Takotsubo syndrome: getting closer to its causes

Abstract

Takotsubo syndrome (TTS) accounts for between 1 and 4% of cases presenting clinically as an acute coronary syndrome. It typically presents as a transient cardiac phenotype of left ventricular dysfunction with spontaneous recovery. More dramatic presentations may include cardiogenic shock or cardiac arrest. Despite progress in the understanding of the condition since its first description in 1990, considerable questions remain into understanding underlying pathomechanisms. In this review article, we describe the current published data on potential underlying mechanisms associated with the onset of TTS including sympathetic nervous system over-stimulation, structural and functional alterations in the central nervous system, catecholamine secretion, alterations in the balance and distribution of adrenergic receptors, the additive impact of hormones including oestrogen, epicardial coronary or microvascular spasm, endothelial dysfunction, and genetics as potentially contributing to the cascade of events leading to the onset. These pathomechanisms provide suggestions for novel potential therapeutic strategies in patients with TTS including the role of cognitive behavioural therapy, beta-blockers, and endothelin-A antagonists. The underlying mechanism of TTS remains elusive. In reality, physical or emotional stressors likely trigger through the amygdala and hippocampus a central neurohumoral activation with the local and systemic secretion of excess catecholamine and other neurohormones, which exert its effect on the myocardium through a metabolic switch, altered cellular signalling, and endothelial dysfunction. These complex pathways exert a regional activation in the myocardium through the altered distribution of adrenoceptors and density of autonomic innervation as a protective mechanism from myocardial apoptosis. More research is needed to understand how these different complex mechanisms interact with each other to bring on the TTS phenotype.

Graphical Abstract

Sympathetic nervous system, Endothelin, Beta-receptors, Oestrogen, Endothelial function

1. The discovery

Takotsubo syndrome (TTS) is a clinical syndrome characterized by the presence of characteristic ventricular regional wall motion abnormalities (RWMA), most commonly apical ballooning, in the absence of attributable coronary artery disease (CAD). The first described case, in 1990, involved a patient with an initial suspected diagnosis of an acute coronary syndrome (ACS), but with evidence of unobstructed coronary arteries on angiography and the appearance of systolic left ventricular (LV) apical ballooning on LV angiography.¹ This latter characteristic led to the name Takotsubo, Japanese for fisherman’s bowl, mimicking the typical LV end-systolic morphology.

Over the years, TTS has been coined under multiple nomenclatures in the medical literature including ‘stress cardiomyopathy’, ‘apical ballooning syndrome’, and ‘broken heart syndrome’. However, as multiple triggers and causes are involved, the term TTS appears most appropriate.

2. Epidemiology

Historically, it has been difficult to accurately quantify the prevalence of TTS due to a high frequency of under-diagnosis as well as the presence of an often transient cardiac phenotype; nevertheless, with increasing recognition of the condition TTS reportedly accounts for 1–4% of all patients presenting clinically as ACS with chest pain and/or dyspnoea together with electrocardiographic (ECG) changes and elevated cardiac biomarkers.^2–5

There is a significant female pre-ponderance with up to 90% of cases occurring in females.^6–9 TTS frequently but not exclusively manifests in middle to late age, especially in those over 50 years of age and in post-menopausal women, although the youngest reported case was in a premature neonate.^7,10

Although TTS was first identified in Japanese patients, it is a pan-ethnic condition.¹¹ There are reported variations in prevalence across ethnicities with an increased incidence in Asians and Caucasians compared to Hispanics or Afro-Caribbean populations.¹² Furthermore, there are substantial differences in in-hospital outcomes and prognosis between Japanese and European TTS patients as well as in between African-American and Caucasian cohorts.^13,14 Importantly, ethnicity in isolation does not appear to independently confer adverse outcomes in TTS patients, with differences in outcomes driven by variations in the type of trigger (physical vs. psychological) and by disparities in comorbid conditions and socio-economic factors across ethnicities.^13,14

3. Clinical features

Patients may present with a variety of symptoms that mimic ACS with chest pain and/or dyspnoea being the most common.^9,15 Palpitations and syncope may also be amongst the presenting symptoms. More dramatic presentations can include out-of-hospital cardiac arrest (OOHCA) and cardiogenic shock, particularly in males.^7,16,17 Rarely, TTS is incidentally diagnosed in asymptomatic patients due to ECG abnormalities, troponin elevation, and characteristic imaging findings when under evaluation for alternative medical diagnoses including post-tonic-clonic seizures or acute neurological syndromes such as a cerebrovascular accident (CVA).¹⁸ Whenever a neurological event triggers TTS, the neurological insult and TTS are commonly diagnosed contemporaneously and often on the same day.¹⁹

Given the considerable overlap in signs and symptoms with ACS, TTS is often diagnosed in the catheterization laboratory after invasive coronary angiography excluded the presence of attributable CAD. TTS therefore may be suspected by clinical and ECG characteristics on admission, but confirmed at coronary angiography with typical features on the left ventriculogram or transthoracic echocardiogram (TTE), or less commonly on cardiac magnetic resonance imaging (cMRI; Figure 1). It is important to note that an absence of obstructive CAD at the time of angiography is common, but not mandatory for a TTS diagnosis as patients may have bystander CAD not accounting for the degree of LV systolic dysfunction or the extent of RWMA.^3,20 In fact, up to 25% of TTS patients have concomitant obstructive CAD, the presence of which is an independent predictor of 30-day mortality.²¹ Consequently, type 1 myocardial infarctions (MI) are not mutually exclusive to a TTS diagnosis (Figure 2).

Figure 1

Co-existence of Takotsubo syndrome and coronary artery disease. Images (A–C) Patient 1, (D–F) Patient 2, (G–I) Patient 3, (J–K) Patient 4. Patient 1 and 2 had an inferior STEMI and apical TTS. Patient 3 had a lateral myocardial infarction and mid-ventricular TTS. Patient 4 had an ACS presentation with a finding of unobstructed coronaries and apical TTS on ventriculography. Figures C, F, and I demonstrate wall motion schematics (red: diastole, white: systole, blue dashed line: hypo-/akinesia). Images A-I from Eur Heart J 2020; 41(34): 3255–3268 and Images J–K from Eur Heart J 2016; 37, 2816–2820. Note: Reference to colour is essentiel to understand the figure; permission from OUP obtained.

Figure 2

Arrhythmic complication in takotsubo syndrome. Left ventriculography (antero-posterior view) showing the typical apical ballooning patternwith akinesia of the mid-apical segments and hyperkinesia of the basal segment (A). A 12-lead electrocardiogram recorded at the third day ofhospitalization showing giant negative T-waves in leads aVL, L1, L2, aVF and V4–V6, markedQT prolongation (QTc = 552ms) and ‘R on T’ prematureventricular beats (B). Telemetry recording of a pause-dependent (‘long-short sequence’) torsade-de-pointes/ventricular fibrillation, which requiredelectrical cardioversion (C)." Reprinted with permission from Migliore et al. Int J Cardiol 2013;166:261–263. Note: Permission obtained.¹⁶⁵

Different sub-types of TTS exist based on the location of the RWMA including apical, basal, focal, and mid-ventricular presentations.²² The right ventricle may also be involved in up to a third of TTS patients and rarely may be the only myocardial chamber involved.^22–24

TTS diagnostic criteria have been proposed by many.²⁵ They all share similarities but also have subtle differences. The International (InterTAK) Diagnostic Criteria were developed by international experts to provide universally accepted diagnostic criteria (Table 1).³ The InterTAK Diagnostic Score, based on seven clinical and ECG parameters, has also been designed to allow early and rapid estimation of the probability that a particular clinical presentation is associated with TTS rather than an ACS (Table 2).^3,5

Table 1

The InterTAK Diagnostic Criteria. Adapted from Ghadri et al., Eur Heart J (2018)³

	Characteristics
1	Transient LV systolic dysfunction (akinesia/dyskinesia/hypokinesia) resulting in basal, mid-ventricular, focal, apical ballooning or other LV RWMA with or without right ventricular involvement. The RWMA usually extend beyond a single epicardial coronary vessel distribution however rare cases can exist where RWMA is localized to the myocardial territory of a single coronary artery (focal TTS)
2	Symptoms and signs are preceded by emotional, physical or a combined trigger (not essential)
3	Neurological disorders (seizures, CVA/TIA, subarachnoid haemorrhage) as well as pheochromocytoma may be triggers for TTS
4	New ECG abnormalities (ST-segment elevation, ST-segment depression, T-wave inversion, QTc prolongation) may be present but are not essential as rare cases can exist without ECG changes
5	Patients have moderately elevated cardiac biomarkers (troponin/creatine kinase) in most cases; there are significantly elevated BNP levels commonly
6	The presence of significant CAD does not exclude a diagnosis of TTS
7	Patients have no evidence of infectious myocarditis
8	Post-menopausal women are predominantly affected

Table 2

The scoring system for the InterTAK Diagnostic Score to help differentiate the likelihood of an acute presentation being related to TTS vs. ACS was derived in an all-comer ACS cohort⁵

Points	Criteria
25	Female sex
24	Emotional trigger
13	Physical trigger
12	Absence of ST depression (except in lead aVR)
11	Psychiatric disorders
9	Neurological disorders
6	QTc prolongation

Patients with a score of ≥50 have 95% specificity for a TTS diagnosis and those with a score ≤31 have a diagnosis of ACS with 95% specificity.

4. Arrhythmias

Approximately 7% of patients develop atrial fibrillation and 3% ventricular thrombi or embolic events.²⁶ The presence of an apical-type TTS, elevated white blood cell count on presentation, LV ejection fraction (LVEF) ≤ 30%, and previous vascular disease are independent predictors of thrombo-embolic events.²⁷

Ventricular arrhythmias occur in 10% of cases during both the acute and convalescent phases and are associated with reduced survival.²⁸ The presence of monomorphic ventricular tachycardia (VT) appears to be stronger at predicting mortality compared to polymorphic VT.²⁸ The mechanism of VT in TTS patients may be related to abnormal automaticity of non-pacemaker myocytes in the setting of catecholamine excess, QT prolongation, triggered electrical activity, or due to re-entry mechanisms.²⁹

5. Cardiac mechanisms

The underlying mechanism(s) of TTS remains elusive although various hypotheses have been postulated including sympathetic nervous system (SNS) overstimulation,^7,8,30,31 catecholamine secretion,^31–33 adrenergic receptor modulation,^31,34–37 coronary epicardial or microvascular coronary dysfunction^38,39 or spasm,^40–43 endothelial dysfunction,^44–48 structural and functional central nervous system (CNS) alterations,^{3,6,20,49–52} hormone-mediated changes,^6,7,44,53,54 and genetic predispositions (Figure 3).^55–58

5.1 SNS over-stimulation

Stimulation of the SNS underpins a common pathway in the pathogenesis of TTS given the frequency of preceding stressful physical or emotional triggers before the onset of cardiac signs or symptoms (Figure 3).^3,6,44,59 In a multi-centre study of 324 TTS patients, a triggering event was identified in 76% of females and 86% of males.⁷ An emotional trigger was more common in females, whereas physical ones were more common in men; however, in a considerable proportion of TTS patients, no overt trigger factor was identified.^{7,8,41,60–62} Physical triggers were also more common in Japanese TTS patients compared to their European counterparts.¹³

Further supporting the anecdotal evidence of an association between physical or psychological stressors of increased SNS activity and TTS are case series where the phenotype is re-capitulated by the intravenous (i.v.) administration of catecholamines or β-agonists in human and animal models.^30,31 Similarly, TTS has also been reported after i.v. or intramuscular (i.m.) administration of epinephrine in patients treated for anaphylaxis.^63–65

Primary endocrine disorders, including pheochromocytomas and paragangliomas, resulting in periodic excess catecholamine release and consequent excess SNS activation, can also precipitate a phenocopy of TTS further supporting the concept of SNS overstimulation as a major mechanism.^66–68

Heart rate variability, a surrogate of autonomic tone, is markedly reduced during the acute phase of TTS and slowly recovers in the subacute phase and normalizes within 3 months.⁶⁹ This phenomenon suggests marked SNS activation with simultaneous cardiac parasympathetic depression during the acute phase and recovery of autonomic modulation over time. Microneurography assessments of muscle SNS activity demonstrate significantly increased SNS activity and decreased spontaneous baroreflex control in TTS cohorts compared with age, sex, and LVEF-matched controls.⁷⁰

Using a combination of ¹²³I-meta-iodobenzylguanidine (MIBG) scanning and ^99m-Tc methoxy-isobutyl isonitrile single-photon emission computed tomography (SPECT) imaging, TTS patients demonstrate a decreased heart : mediastinum ratio and increased cardiac washout of radio-isotope on MIBG imaging together with only mild alterations in myocardial perfusion on SPECT imaging suggesting functional alterations in central SNS neurotransmission.⁷¹

However, experiencing physical or emotional stressor is ubiquitous in life and SNS over-stimulation in isolation does not explain why only a relatively small number of individuals exposed to such stressors develop TTS. Similarly, although endocrine disorders such as pheochromocytoma are rare, TTS itself in this population is infrequent despite the periodic surge of catecholamines.⁴⁴ These observations suggest that some individuals may have a particular predisposition to the development of TTS in the context of a catecholamine surge and that other pathophysiological mechanisms must also contribute.⁷²

5.2 Catecholamine secretion

Catecholamines appear to play a cardinal role in TTS. Compared to patients with MI, TTS patients have higher levels of circulating plasma epinephrine, norepinephrine, and dopamine, reinforcing the role of an increased SNS stimulation and catecholamine secretion as a major mechanism (Figure 3).³² Considering that the plasma half-life of epinephrine is only a few minutes and the inevitable time delay from symptom onset to medical presentation, the actual peak of catecholamine levels that the myocardium is exposed to, may be considerably higher than values measured retrospectively after an acute triggering event.³⁷ Patients with TTS also have elevated plasma epinephrine levels in the subacute phase and over follow-up when compared to controls.³³ These findings of elevated plasma levels of catecholamine or their metabolites in patients with TTS have been replicated by others and are corroborated by an increase in circulating norepinephrine levels that occur in patients with subarachnoid haemorrhage developing TTS.^32,73,74

Murine models of TTS have also demonstrated the classical TTS cardiac phenotype after infusion of high doses of epinephrine or isoprenaline.³¹ Similarly, histopathological findings in TTS include the presence of contraction band necrosis and mononuclear infiltration similar to those found in diseases associated with extreme catecholamine surges including subarachnoid haemorrhage or status epilepticus.^75,76

In induced pluripotent stem cell (iPSC)-derived cardiomyocytes from TTS patients, isoprenaline stimulation significantly increases NR4A1 (a cardiac stress-related gene) expression compared to controls, a feature that reverses after catecholamine washout.⁷⁷ Furthermore, β₁-adrenoceptor (AR) signalling and catecholamine-dependent dynamic range cyclic-AMP levels are increased in iPSC-derived cardiomyocytes obtained from TTS patients resulting in reduced calcium decay during catecholamine-induced stress.⁷⁷ These cellular models also demonstrate altered expression of lipid transporters and increased intracellular lipid accumulation.⁷⁷ Increased catecholamine levels may also activate the phosphatidylinositol-3-kinase pathway thereby increasing protein kinase B phosphorylation signalling which has vital roles in myocardial cell survival.^50,78

Despite the convincing findings above, there remains a disparity in the literature, with several studies failing to consistently demonstrate significantly elevated circulating catecholamine levels during the acute phase of TTS.^40,79–81 In one study comparing TTS and ST-segment elevation myocardial infarction (STEMI) populations, plasma concentrations of metanephrine and normetanephrine did not differ between cohorts nor were 24 h urinary catecholamine levels elevated in TTS patients.⁶⁰ These contradictory findings were supported by another small study of 33 TTS patients, where plasma epinephrine and metanephrine levels were normal in a majority and plasma norepinephrine or normetanephrine levels were also normal or only mildly elevated.⁶² These negative findings were reinforced by the SMINC-2 study, which did not demonstrate elevated plasma metanephrine or normetanephrine levels in 86% of TTS patients in both the acute and subacute settings, concluding that excess plasma catecholamines may not be a major cause of TTS.⁸²

¹²³I-MIBG imaging shows a reduction in myocardial activity at the mid-LV and apex in TTS, which are associated with alterations in myocardial glucose metabolism in the same regions on SPECT imaging, despite an absence of elevated systemic circulating catecholamines.^4,80

These negative findings challenge the previous central dogma that TTS is unequivocally related to a massively activated SNS and consequent systemic catecholamine surge. A plausible explanation for this failure to demonstrate consistent systemic catecholamine release may be due to (1) a more localized release of catecholamines into the relevant synaptic clefts with rapid re-uptake and/or (2) the fact that the initial triggering catecholamine surge may have been missed in some studies due to late presentation and/or plasma and urine collection.

Further aspects need consideration: catecholamine biosynthetic enzymes are expressed in myocardial cells in animals and humans. Almost a third of cardiac adrenaline is synthesized by the heart itself through specific enzymes with evidence of a differential localization of phenylethanolamine n-methyltransferase, an adrenaline biosynthetic enzyme, in the myocardium.^83,84 Notably, these catecholamine-producing enzymes are predominantly left-sided and have a differential distribution from base to apex of the LV myocardium.⁸⁴ Thus, there remains the possibility of a localized myocardial catecholamine excess release as a major mechanism in TTS in the absence of elevated systemic catecholamines.⁸⁵ Supporting this hypothesis, are evidence of elevated norepinephrine levels in the coronary sinus compared to the aorta during the acute phase of TTS.⁷⁹

This localized catecholamine release through autocrine or paracrine mechanisms from intrinsic catecholamine stores may overload adrenergic receptors in specific cardiac regions and result in regional myocardial dysfunction as occurs in TTS. Further supportive evidence of an autocrine mechanism of cardiac adrenaline secretion includes a study demonstrating high cardiac adrenaline concentrations despite chemical or surgical denervation.⁸⁶ This localized catecholamine production and storage may also account for the differential myocardial activation and regional abnormalities demonstrated in TTS.

5.3 Adrenergic receptor modulation

The regional abnormalities identified in TTS remain unexplained given a potential systemic SNS trigger. Some studies have hypothesized that TTS may be in part related to alterations in the distribution of myocardial adrenergic receptors given the observed regional alterations in myocardial contractility.

Although a potential systemic catecholamine surge in TTS patients results in positive chrono- and inotropy, this is offset by an associated supply–demand oxygen mismatch creating regions of myocyte hypoxia and predisposing to cellular apoptosis.⁸⁷ Catecholamines released from pre-synaptic sympathetic nerve terminals innervating the myocardium are also thought to be more toxic than systemic circulating catecholamines due to a cyclic AMP-mediated calcium excess.^50,88

Other proposed mechanisms to explain regional variations in myocardial function include ligand or stimulus-directed trafficking, where there is a switch in myocardial epinephrine signalling from the pleiotropic β₂-AR canonical stimulatory G-protein (Gs) cardio-stimulant pathway to an inhibitory G-protein (Gi) cardio-depressor pathway (Figure 3).^31,34,35,89

The LV myocardium contains both β₁- and β₂-AR in a ratio of 4:1.^37,90 Activation of these AR provides increased inotropy and lusitropy.³⁷ Supra-normal levels of epinephrine result in an intense activation of the β₁-AR and β₂-AR Gs-protein pathway and therefore induces a switch from Gs-protein adenyl-cyclase pathway to β₂-AR Gi protein signalling.^91–93 The Gi signalling switch is protective and provides anti-apoptotic and anti-arrhythmic effects, thereby shielding the LV myocardium from the catecholamine surge.³¹ Gi protein signalling in turn activates the p38 MAP-kinase pathway, which has a negative inotropic action possibly through inhibition of L-type calcium channel currents and upregulation of sodium–calcium ion exchangers to reduce myofilament sensitivity.^{37,78,89,94–97}

However, the altered cellular signalling, although biologically plausible does not explain the distinct regional disparities in LV function seen in TTS. Animal studies demonstrated a regional variation in the density of SNS endings across the LV with a 40% increase in the basal compared to apical LV.³⁶ Hypothetically, the presence of fewer SNS endings in the apical LV might be compensated for by an increased density of β₂-AR in the apical myocardium.³⁷ This concept has been supported in animal models involving radio-ligand binding displacement assays demonstrating an increased β₂:β₁ AR ratio in the apex of the heart.³¹ This regional heterogeneity in β₂-AR density from base to apex of the LV may explain the disparity in response of the myocardium to elevated local and/or systemic catecholamine surges induced by physical or emotional stress and explain the onset of diverse TTS phenotypes.³⁷

Biased epinephrine agonism for β₂-AR Gs signalling at normal concentration and Gi signalling for supra-normal concentrations of catecholamines therefore is likely to play an important role in TTS and together with the β₂-AR gradient from LV base to apex may explain the cardio-inhibitory response, the regional diversity in LV systolic dysfunction and the absence of a major infarction in spite of catecholamine excess.³¹ This central role of Gi pathway signalling in TTS is reinforced by the inactivation of the Gi pathway by pertussis toxin, which prevents the onset of the TTS phenotype in rat models.³¹ The inability of norepinephrine to induce a TTS phenotype in animal models also supports the concept that TTS is not mediated by myocardial β₁-AR- or α₁-AR-induced vasoconstriction.³¹

Although alterations in G-protein signalling in the presence of local or systemic catecholamine surge may account in part for the pathophysiology of TTS, it is far from the complete picture. Indeed, the above findings do not explain how different variants of TTS can occur in the same individual over time nor the presence of atypical or reverse TTS phenotpyes.⁹⁸ Furthermore, plasma epinephrine levels are demonstrably higher in subarachnoid haemorrhage patients with reverse TTS presentations compared to those with classical apical TTS, a fact that goes against the apical epinephrine-induced switch in signal trafficking hypothesis (Figure 3).⁹⁹

Figure 3

Potential mechanism underlying the pathophysiology of TTS.

5.4 Coronary epicardial or microvascular spasm

The identification of CAD is not mutually exclusive to a diagnosis of TTS. Most patients, however, have smooth, unobstructed coronaries, or only mild non-obstructive plaques diagnosed on angiography. Initially, it had been hypothesized that TTS might result from acute plaque rupture with spontaneous recanalization and/or spontaneous dissipation of an occluding thrombus by medical anti-platelet therapy prior to urgent angiography. However, this concept is not compatible with the obvious mismatch of coronary anatomical findings and RWMA. Furthermore, intra-coronary (i.c.) imaging studies have failed to demonstrate microthrombi or evidence of acute plaque rupture in the left anterior descending coronary artery of TTS patients using optical coherence tomography.¹⁰⁰

More recently, the concept of definitive myocardial infarction with non-obstructive coronary arteries (MINOCA) has come to light. One study demonstrated a differing patient characteristic profile in TTS compared to MINOCA patients with a reported higher prevalence of female sex, older age, and comorbid psychiatric disorders in TTS populations.¹⁰¹ TTS patients also suffered more in-hospital mortality or complications than MINOCA patients suggesting a more hostile acute phase.¹⁰¹

The phenomenon of transient epicardial coronary spasm has also been postulated as a cause for TTS as myocardial catecholamine excess and coronary spasm are thought to be complementary mechanisms.¹⁰² In one study of 30 patients with STEMI presentations, subsequently identified as TTS without significant CAD, 3 had multi-vessel coronary spasm at angiography.⁴⁰ The presence of single or multi-vessel epicardial coronary spasm at baseline or during i.c. acetylcholine provocation has been noted in a considerable number of TTS patients.^41–43 However, spontaneous or provoked epicardial coronary spasm is not ubiquitous in TTS patients and a majority do not demonstrate any coronary spasm during angiography and lack spasm inducibility on provocation testing.³ In animal models, norepinephrine is unable to induce a TTS phenotype, whilst epinephrine and isoprenaline can, suggesting that coronary spasm is unlikely to be a major mechanism of disease as norepinephrine unlike epinephrine hardly impacts β₁-AR or β₂-AR Gi signalling.³¹

5.5 Endothelial dysfunction

One of the more interesting findings is that TTS patients rarely develop marked troponin elevations as seen in acute myocarditis or MI. They also are much less likely to develop any late gadolinium enhancement or fibrosis on cMRI imaging during the acute or convalescent phase of the illness.³⁸ This suggests that there must be both vascular and myocardial alterations resembling ‘myocardial stunning’ or ‘hibernation’ during an acute episode of TTS.

It has been noted that female patients with TTS are more likely to have a history of Raynaud’s disease or migraine supporting a role for endothelial dysfunction in TTS.^44,45 Endothelial dysfunction is also associated with cardiovascular risk factors including hypertension, hyperlipidaemia, smoking, and other comorbidities often found in TTS patients.^6,50,103 Stress has indeed been demonstrated to affect flow-mediated endothelium-dependent vasodilatation in vascular beds such as the radial artery.^46,47 TTS patients also exhibit transiently impaired brachial artery flow-mediated vasodilatation compared to controls or those with MI.⁴⁸

Females with acute TTS demonstrate higher mean Thrombolysis in Myocardial Infarction (TIMI) frame counts at angiography in one or more major epicardial coronary arteries compared to matched controls despite an absence of obstructive CAD.¹⁰⁴ In another study of consecutive TTS patients, an abnormal TIMI myocardial perfusion grade at angiography was noted in 69% with higher peak troponin levels in those with abnormal TPMG.¹⁰⁵ This effect is reflective of abnormal microvascular perfusion. This is further supported by evidence of decreased coronary flow velocity reserve and shorter diastolic velocity deceleration time in acute TTS which improves on follow-up 3 weeks later.¹⁰⁶ Similarly, there is evidence of reduced microvascular vasodilator capacity with a transient impairment in the coronary flow reserve in acute TTS as assessed non-invasively by Doppler TTE.¹⁰⁷

Contrast echo in TTS and STEMI patients demonstrates similar baseline findings, but an adenosine infusion reverses parameters such as the contrast score index, endocardial length of contrast defect, myocardial wall motion score index, and endocardial length of contractile dysfunction in TTS patients but not in the STEMI cohort.¹⁰⁸ This suggests that the TTS changes are likely secondary to an element of microvascular dysfunction and is further supported by abnormalities evidenced on radio-isotope nuclear myocardial SPECT imaging.³⁹

During mental stress, endothelin (ET) plasma levels are also increased and stress-induced endothelial dysfunction can be inhibited by ET-A receptor anatgonists (Figure 3).⁴⁶ ET-A receptors predominate in the coronary microcirculation and ET has potent, long-lasting vasoconstrictor effects suggesting that prolonged ET-induced microvascular contraction and/or spasm may be relevant.¹⁰⁹ Microvascular spasm may be caused by increased ET-1 levels, which affect voltage-dependent calcium channels and calcium influx intracellularly.²⁰ Furthermore, the microRNA, miR-125a-5p, a regulator of ET-1 expression, is down-regulated in TTS leading to increased plasma levels of the peptide.²⁰ This could explain the reduced microvascular blood flow and coronary flow reserve.^20,110,111 Endomyocardial biopsy samples of TTS patients demonstrates microvascular endothelial cell apoptosis.¹¹²

Ultimately, coronary endothelial dysfunction may result in episodes of transient myocardial ischaemia and associated stunning without significant MI and yield a picture of reversible LV systolic dysfunction.⁵⁰

Again, the literature is contradictory on the endothelial dysfunction hypothesis as rodent models of TTS fail to exhibit altered myocardial perfusion in dysfunctional myocardial segments and therefore more studies are needed in this area.¹¹³

5.6 Structural and functional alterations in the CNS

The strong association of physical or emotional stress contemporaneous with the onset of TTS suggests an important role for the CNS in its pathogenesis. Case–control studies identify a higher prevalence of comorbid psychiatric or neurological disorders in TTS patients and the presence of these comorbidities is independently associated with TTS recurrence.^6,49 Depressed individuals may have impaired CNS norepinephrine reuptake and are more likely to be on serotonin-norepinephrine reuptake inhibitors, which affect local norepinephrine reuptake from the synaptic cleft thereby increasing local catecholamine levels.^114,115 This may contribute to the increased susceptibility to myocardial stunning and contractile impairment in these patients.

Further evidence for a pivotal role of the brain-heart axis in TTS includes a significant upregulation of microRNAs associated with stress, depression, and other neuropsychiatric disorders.²⁰ MicroRNAs are highly conserved, non-coding post-transcriptional regulators of cellular apoptosis, proliferation, and differentiation.^20,116 In a comparative study of TTS patients with healthy controls and STEMI patients, there was evidence of dysregulation of miR-16 and miR-26a microRNAs, both of which are strongly associated with stress and depression-related microRNA.²⁰ This is concordant with findings of elevated levels of these microRNAs in humans after an acute stress stimulus and in stressed mice models.^117,118

Cerebral autonomic centres are known to process sympathetic activation.¹¹⁹ Specific areas of the CNS including the spinal cord, brainstem, reticular and limbic system, and cerebral neocortex are involved in modulating the ‘stress response’.⁵⁰ Emotional or physical stress activates specific cerebral regions including the amygdala, hypothalamus, and cingulate gyrus, which in turn innervates the locus coeruleus (the principal site of norepinephrine synthesis in the brain). Activation of these CNS regions via noradrenergic neurons further stimulates the hypothalamic–pituitary–adrenal axis and results in increased secretion of norepinephrine and epinephrine from adrenal medullary chromaffin cells. This in turn facilitates higher circulating catecholamine levels with stress to enable the physiological ‘fight or flight’ response.^50,120

Evidence of the important role of the CNS in TTS is highlighted by increased cerebral blood flow, a surrogate of activity, in specific areas with regional cerebral activation.⁵⁰ This is evidenced using SPECT imaging in the acute phase of TTS demonstrating elevated cerebral blood flow in the hippocampus, basal ganglia, and brainstem and a reduction in the prefrontal cortex.⁵¹ These changes remain persistent, albeit attenuated, in the chronic phase of the disease process. Some stressors may have a global impact on cerebral blood flow and metabolism, whereas others may only induce a more targeted, regional effect.¹²¹ The exogenous release of epinephrine from the adrenal medulla together with endogenous norepinephrine in the brain act as a common pathway in mediating altered cerebral perfusion and energy metabolism.¹²¹

Notably, injury to the limbic system or insular cortex in acute ischaemic stroke has been associated with the onset of TTS.¹⁸ Given the important role of the amygdala in regulating the response to stressors, ¹⁸F-FDG positron emission tomography combined with cardiac tomography imaging identified higher amygdalar activity in TTS patients at baseline preceding the onset of TTS and a shorter duration between imaging and subsequent TTS onset in those with greater amygdalar activity at baseline.¹²² This heightened amygdalar activity may be a risk factor for the development of TTS in susceptible individuals by potentiating the neurohormonal and physiological effects of stressors.^122,123

Cross-sectional studies have also demonstrated morphological differences in cerebral MRI with reduced anteroventral insular and cingulate cortical thickness, amygdala grey matter volume, and reduced structural connectivity in the limbic system in TTS patients compared to controls.⁵² These anatomical differences are important due to the role of the limbic system in emotional processing through the autonomic nervous system and may result in altered neural signalling.⁵² These anatomical findings were replicated in another study utilizing cerebral MRI during acute TTS, demonstrating a lower grey matter cerebral volume in the right middle frontal gyrus and right subcallosal cortex in TTS patients compared to controls.¹²⁴ There was also significantly reduced functional connectivity in TTS patients, in particular in the right insula, which is involved with autonomic tone.¹²⁴ However, whether these changes are causative or a result of TTS (reverse causality) remains to be clarified.

In summary, structural anatomical abnormalities in tandem with functional regional alterations in cerebral perfusion and metabolism appear to play an important role in TTS. This complex interplay between the limbic system, the locus coeruleus, and the adrenal axis results in neurophysiological changes, abnormal stimulus processing and hormonal changes possibly resulting in the onset of TTS.^123,125

Further work is needed in this developing area, utilizing advanced cerebral anatomical and functional imaging together with biomarker profiling to understand the possible causation and association of CNS changes with TTS and whether these CNS changes are secondary phenomena to the triggering stimulus and nervous system activation.

5.7 Hormone mediated changes—oestrogens

Ninety per cent of TTS cases involve women, almost exclusively after the menopause.^6,7 Given this significant sex bias and that the incidence of TTS increases considerably in the post-menopausal years, there is a suggestion that a biological reduction in oestrogen levels may predispose to TTS and play a pivotal role in its pathogenesis.⁴⁴ This hypothesis would not be incongruous with the catecholamine theory, as females have lower basal plasma catecholamine levels than men due to reduced catecholamine synthesis, increased degradation and reduced basal level release.^37,126 It is therefore possible that females have a greater ability to increase their catecholamine release during sudden acute stress, although this theory has not been substantiated yet.

Oestrogen receptors are expressed in cardiovascular cells and control vasomotor tone through upregulation of endothelial nitric oxide synthase.^50,127,128 Exogenous oestrogen administration has a direct coronary vasodilatory effect through endothelium-dependent and independent mechanisms.¹²⁹ Oestradiol is also protective against catecholamine excess in iPSC models.^53,130 Oestrogen removal increases β₁-AR and L-type Ca²⁺ channel proteins and reduces Na⁺–Ca²⁺ exchange proteins.¹³¹ Accordingly, in ovariectomized rats, oestrogen supplementation reduces c-fos mRNA expression (a marker of cellular activation) in the LV, adrenal glands, and paraventricular hypothalamic nucleus and improves stress-induced cardiac dysfunction by altering the impact of the hypothalamic-sympathetic-adrenal outflow on target organs.⁵⁴

Oestrogen supplementation also up-regulates heat shock protein 70 and atrial natriuretic peptide, both of which are cardioprotective.^50,54 The β₁-AR is oestrogen receptor-dependent and is suppressed by oestrogen supplementation which in turn, reduces the magnitude of myocardial necrosis from sympathetic overactivity from β₁-AR mediated myocardial ischaemia.¹³² Oestrogen in women may also alter the ratio of β₁-AR : β₂-AR signalling in preference of the β₂-AR Gi signalling, which can provide a protective benefit in the setting of increased catecholamine levels.³⁷ This is highlighted by studies demonstrating norepinephrine causing a lower degree of forearm vasoconstriction in women compared to men through greater β₂-AR sensitivity.¹³³ Furthermore, TTS has been averted in rat models by oestrogen supplementation or with simultaneous α- and β-AR blockers.^54,134

Contradicting the oestrogen theory are case–control studies where TTS patients demonstrate no significant differences in parity, history of oophorectomy, and years since the menopause compared to population controls or STEMI cohorts and in fact intercurrent HRT usage was greater in TTS cohorts than controls.⁴⁴ This suggests that despite the post-menopausal female preponderance, oestrogen deficiency is unlikely to be the sole or primary trigger for TTS.

Overall, these studies suggest that sex hormone profile alone is unlikely to solely account for the mechanism of TTS.⁴⁴ Despite the clear sex bias in TTS presentations, a clear central pathophysiological role of oestrogen in TTS has not been demonstrated yet but it may contribute to a predisposition to the syndrome.

5.8 Genetic predisposition

It is unclear whether there is an underlying genetic mutation that causes or predisposes individuals to TTS. Some patients with TTS develop a dilated cardiomyopathy (DCM) phenocopy due to the lack of recoverability in LVEF in the convalescence phase of the condition. However, there are case reports of familial cases of TTS suggesting a possible genetic mechanism in at least some patients.^55,56,135 Variants in genes including BAG3, ADRB1, GRK5, and others have been associated with TTS.^57,58

Given the central role of catecholamine release and AR signalling, studies have also suggested a central role in AR polymorphism in TTS with respect to β₁-AR and β₂-AR that are associated with increased troponin release and lower LVEF in patients with subarachnoid haemorrhage.¹³⁶ Whole-exome gene sequencing studies of TTS cohorts, however, have failed to identify an excess of pathogenic variants in TTS patients compared to control cohorts and demonstrated a similar frequency of functional adrenergic polymorphisms between TTS and controls.¹³⁷

Overall, there is no clear systematic pattern of Mendelian inheritance demonstrated in TTS when compared to genetic forms of DCM. It is likely, just like other risk factors, genetics may play a role in increased susceptibility in combination with other epigenetic factors.

6. Prognosis

In general, TTS is a benign, transient self-limiting condition with spontaneous recovery of LVEF within days to weeks in a majority of patients.^{8,32,33,41,42,72,138,139}

However, more recently there is an increasing awareness that a TTS diagnosis does not confer as favourable a prognosis as previously thought and that there is an association with increased short-term and longer-term mortality.¹⁴⁰ The condition can be life-threatening in around 1 in 20 patients, particularly when presenting with OOHCA, malignant ventricular arrhythmias, or cardiogenic shock.^8,6,9,61 In one meta-analysis of case reports comprising endogenous or exogenous catecholamine-induced TTS, as many as 68% of patients had complications including heart failure, cardiogenic shock, arrhythmia, thrombo-embolism, and respiratory failure.⁹ In this study, mechanical ventilation was required in 26% of cases and in over 20% of patients, mechanical circulatory support was needed, although this considerable rate of adverse events is likely to reflect publication bias.⁹

The reported literature rates of all-cause mortality, however, are as high as 4–8%.^11,141 The 5-year mortality rate is 17%, but no differences in cardiovascular mortality were found between age- and sex-matched populations.⁷²

In a minority, RWMA may be associated with systolic anterior motion of the mitral valve or chordae and consequent mid-cavity or LV outflow tract obstruction (LVOTO). This can be associated with hypotension, particularly if aggravated by inotropes.¹⁴² In the setting of LVOTO, small doses of an i.v. β-blockers or mechanical circulatory support with extracorporeal membrane oxygenation (ECMO) may be beneficial in select circumstances to treat dynamic intra-cavity obstruction and allow stability until myocardial recovery.¹⁴² Rarely, the marked LV wall stress in apical ballooning may lead to free wall rupture of the LV as demonstrated in case reports.^143,144

Physical triggers appear to be an independent predictor of mortality in TTS patients.^140,145 Male patients were noted to have higher mortality rates than females, although this has been attributed to a higher rate of comorbid disease and critical illness on presentation.¹⁴⁶ Patients with a global LV ballooning pattern had more frequent adverse events compared to patients with isolated apical ballooning.⁹ The InterTAK Prognostic Score represents a valuable clinical tool with high sensitivity and specificity to predict outcomes.¹⁴⁷

The recurrence rate of TTS appears to be 1.8% per patient-year spanning from 25 days to over 9 years post-index presentation.⁶ In another study, the recurrence rate of TTS was 11.4% over 4 years from index presentation with a recurrence of symptoms, most notably chest pain over follow-up.⁷² In meta-analyses, TTS recurrence is reportedly higher involving 17% of patients, albeit exclusively in those with paraganglioma-induced TTS.⁹ Notably, comorbid psychiatric and neurological disorders are independent predictors of recurrence. Interestingly, in 35% of patients with recurrent TTS, the phenotype was with a different pattern of LV ballooning at the recurrent presentation compared to the index event.⁴⁹

Studies have also suggested that patients with TTS develop a subclinical cardiac phenotype despite overt LVEF recovery after the acute event and in some cases have persistent myocardial oedema on CMR months post-presentation.^148,149 This suggests the possibility of ongoing cellular and humoral inflammatory cascades as evidenced by higher T2* values in myocardial segments during the acute phase of TTS identified by ultra-small superparamagnetic iron oxide particles-enhanced cMRI.¹⁵⁰ This is indicative of a macrophage-mediated cellular inflammatory process. This inflammatory response is also associated with raised serum interleukin-6 and CXCL1 (growth-regulated protein) chemokine levels compared to control populations at presentation and over follow-up, suggesting a systemic humoral inflammatory component too in the TTS phenotype and may explain the persisting signs of myocardial dysfunction or alterations in the subacute or chronic phases of the condition.¹⁵⁰

7. Potential consequences for therapeutics

Given the association of TTS with neuropsychiatric disorders, the role of therapeutic strategies focusing on the CNS has been postulated. Cognitive behavioural therapy has been associated with a reduced amygdalar activity, which hypothetically could help in preventing TTS.¹⁵¹ This is an unexplored area at present.

There remains uncertainty on whether medical therapy after acute presentation with TTS alters long-term outcomes. Given the postulated role of localized or systemic catecholamine release, the adjunctive use of β-blockers at first glance may appear beneficial in minimizing the impact of catecholamines on β₂-AR. However, it is established that some patients develop TTS despite being on β-blockers pre-morbidly and a considerable proportion develop recurrent TTS despite still being on β-blockers.⁶ Further studies have failed to identify a demonstrable benefit of routine β-blocker initiation in TTS patients with respect to in-hospital mortality.¹⁵²

Large registries have demonstrated that angiotensin converting enzyme inhibitors or angiotensin receptor blockers improve survival in patients with TTS and impaired LV function and have been associated with reduced rates of recurrence.^6,153

ET-A antagonists may help promote vasodilatation during the TTS acute phase, but this hypothesis remains to be tested. In reality, trials need to be developed in patients with TTS to evaluate the role of ET receptor antagonists and α-1 blockers in patients with TTS.

Previous studies have suggested the persistence of subclinical or overt myocardial abnormalities beyond the acute phase of TTS and a potential association with a cellular and humoral-based inflammatory response.¹⁵⁰ The role of anti-inflammatory medications to alter the acute and subacute inflammatory response in TTS and thereby prevent the onset of a heart failure phenotype and adverse long-term outcomes is an area of interest but requires systematic analysis by randomized trials.

In general, the management of heart failure in TTS follows that of algorithm evidence-based conventional heart failure medication used in other patients with ischaemic or non-ischaemic LV systolic dysfunction. There is no consensus on the duration of guideline-directed heart failure medical therapy in TTS patients, but in general these are continued until there is a recovery in LVEF.

Ventricular thrombi are identified in >1% of patients presenting with TTS.⁶ Anticoagulation is occasionally used if patients have an increased risk of thrombo-embolism or have evidence of thrombus in situ to prevent cerebral or systemic thrombo-embolism. Fortunately, LV thrombus formation has a low incidence in TTS cohorts although being more common in females with an apical ballooning pattern of TTS and in those with higher troponin levels. In one study, 3 months of anticoagulation successfully resolved all LV apical thrombi.¹⁵⁴

Although most patients with TTS do not have any evidence of flow-limiting CAD, aspirin is commonly prescribed at discharge with the aim of reducing the risk of future thrombotic events,¹⁵⁵ although there is no evidence that this is associated with a reduced risk of major adverse cardiovascular events at 30-day and 5-year follow-up.¹⁵⁶ These conclusions are supported by another meta-analysis, where anti-platelet therapy in TTS patients did not result in a significant difference in rates of TTS recurrence or progression of CAD or subsequent cerebrovascular events.¹⁵⁷ Thus, routine prescription of anti-platelet drugs should be avoided unless CAD is also present on angiography.

Some patients with TTS present with or develop cardiogenic shock during the course of their admission. This may require inotropic or vasopressor support to achieve adequate blood pressures to maintain end-organ perfusion. The usage of catecholamines in patients with TTS is associated with reduced in-hospital medium- and long-term survival.¹⁵⁸ Case series have suggested that usage of the calcium-sensitizer, levosimendan, however, is safe in TTS patients.¹⁵⁹

β-Blockers such as esmolol infusions, may be useful in a subset of patients that develop acute LVOTO with or without cardiogenic shock as evidenced by its use in patients with hypertrophic cardiomyopathy and LVOTO.¹⁶⁰ In those with LVOTO, caution is recommended with respect to using vasodilator drugs or positive inotropes as they can exacerbate LVOTO. In select refractory cases, phenylephrine may be used to augment vasoconstriction and increase afterload.¹⁶¹ ECMO is an alternative option for mechanical circulatory support in some patients with TTS and cardiogenic shock.^162,163

Ghadri et al. provide a pragmatic and useful treatment algorithm (Figure 4) for patients with TTS.¹⁶⁴

Figure 4

Management of TTS. Courtesy Eur Heart J, Vol. 39, June 2018; pages 2047–2062 (https://doi.org/10.1093/eurheartj/ehy077).¹⁶⁴

8. Conclusion

Although the precise underlying mechanism of TTS remains elusive, it is likely that it is driven by a complex interplay between the different pathophysiological mechanisms discussed above. Physical or emotional stressors activate a structurally and functionally altered neurological CNS substrate, which in turn triggers a complex cascade of intrasynaptic, local, and/or systemic catecholamine excess, switched myocardial AR signalling and altered metabolism in concert with concomitant endothelial dysfunction.

An altered CNS together with direct catecholamine-induced cardiotoxicity results in regional myocardial wall motion abnormalities driven by altered density of autonomic innervation and distribution of adrenoceptors. Protective mechanisms in play to reduce the impact of this catecholamine surge and endothelial dysfunction may contribute to differences in activation to protect from myocardial apoptosis and necrosis.

Although there is an increasing awareness of TTS, considerably more research effort needs to be invested into understanding potential pathophysiological mechanisms and testing hypotheses to ascertain appropriate individualized treatment. Despite advances in cardiac imaging over the past decade, TTS remains underdiagnosed due to delays in echocardiography or a lack of availability in CMR imaging. Even with the availability of these imaging modalities, delays in angiography in non-STEMI TTS presentations mean that the condition is diagnosed later in the natural course of the disease, where serological or imaging diagnostic or prognostic markers and novel treatment strategies may be rendered less effective.

More research is needed to understand how these different complex mechanisms interact with each other and eventually lead to different TTS phenotypes. Larger clinical datasets with longer longitudinal follow-up are required to understand why some patients develop recurrent TTS while others do not, why there is a failure of recovery in LVEF in some, and whether prognostic guideline-directed heart failure medications truly improve short- or long-term outcomes. There is an increased interest in the coronary microcirculation with flow wires able to attribute numerical indices for microvascular resistance. Systematic usage of these in patients presenting with TTS and unobstructed coronaries on angiography may provide supportive evidence of the important role the coronary microcirculation plays in TTS cohorts.

There is also an unmet clinical need for randomized clinical trials using conventional heart failure medications and ET-A antagonists in the acute phase of TTS to determine their impact on the natural course of the condition and subsequent clinical outcomes. The role ofα-receptor or ET antagonists in the treatment of TTS is biologically plausible and can be determined through trial data. Further complete and comprehensive studies are needed with complete clinical, serological, metabolomic, and imaging profiles in TTS patients to help further appreciate underlying pathophysiological mechanisms and to fully appreciate the natural course of disease and the frequency of complications as current values are limited by observational data on smaller TTS cohorts.

Author notes

Conflict of interest: None declared.