Author + information
- Thomas Stiermaier, MD,
- Charlotte Eitel, MD,
- Stefanie Denef, MD,
- Steffen Desch, MD,
- Gerhard Schuler, MD,
- Holger Thiele, MD and
- Ingo Eitel, MD∗ ()
- ↵∗Medical Clinic II, University of Luebeck, University Hospital Schleswig-Holstein, Ratzeburger Allee 160, 23538 Lübeck, Germany
Although prognosis of Takotsubo cardiomyopathy (TTC) is generally thought to be favorable, the initial presentation can be accompanied by various complications, including life-threatening arrhythmias. Previously, the prevalence of ventricular tachycardia (VT), ventricular fibrillation (VF), or complete atrioventricular (AV) block was reported to be approximately 8%, mostly in the setting of QT prolongation (1,2). However, data regarding the exact prevalence, optimal management, and prognostic relevance of arrhythmias in TTC patients are lacking.
This single-center study consisted of 203 consecutive patients with suspected TTC presenting to our institution. Of these, 25 patients were excluded because of cardiovascular magnetic resonance (CMR) evidence of acute myocardial infarction with spontaneous lysis of thrombus and myocarditis. TTC (n = 178) was diagnosed according to the Mayo Clinic criteria and was confirmed via CMR (3,4). Complete recovery of left ventricular (LV) function within 6 months was documented in all surviving patients. Continuous electrocardiographic monitoring for at least 24 h and serial 12-lead electrocardiogram recording was performed. QT prolongation was defined as QTc >450 ms for male and >460 ms for female patients. The occurrence of life-threatening arrhythmias consisting of VT, VF, asystole, and complete AV block was determined during the hospital stay. VTs lasting >30 s were classified as sustained. Mortality 1 year after initial presentation was obtained by telephone contact with the patient or the treating physician and compared between patients experiencing arrhythmias and those without arrhythmias. Follow-up after 1 year was available in 169 patients (95%). The 9 patients lost to follow-up were excluded from survival comparison (Table 1).
The study population consisted of predominantly female patients (87.6%) with a mean age of 71.8 ± 10.4 years. A stressful trigger could be identified in 67.4% of patients, and the majority exhibited the typical apical ballooning pattern (70.8%). Initial LV ejection fraction was 44.2 ± 9.3% and recovered to normal values within 6 months in all patients (60.3 ± 7.0%). CMR (available in 125 patients) demonstrated focal edema in 89.2%. Only a few patients (9.2%) showed subtle fibrosis/necrosis by means of late gadolinium enhancement when using a threshold of 3 SDs above remote myocardium. QTc prolongation was identified in 49.1%.
Life-threatening arrhythmias occurred in 24 patients (13.5%) (Table 1). Patients developing arrhythmias had a higher prevalence of subtle fibrosis/necrosis on CMR (33.3% vs. 7.2%; p = 0.04), a lower LV ejection fraction (39.7 ± 11.0% vs. 44.7 ± 8.4%; p = 0.02), and they showed a significantly longer QTc (476.5 ± 74.4 ms vs. 442.8 ± 40.9 ms; p < 0.01). Although all polymorphic VT occurred in the presence of QTc prolongation, all but 1 patient with monomorphic VT had a normal QTc duration. None of the patients received an implantable cardioverter-defibrillator, but 3 of the 4 patients with complete AV block underwent permanent pacemaker implantation.
Mortality 1 year after the index event was significantly higher in patients experiencing life-threatening arrhythmias (10 [44%]) compared with patients without arrhythmias (14 [10%]; p < 0.01) (Table 1). Cardiovascular death occurred in 4 TTC patients with arrhythmias (17%) and 8 patients without arrhythmic events (5%; p = 0.04). The difference was driven by an increased mortality in patients experiencing VF; sustained, monomorphic VT; or AV block.
Our results indicate that the prevalence of arrhythmias in TTC is higher than expected and, for the first time to our knowledge, clearly links these arrhythmias to worsened survival (1,2). Importantly, there was a clear divergence of mortality curves ≤2 months after initial presentation, which implies that arrhythmias are of higher prognostic value in the convalescent phase of TTC. Therefore, our data illustrate the significance of arrhythmic events as a determinant of outcome and a useful approach to risk stratification in TTC.
Several explanations for the occurrence of arrhythmias in TTC seem reasonable. The high proportion of patients with QTc prolongation indicates a potential role of after-depolarizations in the setting of differences in refractory periods. Consistently, virtually all polymorphic VTs were associated with QT prolongation. However, a considerable number of patients developed monomorphic VT, mostly in the setting of a normal QT duration. CMR studies demonstrate the absence of scar formation in TTC, which makes re-entry an unlikely underlying mechanism for ventricular arrhythmias (4). However, we identified a higher incidence of subtle, diffuse fibrosis/micronecrosis in patients with arrhythmias, which may reflect a substrate for malignant arrhythmias (5). Moreover, almost all patients developed myocardial edema that may result in electrical inhomogeneity predisposing to arrhythmias. Another possible mechanism is catecholamine excess and enhanced sympathetic activity. Secondary increase in vagal tone might explain the occurrence of bradyarrhythmias. However, it cannot be excluded that TTC occurs secondarily due to heart rhythm disease, rather than representing the substrate for the arrhythmia in some cases.
The current published data provides limited information on the management of arrhythmias in the setting of TTC. In particular, the need for implantable cardiac-defibrillators is a matter of debate given the reversible nature of TTC and the substantial number of noncardiovascular deaths. Considering the lack of data with respect to recurrence of arrhythmias, cardiac defibrillator implantation has to be evaluated individually. In conclusion, life-threatening arrhythmias are a common finding in patients with TTC and have a major prognostic effect.
Please note: All authors have reported that they have no relationships relevant to the contents of the paper to disclose.
- 2015 American College of Cardiology Foundation