Takotsubo syndrome (TTS), also known as takotsubo cardiomyopathy, stress-induced cardiomyopathy, or apical ballooning syndrome, is an acute and transient heart failure syndrome originally reported by Dr Sato in 1991 in a Japanese textbook and by Pavin et al. in Europe in 1997 [1, 2]. Chest pain and/or dyspnoea are the most common symptoms at presentation, whereas diaphoresis and syncope are less frequently observed [3]. TTS generally occurs in post-menopausal women, though in several registries and case series up to 10% of TTS patients are male. Although stressor events (emotional or physical) usually trigger the clinical onset of TTS, in about one-third of cases TTS has also been described without a preceding trigger.

Recently, a classification of TTS according to the type of triggering event has been proposed (Table 1) [4].

Electrocardiographic abnormalities (ST-T elevation in the majority of cases) resembling those detectable in acute coronary syndromes are common, as well as increased biomarkers reflecting myocyte damage despite unobstructed coronary arteries. For this reason, TTS has been classified as ‘myocardial injury’ in the last universal definition of myocardial infarction and categorized as myocardial infarction with non-obstructive coronary arteries (MINOCA) [4]. The pathophysiology of TTS is still unclear. Neuroendocrine, metabolic, genetic, and inflammatory factors via increased adrenergic stimulation and high level of catecholamine release seem to be involved in the genesis of the reversible myocardial stunning associated with this fascinating syndrome [3]. Initially, multivessel coronary spasm was suspected as a possible cause of this unique wall motion defect, typically characterized by apical ballooning. Iga et al. described the first echocardiographic findings in eight cases with transient left ventricular (LV) segmental asynergy and suggested asynergy was unrelated to coronary artery disease [6]. Now, it is well recognized that a ‘takotsubo-like’ appearance represents only a part of this syndrome and many variant forms have been reported [7–9]. Despite its transient nature, the acute phase of TTS is characterized by a substantial incidence of adverse events such as acute heart failure, cardiogenic shock, and arrhythmias, and considerable in-hospital death (4–5%) [10, 11]. Even at long term, TTS recurrence, cardiac and non-cardiac disorders, and increased mortality have been reported.

Owing to the wide spectrum of clinical presentations, diagnosis of TTS is often challenging. However, early recognition of this syndrome is key to adopting an appropriate therapy. Multimodality imaging is helpful in reinforcing the clinical suspicion of TTS in the early phase, allowing confirmation of the diagnosis, even retrospectively, after ruling out other clinical entities that should be considered in the differential diagnosis. The aim of this consensus document is to discuss and review the utility of multimodality imaging, including left ventriculography, echocardiography, computed tomography (CT), magnetic resonance imaging, and nuclear imaging, in the diagnostic workup of TTS.

Several diagnostic criteria for TTS have been proposed, including those issued by the Mayo Clinic, the Japanese guidelines [12], the Tako-tsubo Italian Network, the Gothenburg group, and the Heart Failure Association (HFA) TTS Taskforce of the European Society of Cardiology (ESC). Recently, the InterTAK diagnostic criteria have also been developed [3] that incorporate several different aspects: (i) right ventricular (RV) involvement and other atypical wall motion abnormalities (WMAs); (ii) emotional or physical stress are no longer mandatory features; (iii) neurological disorders and pheochromocytoma are considered as potential triggers for TTS; and finally (iv) the possibility of coexisting significant coronary artery disease and TTS has been confirmed (Table 2). According to the 4th universal criteria of myocardial infarction, the diagnosis of TTS should be based on the absence of coronary artery disease, or, if present (about 15% of cases), it should not be sufficient to explain the observed pattern of regional WMAs. Additionally, criteria for diagnosis of types 1, 2, and 4 myocardial infarction should be excluded [5].

Although initial studies suggested that prognosis of TTS is benign (in-hospital death 1–1.7%) [13, 14], recent data have shown higher in-hospital mortality rates (3.5–5%) [9, 15, 16]. Clinical characteristics that were associated with in-hospital adverse events or death included male gender, the presence of physical triggers, acute neurologic or psychiatric disease, the first troponin level, and a LV ejection fraction (EF) of <45% [7]. Multiple unfavourable echocardiographic findings (described in the next sections) were associated with in-hospital complications, with evidence of RV involvement being associated with poor long-term outcomes [15].

Coronary arteries are generally normal or near-normal in TTS. Although echocardiography is the first-line imaging modality in patients with suspicion of TTS, assessment of coronary anatomy is crucial in TTS diagnostic workup to rule out alternative diagnoses. As recommended by current guidelines on acute coronary syndromes (ACS), cardiac catheterization should be performed in all patients with acute cardiac ischaemia and ST-segment elevation or in patients without ST-segment elevation according to the individual risk profile, especially for patients with a low-intermediate probability of TTS [10, 17–21].

Coronary anatomy should be carefully evaluated using multiple angiographic views, ensuring that all lesions have been assessed in at least two orthogonal projections. Obstructive atherosclerotic plaques can be observed in about 1 in 10 of TTS patients [3, 18]. It must be emphasized that obstructive single-vessel coronary lesions are not an absolute exclusion criterion for diagnosis, since LV WMAs usually extend beyond a single epicardial vascular distribution in TTS, whilst equally non-obstructive plaques can cause myocardial infarction (Fig. 1) [10, 18, 19].

The diagnosis of TTS can also be established during cardiac catheterization through LV opacification. Biplane left ventriculography in the right and left anterior oblique projections allows for the assessment of nine LV segments and differentiation of TTS from anterior ST-elevation myocardial infarction (STEMI) in the majority of cases [19, 22]. The ‘apical nipple’ sign, a very small zone with preserved contractility of the LV apex, has been described on left ventriculography in about 30% of patients with TTS and typical apical ballooning (Fig. 2) [23]. This sign can be a useful additional tool to discriminate TTS from acute anterior STEMI, in which the phenomenon is not observed. Similarly, in patients with the mid-ventricular variant, systolic contraction of the apex can configure the hawk’s beak appearance by ventriculography (Fig. 3) [24, 25]. In addition, the LV chamber silhouette depicted by the right anterior oblique view may show the typical apical ballooning or one of the variant morphologic patterns, which are suggestive of TTS diagnosis (Fig. 4) [7, 26]. Given the transient and reversible course of WMAs, left ventriculography should be performed in all patients with suspected TTS, especially if no transthoracic echocardiogram is available or in patients with poor acoustic window [8]. LV opacification may also be useful to identify possible mechanical complications such as acute mitral regurgitation (MR) or LV apical thrombi, which have prognostic implications [14]. Before removal, the catheter should be withdrawn slowly from the LV cavity in order to assess invasively intraventricular pressure gradients and possibly detect LV outflow tract obstruction (LVOTO), particularly in patients with haemodynamic instability [26].

Although not specific for TTS, coronary anatomical variants, including myocardial bridging or long and tortuous arteries, are common findings on CA [27]. Likewise, a high thrombolysis in myocardial infarction frame count is frequently observed and has been associated with coronary microcirculation dysfunction with consequent slow coronary flow [28, 29]. Beyond the identification of significant coronary lesions using conventional angiography, TTS diagnosis requires the exclusion of any other epicardial mechanisms potentially involved in other types of MINOCA such as plaque rupture or erosion and above all spontaneous coronary artery dissection [30]. The use of intravascular imaging modalities, e.g., optical coherence tomography and intravascular ultrasound, may help to clarify alternative aetiologies of myocardial injury [31, 32]. However, given the technical complexity and high costs, the application of these tools should be considered only in selected cases with unclear or inconclusive angiographic findings.

Transthoracic echocardiography (TTE) is the first-line imaging modality for the evaluation of patients with TTS [33]. If suspicion of TTS arises, echocardiography should be promptly performed because it provides useful information about systolic and diastolic function, RV involvement, haemodynamic status, mechanical complications, pulmonary artery systolic pressure, pericardial effusion, and LV intraventricular thrombi [26]. Echocardiographic findings are also helpful to identify patients at higher risk of adverse outcome and to monitor regression of WMAs (Tables 3 and 4).

Strain imaging by speckle tracking echocardiography allows assessment of global and regional myocardial function in all three layers of myocardium, and which can be applied to assess strain in the longitudinal, circumferential, and radial directions. Owing to the ability of strain imaging to objectively depict the transient impairment of myocardial deformation with circular involvement of the opposite LV walls (circumferential pattern), speckle tracking assessment in TTS patients should be performed, whenever possible, even in the acute phase. Furthermore, it is a useful tool to monitor myocardial functional recovery [63–73] (Fig. 7). LV twist is also transiently impaired in the acute phase of typical TTS, mainly due to apical dysfunction, with secondary alterations in the normal rotational pattern of contraction [74] (Fig. 8). LV early untwisting rate, an index of diastolic function, is also decreased in the acute phase. Recent studies have further demonstrated RV involvement in TTS as an important finding associated with worse outcome [75–80]. Traditional echocardiographic parameters used to assess RV function, such as RV fractional area change and tricuspid annular plane systolic excursion, do not account for contraction abnormalities of the mid- or apical RV segments and are instead predominantly affected by basal hyperkinesis. RV free wall longitudinal strain is therefore recommended in order to detect even small areas of myocardial dysfunction hardly recognizable by visual assessment and/or by other traditional echocardiographic parameters. Conversely, RV global longitudinal strain should not be assessed because involvement of the interventricular septum can be misleading [81].

WMAs and LVEF usually recover to normal within a few weeks. However, several studies have shown that significant mechano-temporal alterations affecting both systole (global longitudinal strain and apical circumferential strain) and diastole (untwist rate and time to peak untwisting), reduced exercise capacity, increased native T1 mapping values and impaired cardiac energetics can persist for months after the acute phase, despite normalization of LVEF and volumes [67, 82].

The long-term prognosis of patients who experienced TTS seems to be worse than that in the general population; the underlying mechanism may be partly due to persistence of oedema/inflammation, with residual functional abnormalities observed on strain imaging [16, 46, 68, 74].

Together with other clinical, functional, and metabolic testing, these studies provide evidence that structural and functional myocardial abnormalities often persist after TTS. These findings do not rule out the diagnosis of TTS and emphasize the need for systematic follow-up. At 1 month from TTS onset, a clinical and echocardiographic follow-up is mandatory: in case of persisting WMAs, ECG changes, elevated levels of B-type natriuretic peptide and/or symptoms, a second follow-up visit should be scheduled at 3 months, conversely the patient should be evaluated at 6 months and 1 year, and then annually. Careful evaluation of persistent systolic and diastolic alterations using speckle tracking echocardiography could add important prognostic information and aid clinical decision-making.

Visualization of coronary flow in the distal part of the LAD by transthoracic Doppler can help to differentiate TTS from acute myocardial infarction (AMI) due to LAD occlusion (Fig. 9) [9, 83–85] and also provides non-invasive physiological information on coronary blood flow. Through the ratio of hyperaemic to baseline diastolic peak flow velocity in the distal LAD, this method also allows the assessment of coronary flow reserve after administration of coronary vasodilator agents (e.g. as adenosine or dipyridamole) [86]. In the acute phase of TTS, coronary flow reserve is reduced due to transient coronary microcirculation impairment and recovers prior to LV wall motion normalization [85, 87–90]. The decrease in coronary flow reserve in TTS seems to be milder than that in AMI, although TTS has worse WMAs than AMI, suggesting that microvascular dysfunction is not the only mechanism of TTS [91].

In daily practice, this tool is of limited value, although it may be useful for monitoring the recovery of coronary flow reserve in doubtful cases.

In clinical practice, the most important application of contrast echocardiography is LV cavity opacification for the detection of WMAs [92] (Fig. 10) and the determination of the morphologic pattern of TTS in patients with poor acoustic window. Furthermore, the use of ultrasound contrast agents improves the detection of intraventricular thrombi, directing prompt introduction of anticoagulant therapy [9]. Finally, contrast echocardiography may be used to improve visualization of coronary flow by TTE with colour and pulsed wave Doppler [83, 85,⁸⁶].

Of note, myocardial contrast echocardiography performed within 24 h after TTS onset revealed reversible perfusion defects in involved segments, which improved faster than WMAs [91].

3D echocardiography is a useful method to quantify chamber measurements with higher accuracy and reproducibility than 2D echocardiography [93] (Figs. 11 and 12). 3D visualization may enable more objective assessment of regional WMAs and EF, especially in the right ventricle because of its complex morphology [94]. As 2D echocardiography has a limited ability to detect RV involvement [79], 3D echocardiography might help identify RV involvement [95] and improve risk stratification in patients with TTS. However, there is a paucity of data and further studies are warranted.

These ‘advanced’ echocardiography techniques, including strain imaging, coronary flow assessment, contrast echocardiography, and 3D echocardiography, are emerging and hold promise in improving our understanding of the pathophysiology underlying TTS. However, it should be acknowledged that the clinical utility of these techniques has not been fully established and physicians should not make clinical decisions solely on this basis.

Although echocardiography and left ventriculography have mainly been used to diagnose LV WMAs in patients with TTS, CMR plays an important role in the comprehensive assessment of the functional and structural changes that occur in patients with clinical suspicion of TTS [96]. CMR represents the reference standard for qualitative and quantitative assessment of regional WMAs and accurate quantification of right and LV volumes and function. Other additional abnormalities, including pericardial effusion, pleural effusion, and intraventricular thrombi, can also be easily appreciated. However, the major strength of this imaging technique is myocardial tissue characterization, due to its unique ability to provide different markers of myocardial injury. The detection of oedema/inflammation rather than fibrosis or necrosis is useful to distinguish reversible from irreversible myocardial damage and is helpful in guiding the differential diagnosis between TTS, other MINOCA presentations and ACS.

Technological developments over recent years have emphasized the role of coronary computed tomography angiography (CCTA) in the evaluation of patients with suspected TTS [109–111].

An advantage of both CCTA and CMR is the ability to evaluate the heart in any potential plane, overcoming the limitations imposed by suboptimal acoustic windows on echocardiography [112]. Moreover, unlike CA, this imaging modality is not invasive and, compared with CMR, can be readily used in the emergency setting due to its accessibility and fast acquisition time. The main indication for CCTA in TTS patients relates to the assessment of the epicardial coronary arteries and the ruling out of coronary artery disease, according to local expertise and availability [113].

CCTA has been proposed as a non-invasive alternative to CA in stable and pain-free patients with no ST-segment elevation at onset and convincing clinical (post-menopausal woman, trigger event) and echocardiographic (apical ballooning, circumferential pattern, distal LAD flow visualization) findings consistent with TTS. This approach could be advantageous, particularly in patients with delayed presentation (>48 h after chest pain onset) or when CA is not available on site [8, 17].

Cardiac CT, beyond coronary artery disease, allows assessment of LV function and WMAs. Nowadays, cardiac CT also provides myocardial characterization with late iodine enhancement imaging comparable with that obtained from CMR [114].

Owing to multiplanar imaging [115, 116] that overcomes the limitations of echocardiography, multidetector CT permits a better identification of LV apical thrombi, a not so rare complication of TTS. Cardiac CT plays an important role in the diagnostic workup of patients with acute chest pain and doubtful TTS diagnosis to exclude other conditions such as pulmonary embolism and aortic dissection [117, 118]. Recently, a comprehensive assessment of the coronary arteries, LV function, and myocardial late iodine enhancement has been performed in patients with suspected TTS [119], thus pointing to the growing importance of CCTA.

Up to date there is no evidence for delaying invasive CA in patients with ongoing acute chest pain and acute ECG changes. In such a clinical scenario, patients should be managed according to current guidelines for STEMI or ACS until culprit coronary artery disease has been ruled out [20, 21].

CCTA may be chosen instead of CA:

Moreover, CCTA can be considered as a non-invasive valuable alternative in life-threatening conditions where CA is highly likely to cause complications (e.g. terminal malignancy and advanced age with frailty).

Although the role of nuclear imaging in TTS has not yet been well established in clinical practice, nuclear imaging can provide assessments of myocardial perfusion and metabolic activity, improving our understanding of the pathophysiology of TTS and potentially aiding diagnosis.

Some studies have reported that fatty acid metabolism depicted by ¹²³I-beta-methy-iodophenyl pentadecanoic acid (¹²³I-BMIPP) single-photon emission computed tomography (SPECT) was more severely impaired than myocardial perfusion depicted by ²⁰¹thallium scintigraphy in acute TTS patients [28, 120]. Reductions in perfusion tracer counts occur as a result of regional myocardial wall thinning at the apex, due to both artefacts and partial volume effects, and may mimic ACS. Combined perfusion and metabolic imaging can help identify this phenomenon and aid the diagnosis of TTS (Fig. 18).

¹²³I metaiodobenzyl guanidine (MIBG) scintigraphy can assess myocardial sympathetic nerve terminal activity and detect adrenal and ectopic pheochromocytoma. Even in the subacute phase, the adrenergic hyperactivity observed in TTS results in decreased uptake and increased washout of ¹²³I-MIBG from the heart (Fig. 19) [26, 121]. Indeed, myocardial ¹²³I-MIBG uptake is impaired for months as a consequence of regional disturbances in sympathetic neuronal activity [122]. The combination of ¹²³I-MIBG and myocardial perfusion scintigraphy is also useful for distinguishing TTS from ACS when innervation and myocardial scarring are matched.

Cardiac positron emission tomography (PET) using [¹⁸F]2-fluoro-deoxy-glucose (FDG) shows abnormal glucose metabolism alongside normal myocardial perfusion in TTS. Recently, some studies have been conducted at different phases during the TTS time course [123, 124]. In the acute and subacute phases, similar defects of MIBG and FDG uptake were found, despite slightly reduced perfusion in TTS. Although rapid normalization of myocardial perfusion was observed, TTS patients had delayed recovery of both LV glucose metabolism and sympathetic innervation. In the context of normalized LV wall motion, the delayed recovery of glucose metabolism and sympathetic innervation may allow diagnosis in patients with delayed presentation, even within a few months of the suspected triggering episode.

The pathophysiology of TTS remains challenging and controversial. Several hypotheses have been proposed to explain this unique cardiac disorder [52, 125–128], which can be broadly categorized as vascular or myocardial causes. Temporal evolution of MIBG and FDG abnormalities has provided new insight into functional myocardial changes and their time course in TTS; however, the mechanisms of TTS remain undetermined. The diagnostic criteria of TTS have been updated, such that coronary artery disease and pheochromocytoma are included in the new criteria. Therefore, nuclear imaging seems to be an ever more important imaging modality for the assessment of TTS, not only for diagnostic purposes, but also as a research tool that can allow better understanding of the pathophysiology of TTS.

Coronavirus disease 2019 (COVID-19) is a tremendous infectious disease firstly recognized in Wuhan, Hubei, China, that has spread throughout to other provinces in China and impetuously involved worldwide many countries causing a pandemic and a public health crisis of global proportions in 2020. Cardiovascular involvement, including acute coronary syndromes, myopericarditis, and pulmonary embolism, has been associated to the clinical course of a substantial proportion of patients affected by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Myocardial injury emerged as an important predictor of worst outcome. Shi et al. report a prevalence of 19.7% of cardiac injury, defined by high blood levels of high-sensitivity troponin I, among 416 patients with confirmed diagnosis of SARS-CoV-2 [129]. In this study, cardiac injury was associated with higher risk of in-hospital mortality. Their results are consistent with another study by Guo et al. on 187 Chinese patients, which highlighted the association between myocardial injury and fatal outcome in SARS-CoV-2 patients [130].

As with other coronaviruses, an intense systemic inflammatory response secondary to SARS-CoV-2 may produce demand ischaemia and/or possible coronary plaque disruption. Moreover, the detection of increased levels of multiple cytokines and chemokines in patients hospitalized in intensive care unit, suggests a direct myocardial damage as an alternative potential mechanism [131].

Noteworthy septic status, hypoxaemia, and metabolic acidosis; complications often detected in SARS-CoV-2 patients, are established triggers for secondary TTS.

Owing to the detection of myocardial inflammatory infiltrates documented in TTS patients, possible shared pathophysiological pathways in occurrence of myocardial dysfunction might be hypothesized in patients with SARS-CoV2 experiencing myocardial injury [82]. TTS and its possible complications such as acute heart failure, cardiogenic shock and life-threatening arrhythmias should be taken into consideration in differential diagnosis among various cardiovascular presentations associated with SARS-CoV-2 [132].

Imaging diagnostic tests are not routinely used in the emergency context of pandemic disease, the European Association of Cardiovascular Imaging recommend performing echocardiography at least in patients with laboratory or electrocardiographic signs of cardiac injury [133]. The suspect of TTS should arise in patients with echocardiographic evidence of extensive myocardial dysfunction especially if typical apical ballooning pattern is detected [134]. Clinicians should consider combining coronary CT with lung scanning (usually used for pneumonia disease) to exclude coronary artery disease. Multislice computed tomography (MSCT), being able to detect late iodine enhancement and myocardial fibrosis, is an additional tool to exclude myocarditis in this clinical context. Compared with MSCT, CMR is able to recognize myocardial oedema, a hallmark sign of TTS playing a key role in differential diagnostic work-up [135].

Although the current challenging clinical scenario makes serial invasive and not invasive diagnostic testing difficult, whenever possible multimodality imaging should be performed in patients with SARS-CoV-2 and cardiac injury to distinguish type II myocardial infarction, myocarditis, and TTS, and to adopt the most appropriate therapeutic strategy.

TTS is a complex clinical entity, with a wide spectrum of clinical presentations and varying levels of risk and haemodynamic stability, which requires tailored management strategies depending on the different scenarios. Multimodality imaging plays a key role in establishing the diagnosis, guiding therapy and stratifying prognosis in both the acute and post-acute phases of TTS (Figs. 20 and 21; Table 5). In clinical practice, echocardiography is the first-line imaging tool due to its widespread availability and feasibility in the acute care setting and the ability to easily and rapidly monitor changes in cardiac function over time. In asymptomatic or paucisymptomatic patients, the detection of unexplained extended apical WMAs, despite mild ST-T changes at ECG and low troponin elevation, is highly suggestive of TTS diagnosis, especially if predisposing factors exist. Patients presenting with acute chest pain and/or dyspnoea, after ECG and first clinical evaluation, can be screened by the InterTAK score. If the score is >70, echocardiographic findings compatible with apical or mid-ventricular ballooning, circumferential pattern and antegrade LAD flow, will orient the diagnosis towards TTS. In this case, coronary anatomy may be assessed invasively by CA or non-invasively by CCTA. Once CCTA and/or CA have confirmed the absence of culprit atherosclerotic plaques, the patient should be hospitalized and monitored in the intensive care unit for at least 48–72 h. TTE should be performed earlier in order to assess EF, cardiac output and, especially in unstable patients, to detect the onset of functional and mechanical complications (i.e. LVOTO, severe MR, and RV involvement). FU echocardiography is recommended daily or every two or three days during the first weeks, and at longer intervals after the acute phase. TTE allows monitoring of the recovery of transient WMAs and can stratify patients into ‘low risk’ and ‘high risk’, the latter requiring specific therapeutic measures. Being able to provide more detailed information about LV morphology and function, RV involvement, intraventricular thrombi, and pericardial effusion, CMR can be useful in the acute phase but its use is conditioned by the more limited availability, compared with echocardiography, and the difficulties in imaging unstable patients. Conversely, owing to its ability to detect the presence of oedema, a hallmark finding in TTS diagnosis, CMR is crucial during the post-acute phase to rule out other common pathologies (e.g. ACS and myocarditis) (Table 6). It should be performed in doubtful cases or in patients with persistent WMAs, even after discharge. Of note, recent studies have demonstrated that long-term FU in TTS patients is not uneventful. Further studies are needed to verify if assessment of subclinical myocardial dysfunction by speckle tracking echocardiography and other more sophisticated nuclear techniques (e.g. SPECT-MIBG, PET) can help identify patients that may be more prone to develop recurrences, life-threatening arrhythmias, or heart failure phenotype.