Echocardiography is frequently the initial noninvasive imaging modality used to assess patients with takotsubo cardiomyopathy (TTC). Standard transthoracic echocardiography can provide, even in the acute care setting, useful information about left ventricular (LV) morphology as well as regional and global systolic or diastolic function. It allows the differentiation of different LV morphologic patterns according to the localization of wall motion abnormalities. A “circumferential pattern” of LV myocardial dysfunction characterized by symmetric wall motion abnormalities involving the midventricular segments of the anterior, inferior, and lateral walls should be considered suggestive of TTC and included in the differential diagnosis of acute coronary syndromes. Moreover, advanced echocardiographic techniques, including speckle-tracking, myocardial contrast, and coronary flow studies, are providing mechanistic and pathophysiologic insights into this unique syndrome. Early identification of any potential complications (i.e., LV outflow tract obstruction, reversible moderate to severe mitral regurgitation, right ventricular involvement, thrombus formation, and cardiac rupture) are crucial for the management, risk stratification, and follow-up of patients with TTC. Because of the dynamic evolution of the syndrome, comprehensive serial echocardiographic examinations should be systematically performed. This review focuses on these aspects of imaging and the increasing understanding of the clinical and prognostic utility of echocardiography in TTC.
Takotsubo cardiomyopathy (TTC), also known as stress cardiomyopathy, left ventricular (LV) apical ballooning syndrome, or broken heart syndrome, is a unique cardiac syndrome characterized by transient LV systolic dysfunction often mimicking an acute myocardial infarction (AMI). It is usually precipitated by acute emotional and/or physical stress and is characterized by three distinctive features: (1) the presence of acute LV wall dysfunction, (2) the absence of significant obstructive coronary artery disease and (3) the rapid improvement of LV systolic function within a few days or weeks.
The development of TTC most likely reflects the cardiac response to a surge of catecholamines, often triggered by a stressful event. In fact, serum epinephrine levels have been shown to be significantly higher in patients with TTC than in patients with acute ischemic heart failure, suggesting an excessive hypothalamic-pituitary-adrenal axis response to stress. Of note, also the infusion of epinephrine or dobutamine for therapeutic or diagnostic purposes has been reported to precipitate TTC. A number of additional mechanisms have been proposed to explain the classical apical dysfunction observed in TTC. These include multivessel coronary vasospasm, aborted myocardial infarction with plaque rupture, thrombus formation and dissipation, vasospasm of a large “wrap-around” left anterior descending coronary artery (LAD) anatomy, or direct catecholaminergic effects on the myocardium. However, the exact pathophysiology of TTC remains to be clearly defined.
All available studies report a marked female predominance in TTC, affecting postmenopausal women in 90% of cases, with an age range of 60 to 75 years. Men are in a similar age range and account for about 10% of the patient population. Overall, <10% of TTC patients are below 50 years of age. The exact incidence of TTC is unknown, at least in part as a result of widespread underdiagnosis. Among patients undergoing coronary angiography for suspected acute coronary syndrome, about 2% are diagnosed as having TTC ( Figure 1 ).
Patients with TTC have generally a good short-term prognosis, with a rapid improvement of LV systolic function within a few days or weeks. A variety of complications may occur in the acute phase of the disease, such as acute heart failure with pulmonary edema or cardiogenic shock, intraventricular pressure gradients, acute functional mitral regurgitation (MR), right ventricular (RV) dysfunction, intraventricular thrombi resulting in stroke or arterial embolism, atrial fibrillation, or malignant ventricular arrhythmias. Rarely, perforation of the LV free wall or the interventricular septum has been described. In large studies, in-hospital mortality was observed in about 2% of patients with TTC. On the other hand, newer studies suggest that both short- and long-term mortality may be substantial and similar to that of patients with acute coronary syndromes.
Notwithstanding this, patients with TTC show more favorable outcomes than those with AMIs with a similar degree of LV dysfunction at presentation. The reported frequency of TTC recurrence ranges from 0% to 11.4%.
Prompt imaging of ventricular function is the key to making the diagnosis. Because of its widespread availability and feasibility in the acute care setting, echocardiography is frequently the first noninvasive imaging modality used to assess patients with TTC. It can provide useful information on LV morphology and regional and global systolic or diastolic function. In addition, features that have been associated with TTC, such as LV outflow tract obstruction (LVOTO), MR, and RV involvement, can also be detected. Echocardiography also allows the noninvasive assessment of coronary microcirculation impairment during the acute phase of TTC. In this review, we discuss each of these topics, with a special focus on clinical and prognostic implications.
Standard Echocardiographic Assessment
LV Systolic Function
In the acute phase, transthoracic echocardiography (TTE) usually identifies the characteristic LV morphology associated with TTC. Standard TTE allows the detection of the different LV morphologic patterns according to the localization of wall motion abnormalities (WMAs). In the majority of cases, WMA typically involve the apical and midventricular segments, which appear akinetic or dyskinetic (defined as “apical ballooning”) in contrast to the basal segments, which are often hyperkinetic. Several variant forms, such as “midventricular” or “inverted” TTC, have also been described. Midventricular TTC is characterized by akinesis of the midventricular segments, mild hypokinesis, or normal contraction of the apical segments and hypercontractility of the base. Inverted TTC is characterized by two different forms, the first defined as “apical sparing” with preserved apical function and severe hypokinesis of the remaining walls, and the second as “basal or reverse’ TTC with hypokinesis confined to the basal segments. The prevalence of these variant forms remains uncertain. In a large cohort of patients with TTC enrolled in the Tako-Tsubo Italian Network (TIN), midventricular and inverted TTC variants were detected in 18.1% and 4.8% of cases, respectively.
Kurowski et al . found no differences in demographic, clinical, angiographic, and laboratory parameters or outcomes in patients with typical apical TTC compared with those with midventricular TTC. Mansencal et al . reported that patients without apical involvement were younger, with less impaired systolic function and fewer signs of heart failure. Nevertheless, there were no differences among the several forms of TTC in LV function recovery at early and late follow-up.
In the acute phase, LV ejection fraction (EF) is reduced in patients with TTC and recovers with resolution of myocardial stunning. The degree of EF reduction varies according to the severity of myocardial impairment, the presence of comorbidities, and age. The magnitude of myocardial dysfunction is wide and irrespective of single coronary artery territory distribution, while the degree of biomarker release is quite small in proportion to the extent of WMAs. In fact, there is a discrepancy between the mild elevations in troponin levels and the extent of myocardial dysfunction. This contrasts with the situation in acute ST-segment elevation myocardial infarction, in which the peak troponin value is usually high in relation to the extent of regional dysfunction. On this ground, the product of peak troponin I level and echocardiographically derived LV EF (≥250) has recently been used as an index to differentiate TTC from ST-segment elevation myocardial infarction, with sensitivity and specificity of 95% and 87%, respectively. In a previous study, Dib et al . compared patients with TTC according to different electrocardiographic patterns at presentation and found no differences in clinical characteristics, EF, and outcomes.
In a systematic review of 28 case series, Pilgrim and Wyss reported a marked depression of LV EF on admission, followed by substantial improvement after 18 days on average (mean time range to recovery, 7–37 days; EF 20%–49.4% vs 59%–76% after recovery) in patients affected by TTC, suggesting the usefulness of this index in monitoring LV systolic function recovery. A marked reduction in LV EF on admission with improvement at short-term follow-up was also reported in the TIN registry (mean EF, 37.5 ± 5.2% vs 55.5 ± 7.1% at short-term follow-up ; Table 1 ). Furthermore, LV EF seems to be an independent predictor of major complications, providing additional information for the early identification of patients at higher risk, in particular those aged ≥75 years ( Table 2 ). EF < 40%, together with age > 70 years and the presence of a physical stressor, are the three criteria in the Mayo Clinic risk score for acute heart failure in TTC. Elderly patients demonstrate significantly delayed and lower LV systolic function recovery compared with younger patients.
|LV end-diastolic volume (mL)||91.9 ± 22.7||86.2 ± 22.3||.015|
|LV end-systolic volume (mL)||57.3 ± 15.2||38.2 ± 11.0||<.001|
|Indexed LV end-diastolic volume (mL/m 2 )||53.8 ± 14.3||49.9 ± 13.3||.002|
|Indexed LV end-systolic volume (mL/m 2 )||33.3 ± 9.1||21.8 ± 6.1||<.001|
|LV EF (%)||37.5 ± 5.2||55.5 ± 7.1||<.001|
|Wall motion score index||1.8 ± 0.2||1.1 (1–1.3)||<.001|
|Left atrial volume (mL)||43.7 ± 8.2||41.5 ± 7.9||<.001|
|E peak velocity (cm/sec)||75.2 ± 18.0||67.5 ± 15.1||<.001|
|A peak velocity (cm/sec)||72.6 ± 21.1||84.6 ± 19.5||<.001|
|E/A ratio||1.0 (0.7–1.3)||0.7 (0.6–0.9)||<.001|
|DTE (msec)||202.0 ± 55.9||230 (190–254)||<.001|
|e′ peak velocity (cm/sec)||7.0 ± 2.3||9.3 ± 2.0||.012|
|E/e′ ratio||11.0 ± 4.0||7.1 ± 2.2||<.001|
|sPAP (mm Hg)||40.3 ± 10.4||28. ± 7.1||.030|
|TAPSE (mm)||19.2 ± 3.2||21.1 ± 2.8||<.001|
|RV area change (%)||38.7 ± 7.1||41.4 ± 3.8||.001|
|Moderate to severe MR||49 (21.5%)||10 (7.6%)||<.001|
|LVOTO||29 (12.8%)||3 (1.3%)||<.001|
|RV involvement||33 (14.5%)||2 (0.9%)||<.001|
|Variable||Univariate analysis||Multivariate analysis|
|Wald χ 2||P||HR||95% CI||Wald χ 2||P||HR||95% CI|
|Moderate to severe MR||23.532||<.001||5.916||2.885–12.133||5.049||.025||3.254||1.163–9.109|
Extensive apical myocardial dysfunction and reduced intraventricular systolic flow velocity are predisposing factors for thrombus formation. Mural or pedunculated thrombi can be visualized at the apex in 1% to 2% of patients with TTC during the first 2 days, causing stroke or systemic embolization (renal or lower limb embolism) in approximately one-third of them. Once thrombus has been documented, therapy with heparin followed by oral anticoagulants should be instituted, and serial TTE should be performed until thrombus resolution and myocardial contractility recovery. Furthermore, in case of diagnostic uncertainty, use of contrast agents or real-time three-dimensional echocardiography (RT3DE) may be helpful, especially in detecting small thrombi.
Wall motion pattern in TTC tends to involve the apical and midventricular myocardial segments circumferentially in contrast to AMI, in which only the territories supplied by the “culprit” coronary artery are primarily involved. Such a “circumferential pattern” of LV myocardial dysfunction characterized by symmetric WMAs involving the midventricular segments of the anterior, inferior, and lateral walls should be considered peculiar to TTC and included in the differential diagnosis between TTC and acute coronary syndromes. These findings support the hypothesis of diffuse ventricular dysfunction secondary to myocardial stunning underlying the pathogenesis of TTC.
Contractility of the apical anterolateral segments seems to recover earlier. Complete recovery may occur as early as at the time of hospital discharge but generally within few months. Despite the absence of significant obstructive coronary artery disease at coronary angiography, in case of persistent WMA at mid- or long-term follow-up, cardiac magnetic resonance (CMR) may be useful to rule out myocardial necrosis.
LV Diastolic Function
Although LV diastolic dysfunction in TTC has been reported in several studies, recent experimental and human data have demonstrated low LV end-diastolic pressure and systemic vascular resistance in patients with TTC. Global and regional diastolic dysfunction has been observed in some patients during the early phase of TTC, as evidenced by impaired LV untwisting and increased E/e′ ratio. Despite controversial data in patients with heart failure, E/e′ ratio may be a practical and reproducible echocardiographic index for assessment of LV filling pressure. In the TIN registry, E/e′ ratios were higher in patients with major adverse events than in those without and were found to be a significant independent predictor of acute heart failure and in-hospital mortality on multivariate analysis ( Table 2 ). Keeping in mind that acute heart failure is the most common early complication of TTC, E/e′ ratio should be assessed early and systematically to identify patients at higher risk for hemodynamic instability and to guide appropriate management. Furthermore, given that diastolic dysfunction is transient and reversible, the improvement in E/e′ ratio at follow-up may be considered an additional useful indicator of LV function recovery ( Table 3 , Figure 2 ).
|Coronary flow in distal LAD|
|Moderate to severe MR|
|Intraventricular thrombus detection|
LV Outflow Tract Obstruction and Mitral Regurgitation
In older postmenopausal women with small left ventricles and septal bulge, LVOTO may result from basal hypercontractility, as occurs in the typical forms of TTC, and may be precipitated or augmented by catecholamine administration. Echocardiographic evidence of significant LVOTO (defined as an intraventricular gradient ≥ 25 mm Hg) ( Figure 3 ) in patients with TTC has important therapeutic implications, in particular for patients with advanced systolic heart failure. In this patient subset, inotropic agents causing enhanced basal contractility and diuretics inducing volume depletion may increase the intraventricular pressure gradient with subsequent hemodynamic instability, ultimately leading to cardiogenic shock. Thus, in patients with cardiogenic shock and concomitant LVOTO, inotropic agents and excessive dehydration should be avoided. In such a cohort, monitoring of hemodynamic conditions and systemic vascular resistance may be helpful to guide therapy. In case of elevated adrenergic tone, the use of low-dose β-blockers should be considered with caution. Conversely, the use of phenylephrine, an α 1 -receptor agonist, may be an option in patients with associated low vascular resistance.
Although intra-aortic counterpulsation usually results in reduced afterload with subsequent improvement of cardiac output, its use should be considered with caution. Afterload reduction, particularly at the onset of systole, may favor the occurrence of unfavorable conditions in this population, which is susceptible to dynamic LV outflow tract gradient because of the hyperdynamic function of the LV base.
In some fulminant cases with rapid worsening of hemodynamic conditions and ensuing multiple-organ failure, venoarterial extracorporeal membrane oxygenation or the implantation of an LV or a biventricular assist device as “bridge therapy” may be a useful option.
LVOTO is associated with increased LV afterload and systolic wall stress leading to subendocardial ischemia and acute myocardial stunning. Initially, LVOTO was thought to be the unique pathogenetic mechanism of TTC characterized by apical ballooning, but this hypothesis was later disproved in large series of patients with TTC, demonstrating a low prevalence of LVOTO (12.8%–25%). More attention should be paid in patients with TTC and severe LVOTO because of the occurrence of life-threatening arrhythmias and fatal LV wall rupture. The degree of reversible LVOTO is variable and depends on loading conditions and may be associated with systolic anterior motion of the mitral valve (SAM), leading to MR. In the TIN registry, LVOTO was more common in patients with major adverse events ( P = .006; Table 1 ). Of note, 17 patients had concomitant SAM and significant MR. Moderate or severe reversible MR was detected in several studies in about one-fifth of patients with TTC ( Figure 4 ). Parodi et al . first described reversible MR and its association with advanced Killip class in patients with TTC. LV EF on admission and SAM were the only independent predictors of acute MR.
The mechanisms underlying MR in TTC are not completely elucidated. SAM has been reported in 33% to 50% of patients with TTC with significant MR. The occurrence of SAM, especially in case of an intraventricular pressure gradient, further worsens LV pump failure and often induces LVOTO. Significant reversible MR has also been described in patients with severe reductions in LV EF and higher LV volumes as well as impaired wall motion score indexes in the absence of SAM. These findings suggest that symmetric tethering of the mitral leaflets secondary to papillary muscle displacement in a dilated and dysfunctional apical left ventricle may have a role in the genesis of MR.
Although in a previous small series, MR appeared not to significantly influence patient outcomes, in the TIN registry, MR definitely emerged as a powerful prognostic marker associated with cardiogenic shock and in-hospital mortality ( Table 2 ).
Because of the negative impact on hemodynamic stability and its therapeutic and prognostic implications, LVOTO and significant MR should be ruled out in the early phase of TTC, especially if a new murmur is audible. In critically ill patients with TTC with poor acoustic windows, transesophageal echocardiography should be performed.
Because of its negative impact on hemodynamics and cardiac morbidity, special attention should be devoted to detect RV involvement in TTC, particularly in patients with electrocardiographic signs of “RV strain pattern.”
The prevalence of RV involvement in the TIN registry was 14.5% ( Table 1 ). However, its real incidence is probably underestimated. To overcome the intrinsic limitations of conventional echocardiography in assessing the complex RV anatomy, multiple echocardiographic windows and even off-axis views should be acquired. In patients with “biventricular ballooning,” the pattern of RV contraction mirrors that of LV walls. This is the opposite of the classic echocardiographic appearance of RV apical hyperkinesis and basal akinesis, known as McConnell’s sign, that typically occurs in patients with acute and massive pulmonary embolism. For this reason, Liu and Carhart named the peculiar pattern of RV involvement in TTC “reverse McConnell’s sign” ( Figure 5 ). RV dysfunction represents an additional finding that can assist in differentiating TTC from anterior AMI.
Although RV involvement has previously been associated with a higher incidence of congestive heart failure, cardiopulmonary resuscitation, use of intra-aortic counterpulsation, and longer hospital stays, no association with short-term cardiac morbidity or mortality was reported by Fitzgibbons et al . In the TIN experience, RV involvement was significantly more prevalent in patients with in-hospital major complications (28.8 vs 9.5%, P < .001). Of note, no significant differences in tricuspid annular plane systolic excursion between patients with and without major adverse events were found.
Because of hyperkinesis of the basal RV segments, serial evaluation of two echocardiographic indexes commonly used in daily practice, such as tricuspid annular plane systolic excursion and tissue Doppler velocity of the tricuspid annulus, could fail to detect impairment of RV function.
Advanced Echocardiographic Assessment
Myocardial Deformation Imaging
The profound transient LV dysfunction seen in TTC disturbs LV mechanics as well ( Figures 6–8 ). Most data on myocardial deformation imaging in TTC are provided by two-dimensional strain with speckle-tracking echocardiography, which allows the assessment of multidirectional LV deformation due to the complex myocardial architecture (longitudinal and circumferential shortening, radial thickening, and twisting). During the acute phase, there is a circumferential impairment of apical and midlongitudinal strain in typical TTC with a base-to-apex gradient of strain, as well as a circular injury of midradial strain. In comparison with anterior AMI, middle and apical longitudinal strain is more severely impaired in TTC as a result of lower values of strain in the inferior, posterior, and lateral segments ; this difference with anterior AMI accounts for the extent of LV dysfunction far beyond the LAD territory in TTC. However, this contractility impairment become asymmetric after some degree of recovery has occurred few days later, making the distinction from AMI more challenging at this stage. Given its higher sensitivity in detecting subtle abnormalities compared with more traditional parameters such as LV EF and wall motion score index, speckle-tracking echocardiography can display altered LV mechanics at the base of the heart, highlighting that myocardial impairment extends beyond the segments of WMAs even in typical TTC. All components of LV twist mechanics involving the systolic and diastolic phases (LV twist, LV untwisting rate) are impaired in the acute phase of TTC, as in anterior AMI, but the damage, due mainly to a loss of apical function, seems less profound in the latter group, the difference between TTC and AMI being significant in patients with ST-segment elevation at presentation. Counterclockwise rotation of the apex is the normal pattern. In several cases of TTC, there is a loss of the normal pattern of apical counterclockwise rotation with an abnormal clockwise rotation, thus severely impairing LV twist and untwisting. This remarkable alteration observed in TTC is entirely reversible, indicating predominant functional impairment, whereas in AMI, such impairment suggests structural injury with irreversible damage. The efficiency of LV twist is important not only for optimizing LV systolic function but also for boosting LV filling by subsequent untwisting. Early diastolic untwisting rate, which occurs during LV relaxation, is transiently diminished in TTC, suggesting that at least regional diastolic dysfunction is also present at the acute phase of TTC. This finding is also illustrated by the vector velocity imaging technique, which showed a transient injury of diastolic strain rate in involved segments in five patients with TTC. However, compared with patients with AMI, those with TTC had better regional diastolic function in the acute phase despite worse systolic dysfunction, as assessed by apical early diastolic strain rate on color-coded Doppler tissue imaging. To summarize, there is a transient multidirectional impairment of myocardial deformation in the acute phase of TTC with longitudinal, circumferential, and radial damage as well as LV twist mechanics. This transmural dysfunction, which is more profound than in AMI in some aspects, is entirely reversible, whereas in AMI, strain impairment predicts myocardial viability and adverse LV remodeling. These findings suggest that the mechanisms of myocardial stunning in TTC differ from the scenario observed in AMI. Finally, myocardial deformation has demonstrated its prognostic value in AMI, whereas such predictive information is still lacking in TTC.