Takotsubo syndrome (TTS) typically presents with symptoms, electrocardiographic changes and mild elevation of cardiac enzymes that mimic myocardial infarction. It is characterized by transient left ventricular (LV) dysfunction extending beyond a single coronary territory, in the absence of obstructive coronary artery disease.1–3 The classic echocardiographic description is of apical akinesia, with or without mid-ventricular involvement, associated with basal hypercontractility; alternative patterns of regional dysfunction have been described, although less frequently.4,5 These features, which most frequently follow an emotional or physical trigger, were first described by Sato and colleagues in the 1990s.6 Initially believed to be a benign entity, it has been shown to be associated with non-negligible complications and an estimated 4-5% mortality.4,7–9 Its etiology is still somewhat elusive, but the most widely accepted hypotheses focus on catecholamine increase owing to sympathetic stimulation, which could cause myocardial dysfunction through adrenoceptor-mediated damage, direct toxicity, and endothelial dysfunction.10–12

Myocardial deformation analysis with speckle-tracking echocardiography has been shown to be more sensitive to lesser degrees of dysfunction than parameters such as ejection fraction, and is nowadays part of standard practice for detection of subclinical dysfunction in settings such as chemotherapy.13,14 Its use in TTS patients has been described in small reports, mainly focusing on serial assessment and possible persisting subclinical dysfunction.15–17 Some groups have also described abnormalities particular to this entity,18,19 however these findings have not been reproduced in larger studies.

Our goal was to characterize global and regional myocardial dysfunction in TTS using standard parameters as well as strain assessment with speckle-tracking echocardiography. A comparison with ST-segment elevation myocardial infarction (STEMI) sought to find patterns distinguishing between the two entities.

A total of 17 TTS patients were compared with 20 STEMI controls. Baseline demographic, clinical and laboratory features are summarized in Table 1. All TTS patients were female, with a mean age of 64 years (ranging from 33 to 84 years), while 19 of the STEMI patients were male, with mean age 57 years (37-76 years). The overall cardiovascular risk factor profile of the two groups was similar, except for current or previous smoking, which was more prevalent in STEMI than in TTS patients.

In the TTS group, 11 patients developed symptoms after an emotional stressor, two after a physical stressor, and four had no identifiable trigger; 16 presented with primary TTS and only one with secondary TTS. The electrocardiogram at hospital admission showed ST-segment elevation in six patients, T-wave inversion in four, ST-segment depression in one and other nonspecific changes in the remaining six patients.

All STEMI patients underwent emergent coronary angiography which showed a culprit left anterior descending (LAD) artery lesion (11 in the proximal segment, eight mid and one distal), with single-vessel disease in 11 of them (55%). Primary angioplasty with stent implantation was performed in 18 patients (90%), the others being two cases of spontaneous reperfusion, one with no need for intervention and the other with three-vessel disease for deferred surgical revascularization.

Laboratory tests showed that peak troponin I levels were higher in STEMI patients and peak brain natriuretic peptide (BNP) was higher in TTS patients. The BNP/troponin I ratio was also significantly higher in the latter, as previously reported.21 By contrast, neutrophil count and acute-phase reactant levels were not significantly different.

Transthoracic echocardiography on admission showed significant LV systolic dysfunction in TTS and STEMI patients, and ejection fraction was similar in the two groups (Table 2). Diastolic function parameters were also comparable in patients in both groups.

As with ejection fraction, LV global longitudinal strain was impaired in both groups, although with no statistically significant difference (-10.3±2.9 in TTS vs. -10.1±3.7% in STEMI, p=0.9).

Regarding regional function, longitudinal strain was worse in TTS patients in basal inferolateral and mid anterolateral segments and, conversely, worse in STEMI patients in the apical anteroseptal segment (Table 3).

Concerning the three LV slices, mean mid and apical longitudinal strain did not differ significantly between the two groups (Table 4). Basal longitudinal strain, however, was worse in TTS than in STEMI patients (-9.8±2.9 vs. -12.4±4.1%, p=0.010).

Analyzing the differences in dysfunction throughout the LV axis, the ratio between mean longitudinal strain in the basal and apical slices was significantly lower in TTS patients as well (1.51±0.86 vs. 2.94±1.88, p=0.006).

A subanalysis comparing TTS and STEMI patients with non-proximal LAD culprit lesions showed similar results, with worse basal longitudinal strain (-9.8±2.9 vs. -13.9±2.7%, p=0.002) and lower basal/apical ratio in the former (1.51±0.86 vs. 3.23±2.51, p=0.016).

A trend towards worse basal longitudinal strain was also seen in TTS compared to STEMI patients with proximal LAD culprit lesions, although this did not reach statistical significance (-9.8±2.9 vs. -11.3±3.0%, p=0.19). By contrast, proximal LAD patients had worse basal longitudinal strain than TTS (-8.4±5.2 vs. -4.9±2.2, p=0.046); moreover, this resulted in a still statistically significant lower basal/apical ratio in TTS patients (1.51±0.86 vs. 2.70±1.23, p=0.005).

Our study shows that, while LV dysfunction is obviously a key feature in both TTS and myocardial infarction patients, the former have worse mean longitudinal strain in basal segments. This finding challenges the eyeball appearance of basal hypercontractility, possibly related to the contrasting markedly akinetic mid and apical segments with a ballooning appearance. The basal strain impairment could be explained, first of all, by unfavorable loading effects and geometrical changes due to remote dysfunctional segments, as has been described in acute myocardial infarction.22 Additionally, the trend toward worse longitudinal strain in TTS patients in inferior, inferolateral and anterolateral segments, similar to the findings of Heggemann et al.,18 suggests more global myocardial dysfunction, as would be expected considering the different pathophysiological mechanisms. Still, we would be inclined to interpret these segmental results with caution, due not only to the number of patients but also to the statistical limitations of using multiple comparisons.

The difference between the two groups lost significance when STEMI patients with proximal LAD culprit lesions only were analyzed, which could be due to the limited sample size, as the numerical difference persisted. Nevertheless, the resulting lower basal/apical longitudinal strain ratio in TTS remained statistically significant both in the entire sample and in each STEMI subgroup, translating into a smaller ‘dysfunction gradient’, providing further evidence of more generalized myocardial impairment in these patients (Figure 1).

Figure 1.

Examples of polar maps of peak systolic longitudinal strain obtained from a Takotsubo syndrome patient (left) and an anterior ST-segment elevation myocardial infarction patient (right).

(0.16MB).

Unlike previous studies, we did not identify worse global LV systolic function in TTS than in STEMI patients as assessed by ejection fraction or global longitudinal strain.18,23 This might help explain why we did not find the more severe midventricular and apical impairment in TTS described by the same authors; nevertheless, this could also be an indicator of better comparability between the two groups, and thus increase confidence in the differences found. Neither do our findings support the existence of the ‘evil eye’ pattern of symmetrical apical involvement in TTS described by Sosa and Banchs.19 On the contrary, there was a trend towards worse longitudinal strain in all apical segments in STEMI patients.

Some limitations in our work should be acknowledged, first of all those inherent to any study with a retrospective design. Additionally, owing to the inclusion criteria used, our findings are only valid for classical TTS. The lack of gender matching is self-evident, but it is closely linked to the contrasting epidemiology of these diseases. The relatively small sample is also a limitation, as previously stated, that is related not only to the relative rarity of this diagnosis but also to the need to use information on patients whose echocardiograms were obtained from only one manufacturer, as speckle-tracking strain analysis is still not fully standardized across different platforms. Nevertheless, we believe these limitations are overcome by the importance of adding to knowledge on a disease about which our understanding still has important gaps.

One of the ultimate goals in TTS research is the ability to quickly and safely distinguish it from STEMI, thus enabling appropriate decisions on the appropriate timing for coronary angiography. Echocardiography is a readily accessible tool in clinical practice and, with appropriately validated specific features, would be the ideal means to this end. Right now, we believe that identifying potential parameters of interest and assessing them in different settings, as we sought to do, is essential for this objective, and will enable adequately powered prospective studies to be performed in the near future. Moreover, the identification of such distinctive characteristics is of great value in advancing our knowledge on the pathophysiology of TTS, which in turn is inextricably linked to research into new therapeutic strategies.

Our main findings are the worse mean longitudinal strain in basal LV segments of TTS compared with STEMI patients, and the smaller basal/apical longitudinal strain ratio in the former group. This smaller ‘dysfunction gradient’ compared to the localized phenomenon that is coronary artery disease is indicative of more generalized myocardial impairment than initially realized. If confirmed by further studies, it could help identify echocardiographic patterns specific to TTS and enhance both our knowledge of this disease and the clinical approach.

The authors have no conflicts of interest to declare.