stress cardiomyopathy, with findings highly suggestive of the characteristics of myocardial damage. We tested the hypothesis that the quantitative analysis of electrocardiograhic changes can predict complications in stress cardiomyopathy. The study subjects were 85 patients with stress cardiomyopathy. A total of 34 patients developed ≥1 in-hospital complications (heart failure, intraventricular pressure gradient [>30 mm Hg], cardiogenic shock, ventricular tachycardia/fibrillation, and embolism). Patients with complications were likely to have a higher heart rate (96 ± 25 vs 76 ± 17 beats/min, p <0.001), larger sum of ST-segment elevation in 12 leads (median 10.5 mm; interquartile range 5.0 to 17.5 vs 3.0 mm, interquartile range 0 to 7.0; p <0.001) and extension of ST-segment elevation to limb leads (50% vs 12%, p <0.001) than those without complications. Multivariate logistic regression analysis identified heart rate (odds ratio 1.05, 95% confidence interval 1.02 to 1.07, p = 0.001) and sum of ST-segment elevation in 12 leads (odds ratio 1.24, 95% confidence interval 1.11 to 1.39, p <0.001) as significant and independent predictors of complications. Receiver operating characteristic analysis selected 5.5 mm as the best cutoff value of sum of ST-segment elevation in 12 leads for the prediction of complications, with a sensitivity and specificity of 74% and 73%, respectively, and area under the curve of 0.81 (95% confidence interval 0.72 to 0.90, p <0.001). The results suggest that the extent and magnitude of ST-segment elevation on the electrocardiogram are potentially useful predictors of in-hospital complications in patients with stress cardiomyopathy.
The 12-lead electrocardiogram (ECG) is a fundamental and accessible tool for the diagnosis of stress (takotsubo) cardiomyopathy. The cardinal electrocardiographic findings for stress cardiomyopathy are initial ST-segment elevation and subsequent T-wave inversion with QT prolongation. Several studies have reported that the number of leads that exhibit ST-segment elevation and sum of ST-segment deviation correlate with the myocardial infarct size, reduced left ventricular ejection fraction (LVEF), and prognosis of patients with ST-segment elevation myocardial infarction (STEMI). However, the prognostic significance of the electrocardiographic changes has not been fully elucidated in patients with stress cardiomyopathy. In the present study, we tested the hypothesis that serial changes and quantitative analysis of electrocardiographic changes can predict future in-hospital complications in patients with stress cardiomyopathy.
Methods
We retrospectively analyzed the data from 85 patients diagnosed with stress cardiomyopathy from June 2002 to November 2010 at 3 cardiovascular centers in Kumamoto, Japan (Kumamoto University Hospital, Saiseikai Kumamoto Hospital, and Japanese Red Cross Kumamoto Hospital). We evaluated the electrocardiographic changes at hospital admission and ≤24 hours after admission, clinical characteristics, and the development of complications. The diagnostic criteria for stress cardiomyopathy were according to the Mayo Clinic proposed diagnostic criteria : (1) transient akinesia, hypokinesia, or dyskinesia of the left ventricular mid-segments with or without apical involvement, and regional wall motion abnormalities extending beyond a single epicardial vascular distribution; (2) absence of obstructive coronary disease or angiographic evidence of acute plaque rupture; (3) new electrocardiographic abnormalities at the episode (either ST-segment elevation and/or T-wave inversion) or elevated cardiac troponin; and (4) absence of pheochromocytoma or myocarditis. The coronary arteries were evaluated by coronary angiography (n = 81) or 64-row multidetector computed tomographic angiography (n = 4). We excluded patients with bundle branch block, pacemaker rhythm, and cerebral vascular accidents. These data were examined by review of the medical records.
The 12-lead ECG was recorded at a paper speed of 25 mm/s and an amplification of 10 mm/mV. ST-segment deviation from baseline was manually measured at the J-point according to the universal definition of myocardial infarction. ST-segment elevation was considered present if the deviation was >0.5 mm in the limb leads and >1.0 mm in ≥2 contiguous precordial leads. T-wave inversion was considered present if T-wave inversion was >1.0 mm in ≥2 contiguous leads. The sum of ST-segment elevation level (ΣSTE) represented the sum of ST-segment deviation in each of the 12 ECG leads. The QTc interval was calculated using Bazett’s formula.
Among the 68 patients (80%) with ST-segment elevation on admission, 62 had a 12-lead ECG taken at 24 hours after admission that was suitable for evaluation. We defined ST-segment resolution as the percentage of reduction in the ΣSTE calculated at admission and 24 hours after admission. We also evaluated the relation between the complications and ST-segment resolution.
Creatinine kinase was measured sequentially until it peaked, and the maximum value was recorded. Other laboratory tests were conducted on admission. Cardiac troponin T was evaluated by qualitative or quantitative assay. A positive reaction or level >0.1 ng/ml was considered troponin T positive. The glomerular filtration rate was calculated using the level-modified Modification of Diet in Renal Disease modified for Japanese. The LVEF at admission was assessed by left ventriculography or 2-dimensional echocardiography using Simpson’s biplane method.
The following complications were analyzed. Heart failure was diagnosed clinically and by overt lung edema on chest X-ray and dyspnea or oxygen desaturation requiring intravenous therapy, such as diuretics or mechanical support. Ventricular tachycardia/fibrillation included fatal sustained ventricular tachycardia or fibrillation. The interventricular pressure gradient was defined by a transit interventricular pressure gradient (>30 mm Hg) by left ventricular outflow tract obstruction during the acute period. The pressure gradient was measured directly with a catheter or by continuous wave Doppler echocardiography. Acute severe mitral regurgitation (transit and acute mitral regurgitation detected by echocardiography or left ventriculography) was also analyzed. Mitral regurgitation was assessed by a semiquantitative scoring system using color Doppler images or Sellers classification. Embolism was defined as de novo embolism during hospitalization or overt left ventricular thrombosis confirmed by echocardiography. Finally, cardiogenic shock was defined as systolic blood pressure <90 mm Hg with signs of tissue hypoperfusion requiring catecholamine or fluid therapy.
Continuous data were expressed as the mean ± SD or median (interquartile range). Differences between groups were examined using the Student’s t test or Mann-Whitney U test for unpaired data. Categorical values were presented as numbers (percentage) and were compared using the chi-square test or Fisher exact test, as appropriate. Multivariate logistic regression analysis was performed to identify the independent predictors of complications. We entered the following variables age, gender, ΣSTE, ST-segment elevation in the limb leads, T-wave inversion, interval to ECG <6 hours, heart rate, and systolic blood pressure, into a multivariate logistic regression model using the forward selection method. The odds ratio with 95% confidence intervals and p values were presented. Receiver operating characteristic curve analysis was performed to discriminate between patients with and without complications. A 2-tailed value of p <0.05 was considered statistically significant. All statistical analyses were performed with SPSS, version 19 (SPSS, Chicago, Illinois).
The human ethics review committee of Kumamoto University approved study protocol, and each subject provided a signed consent form.
Results
The baseline characteristics are listed in Table 1 . The examination of the ECG at admission showed ST-segment elevation and T-wave inversion in 68 (80%) and 34 (40%) patients, respectively, and ST-segment elevation with T-wave inversion in 17 patients (20%). Among patients with ST-segment elevation, ST-segment elevation in the limb leads, as well as in the precordial leads, was observed in 23 patients (27%). The median ΣSTE was 5.0 mm (range 2.0 to 11.3). The median interval between the appearance of symptoms and the first ECG on admission was 4.6 hours. The ECG was recorded within 6 hours of symptom onset in 46 patients (54%). Typical apical ballooning was observed in 79 patients (93%) and 6 other patients presented with midventricular or basal ballooning. The LVEF was available for 73 patients, and the mean LVEF was 48.0 ± 12.0% at admission.
| Variable | Patients (n = 85) |
|---|---|
| Age (years) | 73 ± 9 |
| Women | 72 (85%) |
| Emotional stress | 27 (32%) |
| Physical problems | 19 (22%) |
| Hypertension ⁎ | 54 (64%) |
| Diabetes mellitus | 15 (18%) |
| Dyslipidemia † | 25 (29%) |
| Previous heart failure | 5 (6%) |
| Previous coronary artery disease | 8 (9%) |
| Current smoker | 7 (8%) |
| Heart rate (beats/min) | 84 ± 22 |
| Systolic blood pressure (mm Hg) | 133 ± 31 |
| Diastolic blood pressure (mm Hg) | 75 ± 19 |
| Electrocardiogram findings | |
| ST-segment elevation | 68 (80%) |
| T-wave inversion | 34 (40%) |
| ST-segment elevation with T-wave inversion | 17 (20%) |
| Sum of ST-segment elevation level (mm) | 5.0 (2.0–11.3) |
| ST-segment elevation in limb leads | 23 (27%) |
| QTc (ms) | 458 ± 73 |
| Symptom onset to electrocardiogram interval (hours) | 4.6 (1.5–24) |
| Symptom onset to electrocardiogram interval <6 hours | 46 (54%) |
| Medications at admission | |
| No medication | 37 (44%) |
| Calcium channel blocker | 24 (28%) |
| Angiotensin-converting enzyme inhibitor | 14 (16%) |
| Angiotensin II receptor blocker | 20 (24%) |
| β Blockers | 4 (5%) |
| Nitrate | 8 (9%) |
| Laboratory and imaging findings | |
| Peak creatinine kinase (IU/L) | 173 (107–298) |
| Hemoglobin (g/dl) | 12.1 ± 1.6 |
| Estimated glomerular filtration rate (ml/min/1.73 m 2 ) | 71.3 (51.2–86.2) |
| Cardiac troponin, positive | 53/66 (80%) |
| Left ventricular ejection fraction (%) | 48.0 ± 12.0 |
⁎ Treated with antihypertensive drugs or blood pressure ≥140/90 mm.
† Treated with lipid-lowering drugs or low-density cholesterol level ≥140 mg/dl and/or serum triglycerides ≥150 mg/dl and/or high-density cholesterol <40 mg/dl.
The clinical outcomes are listed in Table 2 . A total of 22 patients (26%) developed heart failure. Among them, 4 patients required intra-aortic balloon pumping and 2 mechanical ventilation. Interventricular pressure gradient and severe acute mitral regurgitation were diagnosed in 11 (13%) and 9 (11%) patients, respectively. The mean intraventricular pressure gradient was 67 ± 34 mm Hg. Interventricular pressure gradient was noted in all patients with severe acute mitral regurgitation. Cerebral embolism was observed in 3 patients. Ventricular tachycardia and ventricular fibrillation occurred in 2 and 1 patient, respectively. Cardiogenic shock occurred in 9 patients, and 2 of these patients died immediately after hospital admission.
| Outcome | n (%) |
|---|---|
| Complication | |
| Heart failure | 22 (26%) |
| Intra-aortic balloon pumping | 4 (5%) |
| Mechanical ventilation | 2 (2%) |
| Interventricular pressure gradient | 11 (13%) |
| Severe acute mitral regurgitation | 9 (11%) |
| Embolism | 3 (4%) |
| Ventricular tachycardia/fibrillation | 3 (4%) |
| Cardiogenic shock | 9 (11%) |
| In-hospital death | 2 (2%) |
| Number of above complications | |
| 1 | 20 (24%) |
| ≥2 | 14 (16%) |
The study patients were classified into the noncomplication group (n = 51) and complication group (n = 34), representing those without and with ≥1 of these complications. The differences between the 2 groups are listed in Table 3 . The patients in the complication group were more likely to have higher heart rate, ΣSTE, ST-segment elevation in the limb leads, frequency of symptom onset to ECG of <6 hours, lower diastolic blood pressure, frequency of T-wave inversion, and lower LVEF, than the patients in the noncomplication group.
| Variable | Complication Group (n = 34) | Noncomplication Group (n = 51) | p Value |
|---|---|---|---|
| Age (years) | 75.1 ± 8.5 | 71.5 ± 9.3 | 0.08 |
| Women | 28 (82%) | 44 (86%) | 0.62 |
| Preceding stress ⁎ | 17 (50%) | 29 (57%) | 0.53 |
| Hypertension | 22 (65%) | 32 (63%) | 0.85 |
| Diabetes mellitus | 7 (21%) | 8 (16%) | 0.56 |
| Dyslipidemia | 8 (24%) | 17 (33%) | 0.33 |
| Previous heart failure | 2 (6%) | 2 (6%) | 1.00 |
| Previous coronary artery disease | 5 (15%) | 3 (6%) | 0.26 |
| Heart rate (beats/min) | 96 ± 25 | 76 ± 17 | <0.0001 |
| Systolic blood pressure (mm Hg) | 126 ± 37 | 138 ± 26 | 0.08 |
| Diastolic blood pressure (mm Hg) | 70 ± 21 | 79 ± 17 | 0.04 |
| No medication | 13 (38%) | 24 (47%) | 0.42 |
| ST-segment elevation | 33 (97%) | 35 (69%) | 0.001 |
| T-wave inversion | 8 (24%) | 26 (51%) | 0.01 |
| Sum of ST-segment elevation level (mm) | 10.5 (5.0–17.5) | 3.0 (0–7.0) | <0.0001 |
| QTc (ms) | 443 ± 53 | 467 ± 82 | 0.13 |
| ST-segment elevation in limb leads | 17 (50%) | 6 (12%) | <0.0001 |
| Symptom onset to electrocardiogram interval <6 hours | 23 (68%) | 23 (45%) | 0.04 |
| Left ventricular ejection fraction (%) | 44.1 ± 10.5 | 50.8 ± 12.4 | 0.02 |
| Hemoglobin (g/dl) | 12.0 ± 1.8 | 12.1 ± 1.4 | 0.63 |
| Estimated glomerular filtration rate (ml/min/1.73 m 2 ) | 62.7 (36.9–87.4) | 72.3 (58.8–85.5) | 0.20 |
| Peak creatinine kinase (IU/L) | 180 (115–320) | 168 (95–313) | 0.41 |
| Cardiac troponin T, positive | 22/25 (88%) | 31/41 (76%) | 0.22 |
⁎ Emotional and physical stress preceding symptoms of stress cardiomyopathy.
Of the 62 patients who could be evaluated for resolution of ST-segment elevation at 24 hours after admission, complications occurred in 32 patients (52%). The complication group was not different from the noncomplication group with respect to the frequency of ST-segment elevation at 24 hours (63% vs 53%, p = 0.61) or ΣSTE at 24 hours (4.0 mm, range 0 to 8.8, vs 2.0 mm, range 0 to 4.5; p = 0.13). Similarly, no significant differences were seen between the 2 groups in resolution of ST-segment elevation (65%, range 31% to 100%, vs 65%, range 0% to 100%; p = 0.30) and resolution of ST-segment elevation of <50% at 24 hours (38% vs 43%, p = 0.80).
Multivariate logistic regression analysis identified heart rate (odds ratio 1.05, 95% confidence interval 1.02 to 1.07, p = 0.001) and ΣSTE (odds ratio 1.24, 95% confidence interval 1.11 to 1.39; p <0.001) as independent and significant predictors of complications ( Table 4 ). Receiver operating characteristic curve analysis showed that for a ΣSTE cutoff value of 5.5 mm, the sensitivity and specificity for predicting the development of complications was 74% and 73%, respectively, with an area under the curve of 0.81 (95% confidence interval 0.72 to 0.90; p <0.001; Figure 1 ) .
