The case
A 54-year-old postmenopausal woman presented to the emergency department with retrosternal chest pressure, nausea, and vomiting of 4 h duration. Her medical history included hypertension (treated with metoprolol 12.5 mg twice per day and ramipril 2.5 mg twice per day), hyperlipidemia (treated with atorvastatin 40 mg per day), and depression (treated with fluoxetine 40 mg twice per day). A few hours before symptom onset, she had witnessed an accident in which her sister sustained serious injuries. The patient's symptoms began while she was visiting her sister, who was in critical condition in the hospital.
On examination the patient was alert and oriented. Her heart rate was 65 bpm, blood pressure 138/71 mmHg, and oxygen saturation 98% while breathing room air. Cardiac examination revealed normal heart sounds. There was no jugular venous distention. The remainder of the examination was normal. No electrocardiographic abnormalities were noted initially; however, cardiac biomarkers were elevated—troponin T and creatine kinase MB fraction were 0.4 ng/ml and 10.1 ng/ml, respectively. An acute coronary syndrome was suspected. The patient was treated with sublingual nitroglycerin (0.4 mg), intravenous morphine (4.0 mg), and intravenous metoprolol (15.0 mg), with symptomatic relief.
The patient underwent urgent coronary angiography, which demonstrated mild coronary atherosclerosis without evidence of acute plaque rupture. However, left ventriculography revealed dilatation, ejection fraction of 48%, and akinesis of the mid-anterior, mid-septal, mid-inferior, and mid-lateral segments, and mild hypokinesis of the basal-anterior, basal-septal, and basal-inferior segments (Figure 1). The patient was diagnosed with stress-induced (Takotsubo) cardiomyopathy (apical-sparing variant) on the basis of the following criteria: cardiac biomarker elevation, regional wall motion abnormality involving multiple coronary territories, absence of obstructive coronary artery disease, and absence of alternative diagnoses.
Figure 1 | Left ventriculogram.
a | Diastole. b | Systole. These images demonstrate systolic dysfunction (ejection fraction 48%) with akinesis of the midventricular segments (arrows).
On the second day of hospitalization, the patient underwent PET imaging with 13N ammonia (12.09 mCi) and 11C hydroxyephedrine (16.68 mCi) as previously described.1 The 13N ammonia and corresponding 11C hydroxyephedrine images normalized to the highest count are shown in Figure 2. By visual analysis, tracer uptake was homogeneous throughout most of the left ventricular (LV) myocardium on the 13N ammonia images, indicating preserved perfusion. Tracer uptake was, however, reduced in the majority of the mid segments and several of the apical and basal segments on the 11C hydroxyephedrine images, suggesting sympathetic abnormalities. Absolute regional myocardial blood flow (MBF) and myocardial 11C hydroxyephedrine retention index were also calculated.1, 2 The hydroxyephedrine retention index was normalized to regional MBF to exclude flow-dependent changes in hydroxyephedrine uptake.3 In all segments, absolute regional MBF at rest was in the range (0.6–1.4 ml/min/g) reported for healthy individuals4 (Figure 3). However, the majority of segments exhibited reduced normalized regional hydroxyephedrine retention index, similar to levels reported in patients with NYHA class II and III heart failure1 (Figure 3). Lower hydroxyephedrine retention index values were noted in regions with contractile dysfunction, and were lowest in mid and basal segments and highest in apical segments (Figure 3).
Figure 2 | Short-axis images in the patient at 2 days and 6 weeks after presentation with stress-induced cardiomyopathy.
a | Myocardial blood flow (13N ammonia). b | 11C hydroxyephedrine PET.
Figure 3 | Physiological parameters calculated from the patient's PET imaging studies.
a | Absolute regional myocardial blood flow and b | normalized regional 11C hydroxyephedrine retention index at 2 days and 6 weeks after presentation.
Serial electrocardiograms revealed T-wave inversions in leads I, aVL, and V1–V3 as well as prolongation of the QTc interval (510 ms). Cardiac biomarkers decreased on serial measurements. Plasma and 24 h urine levels of fractionated metanephrines and catecholamines were normal. The patient was treated with aspirin (81 mg per day), metoprolol (12.5 mg twice per day), and ramipril (2.5 mg twice per day) and was discharged after 4 uneventful days in hospital.
At the 6 week follow-up visit, electrocardiography demonstrated resolution of the T-wave abnormalities and a normal QTc interval. Echocardiography showed complete resolution of LV systolic dysfunction. The calculated ejection fraction was 62%. PET 13N ammonia (17.63 mCi) and 11C hydroxyephedrine (14.17 mCi) was repeated (Figure 2). By visual analysis, there was no notable change in tracer uptake between the acute and follow-up 13N ammonia images, but the 11C hydroxyephedrine images showed higher tracer uptake throughout the LV myocardium at follow-up than had been noted at presentation. Furthermore, a mild, persistent inferior defect was noted on the follow-up 11C hydroxyephedrine image. Regional MBF and normalized hydroxyephedrine retention index were again calculated (Figure 3). The majority of myocardial segments exhibited higher normalized regional hydroxyephedrine retention index than at presentation, most of which were in the range previously reported in normal volunteers.1 The inferior wall did not exhibit consistently lower normalized hydroxyephedrine retention index when compared with other regions. Absolute regional MBF was again within the normal range for all segments. Two years after her admission to hospital, the patient remains well and has not experienced a recurrence of symptoms.
Discussion of diagnosis
Stress-induced cardiomyopathy—also known as Takotsubo cardiomyopathy, apical ballooning syndrome, or ampulla cardiomyopathy—is an acute, reversible condition characterized by LV systolic dysfunction generally involving the mid and apical segments.5 Associated electrocardiographic abnormalities suggest ischemia or myocardial injury, but occur in the absence of obstructive epicardial coronary artery disease. The majority of patients present after an acute emotional or physical stressor, which implicates a catecholamine surge in the pathophysiology.6, 7, 8 Stress-induced cardiomyopathy mimics an acute coronary syndrome and is estimated to account for approximately 1–2% of all presentations with symptoms of myocardial infarction.9 The in-hospital mortality rate is lower than that for myocardial infarction.9 Long-term survival is similar to that of an age-matched and gender-matched population.10 The diagnosis of stress-induced cardiomyopathy in the patient in this case was confirmed by the presence of all four Mayo Clinic diagnostic criteria: transient contractile dysfunction of the mid-LV segments with or without apical involvement, extending beyond a single epicardial vascular distribution; absence of obstructive coronary disease or angiographic evidence of acute plaque rupture; new electrocardiographic abnormalities (either ST-segment elevation, T-wave inversion, or both) or elevated cardiac troponin; and the absence of a pheochromocytoma or myocarditis.9
To our knowledge, this case is the first published report of PET imaging with 11C hydroxyephedrine in stress-induced cardiomyopathy. Although limited to a single patient, the PET findings are compatible with the proposed pathophysiological hypothesis focusing on direct myocardial effects of high catecholamine levels.8, 11
11C hydroxyephedrine is a norepinephrine analog that provides a measure of cardiac presynaptic sympathetic activity.1 Plasma hydroxyephedrine is transported into cardiac sympathetic nerve terminals by the uptake-1 mechanism.12, 13 Unlike norepinephrine, hydroxyephedrine is not metabolized by monoamine oxidase or catecholamine-O-methyltransferase enzymes. Vesicular storage of hydroxyephedrine occurs, but to a lesser extent than norepinephrine, owing to the higher lipophilicity of hydroxyephedrine. Myocardial retention of hydroxyephedrine, therefore, reflects a continuous release and reuptake of hydroxyephedrine in the presynaptic neuron.12
In this patient, the acutely reduced myocardial 11C hydroxyephedrine retention in segments with contractile dysfunction is compatible with the hypothesis of increased sympathetic activity, with associated increased norepinephrine release and competitive inhibition of 11C hydroxyephedrine reuptake. The higher 11C hydroxyephedrine retention at follow-up is indicative of the reversible nature of this effect. An alternative explanation for the acutely reduced 11C hydroxyephedrine uptake is transient dysfunction of the uptake-1 mechanism secondary to myocardial injury. Other imaging studies have demonstrated abnormal myocardial fatty acid and glucose metabolism using single photon emission computed tomography (SPECT) with 123iodine-
-methyl-p-iodophenyl pentadecanoic acid and PET with 18F-fluorodeoxyglucose, respectively, during the acute phase of presentation.7, 14 Impaired MBF could also contribute to these imaging findings, but was excluded in this case and other studies by demonstrating normal perfusion at the time of imaging.7, 14 It remains to be established whether the sympathetic abnormality is a primary event or an epiphenomenon in stress-induced cardiomyopathy.
Evidence to support the pathophysiological role of the sympathetic nervous system in stress-induced cardiomyopathy includes the temporal relationship between the presentation and preceding emotional or physical stress, elevated plasma catecholamine levels in some patients, and the inability to reproduce the syndrome in animal models of stress-induced cardiomyopathy in the presence of adrenergic blockage. Cardiac sympathetic imaging in this condition has been limited to SPECT studies using 123iodine metabenzylguanidine (MIBG), which have reported findings consistent with cardiac sympathetic abnormalities.7, 15, 16 Like the patient in this case, other individuals have exhibited persistent MIBG defects for several months and even at 1 year after presentation with stress-induced cardiomyopathy, indicating prolonged cardiac sympathetic abnormalities despite resolution of contractile dysfunction.16 SPECT with MIBG has two notable limitations when compared with 11C hydroxyephedrine. First, SPECT has lower spatial and temporal resolution than PET, which precludes absolute quantification of regional differences in tracer uptake by kinetic modeling.17 Second, interpretation of MIBG cardiac behavior is also confounded by a considerable amount of non-neuronal tissue uptake.1
Proposed mechanisms by which catecholamines might induce myocardial stunning include direct myocyte effects, such as intracellular calcium overload, or indirect effects, such as epicardial spasm, microvascular dysfunction, and hyperdynamic contractility with midventricular cavity obstruction.9 Epicardial spasm and cavity obstruction are not common findings, but impaired microvascular perfusion has been reported in up to two-thirds of patients at presentation.18 Notably, the acute cardiac sympathetic abnormality in this patient was not accompanied by reduced MBF, as measured by PET, which suggests mechanisms other than widespread microcirculatory dysfunction or resolution by the time of PET imaging. Regional differences in -adrenergic receptor density and sympathetic innervation have been proposed as potential explanations for the unique pattern of contractile dysfunction observed in stress-induced cardiomyopathy, although supporting data in humans are lacking.19 In addition, the apparent female predisposition for the disorder also suggests gender differences in myocardial catecholamine sensitivity.20 Lyon and colleagues have hypothesized that high levels of circulating epinephrine could trigger a switch in intracellular signal trafficking, from Gs protein to Gi protein via the 2-adrenoreceptor, which might be negatively inotropic.8 Although it is not possible to determine whether the observed reduction in hydroxyephedrine retention index in the patient in this case is a direct or indirect effect of catecholamines, our findings clearly demonstrate acute cardiac sympathetic abnormalities in stress-induced cardiomyopathy and provide impetus for further investigation.
Treatment and management
As with the patient in this case, individuals with stress-induced cardiomyopathy generally recover spontaneously. However, approximately 40% of patients develop congestive heart failure, and a minority (
10%) can present with acute cardiogenic shock or major hemodynamic compromise necessitating hemodynamic support.10 As catecholamines could have a causative role in stress-induced cardiomyopathy, therapy with epinephrine, inotropic agents such as dobutamine, or both might lead to worsening of the condition, although in the absence of clinical studies this is a theoretical concern. In the presence of LV outflow tract obstruction, as occurs in some patients, inotropic agents can also be deleterious. Currently, there is no proven therapy that augments recovery or improves outcomes in stress-induced cardiomyopathy. The potential role of catecholamines has prompted the empirical use of -blockers, with the goal of preventing recurrence. Theoretically, some -blockers could promote stimulus trafficking of 2-adrenergic receptors to Gi protein coupling, but whether this is clinically relevant is unknown.8 Finally, it remains to be established whether complete adrenergic blockade with a combined 
-blocker and -blocker, such as labetalol, is superior to -blockade alone.
Conclusions
We describe the use of PET with 11C hydroxyephedrine in a patient presenting with stress-induced cardiomyopathy. Acute and follow-up PET images indicate transient cardiac sympathetic abnormalities and support the pathophysiological role of the sympathetic nervous system in stress-induced cardiomyopathy. Further studies are needed to explore this hypothesis and the therapeutic implications.
