Takotsubo cardiomyopathy
An important differential diagnosis to acute myocardial infarction
An important differential diagnosis to acute myocardial infarction
Lene Hüche Nielsen, 1 MD, Kim Munk,2 MD, Ole Goetzsche,3 DMSc, Evald Høj Christiansen,2 PhD,
and Bjarne Linde Nørgaard,1 PhD
Faktaboks
Methods
From October 2002 to November 2005, patients with a suspected acute myocardial infarction who underwent an acute or subacute coronary angiography at the Department of Cardiology, Aarhus University Hospital, Skejby, and met the following criteria: 1) ballooning of the LV-apex diagnosed by echocardiography or left ventriculography and 2) absence of obstructive coronary disease (no luminal stenosis ≥ 50%) were retrospectively enrolled in this study.
Three patients were directly hospitalized at Skejby Hospital, while four patients were transferred from local hospitals. Cases were reviewed and included in this study if the cardiologists (EHC, BLN) came to agreement about the diagnosis. Information about the clinical presentation, electrocardiographic and echocardiographic features in the acute phase and at follow-up were obtained from medical records and the laboratory database (LAPKA). Blood samples for measurement of troponin T (tnT) and creatine kinase myocardial band (CK-MB) were drawn on admission, after 6-9 hours and after 12-24 hours (Elexis 1010, Roche, respectively).
Electrocardiograms (ECG) on admission, during the hospital stay and at follow-up were analyzed for the presence of Q waves (≥ 1 mm in at least two contiguous leads in aVL, aVF, I, II, V5-6 or Q wave > 30 ms in V1-4), ST-segment elevation (at the J-point ≥ 2 mm in V1-3 or ≥ 1 mm in aVL, aVF, I, II, III, and V4-6 in at least two contiguous leads), ST-segment depression (≥ 1 mm in at least two contiguous leads measured 60 ms after the J-point), T-wave inversion (negative T-wave in I, II or V2-6), and the duration of the heart rate-corrected QT interval (QTc) was calculated by use of Bazett's formula (QTc = QT/square root of the RR interval, where QTc > 440 ms is prolonged). Echocardiography, coronary angiography and left ventriculography were performed according to daily practice. The coronary angiography was analyzed by experienced cardiologists (EHC, BLN).
Results
Overall, 2824 acute coronary angiographies were performed during the study period. Eleven patients (0.4%) met the inclusion criteria. One patient (female, 78 years) was excluded as the review of the coronary angiography revealed significant coronary artery disease. One patient (female, 44 years) was excluded due to the presence of a dissection distally in the left coronary artery (LAD). Further two patients (female, 83 years, male, 81 years) were excluded due to lack of follow-up. This report therefore comprises seven patients (0,2%) with TC.
Clinical characteristics
The clinical characteristics of the patients are presented in Ta ble 1.
The study population comprises six females and one male (mean age of 70, range 53-81 years). All females were in the postmenopausal age. Five patients had emotional or physical stress associated with onset of symptoms. Six patients had chest pain and/or dyspnoea. One patient presented with syncope without other cardiac symptoms.
Complications
During hospitalization five patients had clinical signs of heart failure, of which two patients were in cardiogenic shock requiring inotropic therapy (patient 3: dopamine, patient 7: dobutamine). Three patients had transient episodes of non-sustained ventricular tachycardia. Two patients had or developed pneumonia, which was treated by antibiotics.
Electrocardiographic features
Electrocardiographic data are presented in Table 2 . Figure 2 shows ECGs from two patients on admission, during hospitalization and at follow-up. ST-segment elevation in the admission ECG was documented in five of seven patients whereas Q-waves were present in two. T-wave inversion was present on the admission ECG or on subsequent ECGs in all patients. The QTc interval was prolonged in all patients during the hospital stay. One patient received treatment with a potentially QTc-prolonging drug (cipramil 20 mg/day, patient 1) before admission. Follow-up ECG normalized in six patients except from the presence of discreet T-wave inversion in two patients. Only one patient had sustained Q-waves in the ECG (follow-up at one week). The QTc interval normalized in patients with > 2 weeks of follow-up.
Cardiac catheterization
Four patients with a suspected acute myocardial infarction had coronary angiography performed ≤ 12 hours after symptom onset. The remaining three patients underwent coronary angiography within 24 hours, three and 11 days after the onset, respectively. No obstructive coronary disease was identified in any of the patients. The delay in the first case was due to an attempt of haemodynamic stabilizing the patient with inotropic agents at the local hospital. In the second case the patient was initially hospitalized for surgery because of a fracture of the hip, and subsequently transferred for coronary angiography because of ECG changes. An 81 year-old woman was admitted with pneumonia, and due to continued chest pain a coronary angiography was performed on the eleventh day. Three patients had normal coronary arteries, three had mild luminal irregularities and one had a non-significant stenosis distally in the ramus descendens anterior. Ventriculography was performed in four patients (Figure 1).
Echocardiography
All patients had echocardiography performed during hospital stay. In three patients in whom ventriculography was not performed, LV apical ballooning was assessed by echocardiography. During the acute phase, all patients had mid-ventricular dysfunction and apical hypokinesis or akinesia. A transient dynamic intra-ventricular pressure gradient of 33 mmHg was demonstrated in one patient. This gradient had disappeared at follow-up echocardiography eight weeks later. Left ventricular systolic function was normalized at follow-up in all patients (Table 1).
Cardiac biomarkers
Elevated levels of tnT and CK-MB (Table 1) were observed in all patients. The mean peak (range) tnT and CK-MB were 1.7 (0.5-2.6) and 30 (10.8-46) mg/l, respectively.
Discussion
The etiology of the TC has not been clearly established. Previous reports of the prevalence of TC vary depending on the selection of patients and the definition of the syndrome. In a few reports the prevalence of TC ranged 1-2% among patients presenting with a suspected STEMI [10-14], thus significantly higher than in the current study. This first Danish study of patients with TC confirms previously reported characteristics: a) approximately 90% of the patients are elderly women, b) in more than 50% of the cases an emotional or physical trigger event can be identified before onset of symptoms, c) symptoms most frequently are chest pain or dyspnoea [1, 2, 5, 14]. The clinical picture is, however, heterogeneous varying from fatigue, backache, dizziness, syncope to ventricular fibrillation and shock [2, 5, 14].
In the current study two of the seven patients were in a state of cardiogenic shock and further three had transient clinical heart failure. No sustained malignant arrhythmias were identified.
Information about the temporal ECG changes in patients with TC is sparse. Our data indicate that the syndrome mimics STEMI with acute development, ST-elevation decreasing over days together with evolution of T-wave inversion and QTc prolongation. A previous report demonstrated a longer QTc three days after admission in patients with TC compared to QTc in patients with STEMI [4]. Q-waves in patients with TC may be present during the acute phase. Unlike most patients presenting with STEMI, this study indicates that Q-waves in patients with TC are transient. In general, normalization of the ECG is expected at follow-up. In accordance with previous findings [5, 6, 15] a discrepancy was demonstrated in the current study between the modest rise in cardiac biomarkers and the pronounced ECG changes and LV systolic dysfunction, respectively.
Pathophysiological aspects
When TC was initially described by Japanese groups the syndrome was interpreted as a demographical isolated syndrome [1, 2]. However, several case studies have been reported also in other ethnicities [4-6, 10, 12-14].
The etiology of TC has not been clearly established. Takotsubo cardiomyopathy is often provoked by emotional or physical stress and seems to represent a pathophysiological catecholamine effect on the myocardium mediated by a direct toxic effect or by a macro/ microvascular dysfunction. This thesis is supported by the finding of elevated plasma catecholamine levels during the acute phase in more than 70% of patients with TC as compared to age-matched STEMI-patients [16]. Moreover, similar transient clinical, electrocardiographic and impairment of LV function have been demonstrated in patients with pheochromocytoma and catecholamine crisis [16-20]. An apical sympathetic innervation defect in patients with TC has been demonstrated by single photon emission computed tomography using 123 I-metaiodobenzylguanidine (MIBG) [17].
Sympathetic denervation of the heart leads to hypersensitivity to circulating ligands similar to what has been demonstrated in skeleton muscles [21]. A sympathetic denervation combined with an elevated b-receptor density in the apical myocardium [18] may hypothetically explain a condition in which a sudden rise in plasma catecholamines leads to regional LV stunning. This hypothesis also seems to explain the peculiar distribution of the functional LV disturbance.
Myocarditis has been proposed as a mechanism for the reversible LV dysfunction [19]. However, since cardiac magnetic resonance imaging does not show any sign of regional delayed hyperenhancement which is a stringent finding in myocarditis [19] the latter seems unlikely. A dynamic left ventricular outflow tract obstruction has been proposed in the pathogenesis of TC [1] which is secondary to a catecholamine induced intracardiac pressure gradient between the apical and the basal segments of the heart leading to elevation in wall stress, oxygen requirement, ischemia and therefore apical ballooning. Contradicting this theory is that the prevalence of an intraventricular gradient in patients with TC is reported in only about 16% of the studies [16]. In accordance, this study demonstrated the presence of an intracavitary pressure gradient in one out of seven patients.
Atypical forms of TC have been described with right ventricular involvement, and basal or mid-ventricular akinesia with or without apical hyperkinesis [16]. This variant form of TC might demand for reconsideration of the diagnostic criteria for this syndrome. Furthermore, in a recent animal study it was hypothesized that high levels of circulating catecholamines trigger a reversible switch in intracellular signal trafficking in ventricular cardiomyocytes from Gs protein to Gi protein signaling via b 2 -adrenoceptors causing a phenotype dependent regional negative inotropic response [20]. This hypothesis may have implications for the use of inotropic drugs in the treatment of patients with TC.
Treatment and prognosis
In the acute phase, TC has clinical, electrocardiographic and echocardiographic characteristics indistinguishable from STEMI, thus acute coronary angiography is indicated in most cases. The current knowledge on the treatment of TC is deficient, but in light of the above mentioned catecholamine associated pathophysiological theory, treatment with adrenergic blocking agents, if tolerated, is potentially beneficial [16, 19].
In this report two patients were treated for cardiogenic shock. In one patient dobutamine was administered. The latter approach seems counterintuitive if TC is the result of a catecholamine mediated effect on the myocardium. Left ventricular ejection fraction was significantly lower in this patient as compared to the patient in whom dopamine was administered. An alternative haemodynamic support may be intraaortic balloon-pump counter-pulsation. One can consider short-term treatment with anticoagulation in order to prevent mural thrombus formation in patients with significantly reduced LV function. Although serious complications to TC are described, hospital mortality is low (0-8%), recurrence of the condition is rare and the long-term prognosis is favourable [10].
Limitations
The current study is limited by its retrospective nature and the small number of patients. The limited number of patients does not allow for statistical analysis. In this study it was not possible to determine the actual incidence of TC. Some TC cases may have been neglected because of: 1) a latency between symptom onset and the time of admission, 2) an echocardiography was not always performed in the acute phase, and 3) a ventriculography was not performed systematically in all patients with chest pain and a normal coronary angiography.
Conclusion
Danish patients with TC have clinical, electrocardiographic, biochemical and echocardiographic characteristics similar to what has been described in TC in other populations. As TC in the acute phase is indistinguishable from STEMI, acute coronary angiography usually is warranted. The ECG changes and the LV systolic function normalise within weeks to months. The pathophysiological mechanisms of TC have not been clearly established, but the syndrome appears to be mediated by a pathological catecholamine effect on the myocardium.
References